ceph/src/common/options.cc

   1 // -*- mode:C++; tab-width:8; c-basic-offset:2; indent-tabs-mode:t -*-
   2 // vim: ts=8 sw=2 smarttab
   3
   4 #include "acconfig.h"
   5 #include "options.h"
   6 #include "common/Formatter.h"
   7
   8 // Helpers for validators
   9 #include "include/stringify.h"
  10 #include "include/common_fwd.h"
  11 #include <boost/algorithm/string.hpp>
  12 #include <boost/lexical_cast.hpp>
  13 #include <regex>
  14
  15 // Definitions for enums
  16 #include "common/perf_counters.h"
  17
  18 // rbd feature validation
  19 #include "librbd/Features.h"
  20
  21 namespace {
  22 class printer : public boost::static_visitor<> {
  23   ostream& out;
  24 public:
  25   explicit printer(ostream& os)
  26     : out(os) {}
  27   template<typename T>
  28   void operator()(const T& v) const {
  29     out << v;
  30   }
  31   void operator()(boost::blank blank) const {
  32     return;
  33   }
  34   void operator()(bool v) const {
  35     out << (v ? "true" : "false");
  36   }
  37   void operator()(double v) const {
  38     out << std::fixed << v << std::defaultfloat;
  39   }
  40   void operator()(const Option::size_t& v) const {
  41     out << v.value;
  42   }
  43   void operator()(const std::chrono::seconds v) const {
  44     out << v.count();
  45   }
  46 };
  47 }
  48
  49 ostream& operator<<(ostream& os, const Option::value_t& v) {
  50   printer p{os};
  51   v.apply_visitor(p);
  52   return os;
  53 }
  54
  55 void Option::dump_value(const char *field_name,
  56     const Option::value_t &v, Formatter *f) const
  57 {
  58   if (boost::get<boost::blank>(&v)) {
  59     // This should be nil but Formatter doesn't allow it.
  60     f->dump_string(field_name, "");
  61     return;
  62   }
  63   switch (type) {
  64   case TYPE_INT:
  65     f->dump_int(field_name, boost::get<int64_t>(v)); break;
  66   case TYPE_UINT:
  67     f->dump_unsigned(field_name, boost::get<uint64_t>(v)); break;
  68   case TYPE_STR:
  69     f->dump_string(field_name, boost::get<std::string>(v)); break;
  70   case TYPE_FLOAT:
  71     f->dump_float(field_name, boost::get<double>(v)); break;
  72   case TYPE_BOOL:
  73     f->dump_bool(field_name, boost::get<bool>(v)); break;
  74   default:
  75     f->dump_stream(field_name) << v; break;
  76   }
  77 }
  78
  79 int Option::pre_validate(std::string *new_value, std::string *err) const
  80 {
  81   if (validator) {
  82     return validator(new_value, err);
  83   } else {
  84     return 0;
  85   }
  86 }
  87
  88 int Option::validate(const Option::value_t &new_value, std::string *err) const
  89 {
  90   // Generic validation: min
  91   if (!boost::get<boost::blank>(&(min))) {
  92     if (new_value < min) {
  93       std::ostringstream oss;
  94       oss << "Value '" << new_value << "' is below minimum " << min;
  95       *err = oss.str();
  96       return -EINVAL;
  97     }
  98   }
  99
 100   // Generic validation: max
 101   if (!boost::get<boost::blank>(&(max))) {
 102     if (new_value > max) {
 103       std::ostringstream oss;
 104       oss << "Value '" << new_value << "' exceeds maximum " << max;
 105       *err = oss.str();
 106       return -EINVAL;
 107     }
 108   }
 109
 110   // Generic validation: enum
 111   if (!enum_allowed.empty() && type == Option::TYPE_STR) {
 112     auto found = std::find(enum_allowed.begin(), enum_allowed.end(),
 113                            boost::get<std::string>(new_value));
 114     if (found == enum_allowed.end()) {
 115       std::ostringstream oss;
 116       oss << "'" << new_value << "' is not one of the permitted "
 117                  "values: " << joinify(enum_allowed.begin(),
 118                                        enum_allowed.end(),
 119                                        std::string(", "));
 120       *err = oss.str();
 121       return -EINVAL;
 122     }
 123   }
 124
 125   return 0;
 126 }
 127
 128 int Option::parse_value(
 129   const std::string& raw_val,
 130   value_t *out,
 131   std::string *error_message,
 132   std::string *normalized_value) const
 133 {
 134   std::string val = raw_val;
 135
 136   int r = pre_validate(&val, error_message);
 137   if (r != 0) {
 138     return r;
 139   }
 140
 141   if (type == Option::TYPE_INT) {
 142     int64_t f = strict_si_cast<int64_t>(val.c_str(), error_message);
 143     if (!error_message->empty()) {
 144       return -EINVAL;
 145     }
 146     *out = f;
 147   } else if (type == Option::TYPE_UINT) {
 148     uint64_t f = strict_si_cast<uint64_t>(val.c_str(), error_message);
 149     if (!error_message->empty()) {
 150       return -EINVAL;
 151     }
 152     *out = f;
 153   } else if (type == Option::TYPE_STR) {
 154     *out = val;
 155   } else if (type == Option::TYPE_FLOAT) {
 156     double f = strict_strtod(val.c_str(), error_message);
 157     if (!error_message->empty()) {
 158       return -EINVAL;
 159     } else {
 160       *out = f;
 161     }
 162   } else if (type == Option::TYPE_BOOL) {
 163     if (strcasecmp(val.c_str(), "false") == 0) {
 164       *out = false;
 165     } else if (strcasecmp(val.c_str(), "true") == 0) {
 166       *out = true;
 167     } else {
 168       int b = strict_strtol(val.c_str(), 10, error_message);
 169       if (!error_message->empty()) {
 170         return -EINVAL;
 171       }
 172       *out = (bool)!!b;
 173     }
 174   } else if (type == Option::TYPE_ADDR) {
 175     entity_addr_t addr;
 176     if (!addr.parse(val.c_str())){
 177       return -EINVAL;
 178     }
 179     *out = addr;
 180   } else if (type == Option::TYPE_ADDRVEC) {
 181     entity_addrvec_t addr;
 182     if (!addr.parse(val.c_str())){
 183       return -EINVAL;
 184     }
 185     *out = addr;
 186   } else if (type == Option::TYPE_UUID) {
 187     uuid_d uuid;
 188     if (!uuid.parse(val.c_str())) {
 189       return -EINVAL;
 190     }
 191     *out = uuid;
 192   } else if (type == Option::TYPE_SIZE) {
 193     Option::size_t sz{strict_iecstrtoll(val.c_str(), error_message)};
 194     if (!error_message->empty()) {
 195       return -EINVAL;
 196     }
 197     *out = sz;
 198   } else if (type == Option::TYPE_SECS) {
 199     try {
 200       *out = parse_timespan(val);
 201     } catch (const invalid_argument& e) {
 202       *error_message = e.what();
 203       return -EINVAL;
 204     }
 205   } else {
 206     ceph_abort();
 207   }
 208
 209   r = validate(*out, error_message);
 210   if (r != 0) {
 211     return r;
 212   }
 213
 214   if (normalized_value) {
 215     *normalized_value = to_str(*out);
 216   }
 217   return 0;
 218 }
 219
 220 void Option::dump(Formatter *f) const
 221 {
 222   f->dump_string("name", name);
 223
 224   f->dump_string("type", type_to_str(type));
 225
 226   f->dump_string("level", level_to_str(level));
 227
 228   f->dump_string("desc", desc);
 229   f->dump_string("long_desc", long_desc);
 230
 231   dump_value("default", value, f);
 232   dump_value("daemon_default", daemon_value, f);
 233
 234   f->open_array_section("tags");
 235   for (const auto t : tags) {
 236     f->dump_string("tag", t);
 237   }
 238   f->close_section();
 239
 240   f->open_array_section("services");
 241   for (const auto s : services) {
 242     f->dump_string("service", s);
 243   }
 244   f->close_section();
 245
 246   f->open_array_section("see_also");
 247   for (const auto sa : see_also) {
 248     f->dump_string("see_also", sa);
 249   }
 250   f->close_section();
 251
 252   if (type == TYPE_STR) {
 253     f->open_array_section("enum_values");
 254     for (const auto &ea : enum_allowed) {
 255       f->dump_string("enum_value", ea);
 256     }
 257     f->close_section();
 258   }
 259
 260   dump_value("min", min, f);
 261   dump_value("max", max, f);
 262
 263   f->dump_bool("can_update_at_runtime", can_update_at_runtime());
 264
 265   f->open_array_section("flags");
 266   if (has_flag(FLAG_RUNTIME)) {
 267     f->dump_string("option", "runtime");
 268   }
 269   if (has_flag(FLAG_NO_MON_UPDATE)) {
 270     f->dump_string("option", "no_mon_update");
 271   }
 272   if (has_flag(FLAG_STARTUP)) {
 273     f->dump_string("option", "startup");
 274   }
 275   if (has_flag(FLAG_CLUSTER_CREATE)) {
 276     f->dump_string("option", "cluster_create");
 277   }
 278   if (has_flag(FLAG_CREATE)) {
 279     f->dump_string("option", "create");
 280   }
 281   f->close_section();
 282 }
 283
 284 std::string Option::to_str(const Option::value_t& v)
 285 {
 286   return stringify(v);
 287 }
 288
 289 void Option::print(ostream *out) const
 290 {
 291   *out << name << " - " << desc << "\n";
 292   *out << "  (" << type_to_str(type) << ", " << level_to_str(level) << ")\n";
 293   if (!boost::get<boost::blank>(&daemon_value)) {
 294     *out << "  Default (non-daemon): " << stringify(value) << "\n";
 295     *out << "  Default (daemon): " << stringify(daemon_value) << "\n";
 296   } else {
 297     *out << "  Default: " << stringify(value) << "\n";
 298   }
 299   if (!enum_allowed.empty()) {
 300     *out << "  Possible values: ";
 301     for (auto& i : enum_allowed) {
 302       *out << " " << stringify(i);
 303     }
 304     *out << "\n";
 305   }
 306   if (!boost::get<boost::blank>(&min)) {
 307     *out << "  Minimum: " << stringify(min) << "\n"
 308          << "  Maximum: " << stringify(max) << "\n";
 309   }
 310   *out << "  Can update at runtime: "
 311        << (can_update_at_runtime() ? "true" : "false") << "\n";
 312   if (!services.empty()) {
 313     *out << "  Services: " << services << "\n";
 314   }
 315   if (!tags.empty()) {
 316     *out << "  Tags: " << tags << "\n";
 317   }
 318   if (!see_also.empty()) {
 319     *out << "  See also: " << see_also << "\n";
 320   }
 321
 322   if (long_desc.size()) {
 323     *out << "\n" << long_desc << "\n";
 324   }
 325 }
 326
 327 constexpr unsigned long long operator"" _min (unsigned long long min) {
 328   return min * 60;
 329 }
 330 constexpr unsigned long long operator"" _hr (unsigned long long hr) {
 331   return hr * 60 * 60;
 332 }
 333 constexpr unsigned long long operator"" _day (unsigned long long day) {
 334   return day * 60 * 60 * 24;
 335 }
 336 constexpr unsigned long long operator"" _K (unsigned long long n) {
 337   return n << 10;
 338 }
 339 constexpr unsigned long long operator"" _M (unsigned long long n) {
 340   return n << 20;
 341 }
 342 constexpr unsigned long long operator"" _G (unsigned long long n) {
 343   return n << 30;
 344 }
 345 constexpr unsigned long long operator"" _T (unsigned long long n) {
 346   return n << 40;
 347 }
 348
 349 std::vector<Option> get_global_options() {
 350   return std::vector<Option>({
 351     Option("host", Option::TYPE_STR, Option::LEVEL_BASIC)
 352     .set_description("local hostname")
 353     .set_long_description("if blank, ceph assumes the short hostname (hostname -s)")
 354     .set_flag(Option::FLAG_NO_MON_UPDATE)
 355     .add_service("common")
 356     .add_tag("network"),
 357
 358     Option("fsid", Option::TYPE_UUID, Option::LEVEL_BASIC)
 359     .set_description("cluster fsid (uuid)")
 360     .set_flag(Option::FLAG_NO_MON_UPDATE)
 361     .set_flag(Option::FLAG_STARTUP)
 362     .add_service("common")
 363     .add_tag("service"),
 364
 365     Option("public_addr", Option::TYPE_ADDR, Option::LEVEL_BASIC)
 366     .set_description("public-facing address to bind to")
 367     .set_flag(Option::FLAG_STARTUP)
 368     .add_service({"mon", "mds", "osd", "mgr"}),
 369
 370     Option("public_addrv", Option::TYPE_ADDRVEC, Option::LEVEL_BASIC)
 371     .set_description("public-facing address to bind to")
 372     .set_flag(Option::FLAG_STARTUP)
 373     .add_service({"mon", "mds", "osd", "mgr"}),
 374
 375     Option("public_bind_addr", Option::TYPE_ADDR, Option::LEVEL_ADVANCED)
 376     .set_default(entity_addr_t())
 377     .set_flag(Option::FLAG_STARTUP)
 378     .add_service("mon")
 379     .set_description(""),
 380
 381     Option("cluster_addr", Option::TYPE_ADDR, Option::LEVEL_BASIC)
 382     .set_description("cluster-facing address to bind to")
 383     .add_service("osd")
 384     .set_flag(Option::FLAG_STARTUP)
 385     .add_tag("network"),
 386
 387     Option("public_network", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 388     .add_service({"mon", "mds", "osd", "mgr"})
 389     .set_flag(Option::FLAG_STARTUP)
 390     .add_tag("network")
 391     .set_description("Network(s) from which to choose a public address to bind to"),
 392
 393     Option("public_network_interface", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 394     .add_service({"mon", "mds", "osd", "mgr"})
 395     .add_tag("network")
 396     .set_flag(Option::FLAG_STARTUP)
 397     .set_description("Interface name(s) from which to choose an address from a public_network to bind to; public_network must also be specified.")
 398     .add_see_also("public_network"),
 399
 400     Option("cluster_network", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 401     .add_service("osd")
 402     .set_flag(Option::FLAG_STARTUP)
 403     .add_tag("network")
 404     .set_description("Network(s) from which to choose a cluster address to bind to"),
 405
 406     Option("cluster_network_interface", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 407     .add_service({"mon", "mds", "osd", "mgr"})
 408     .set_flag(Option::FLAG_STARTUP)
 409     .add_tag("network")
 410     .set_description("Interface name(s) from which to choose an address from a cluster_network to bind to; cluster_network must also be specified.")
 411     .add_see_also("cluster_network"),
 412
 413     Option("monmap", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 414     .set_description("path to MonMap file")
 415     .set_long_description("This option is normally used during mkfs, but can also "
 416                         "be used to identify which monitors to connect to.")
 417     .set_flag(Option::FLAG_NO_MON_UPDATE)
 418     .add_service("mon")
 419     .set_flag(Option::FLAG_CREATE),
 420
 421     Option("mon_host", Option::TYPE_STR, Option::LEVEL_BASIC)
 422     .set_description("list of hosts or addresses to search for a monitor")
 423     .set_long_description("This is a comma, whitespace, or semicolon separated "
 424                         "list of IP addresses or hostnames. Hostnames are "
 425                         "resolved via DNS and all A or AAAA records are "
 426                         "included in the search list.")
 427     .set_flag(Option::FLAG_NO_MON_UPDATE)
 428     .set_flag(Option::FLAG_STARTUP)
 429     .add_service("common"),
 430
 431     Option("mon_dns_srv_name", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 432     .set_default("ceph-mon")
 433     .set_description("name of DNS SRV record to check for monitor addresses")
 434     .set_flag(Option::FLAG_STARTUP)
 435     .add_service("common")
 436     .add_tag("network")
 437     .add_see_also("mon_host"),
 438
 439     Option("container_image", Option::TYPE_STR, Option::LEVEL_BASIC)
 440     .set_description("container image (used by cephadm orchestrator)")
 441     .set_flag(Option::FLAG_STARTUP)
 442     .set_default("docker.io/ceph/ceph:v15"),
 443
 444     Option("no_config_file", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
 445     .set_default(false)
 446     .set_flag(Option::FLAG_NO_MON_UPDATE)
 447     .set_flag(Option::FLAG_STARTUP)
 448     .add_service("common")
 449     .add_tag("config")
 450     .set_description("signal that we don't require a config file to be present")
 451     .set_long_description("When specified, we won't be looking for a "
 452                           "configuration file, and will instead expect that "
 453                           "whatever options or values are required for us to "
 454                           "work will be passed as arguments."),
 455
 456     // lockdep
 457     Option("lockdep", Option::TYPE_BOOL, Option::LEVEL_DEV)
 458     .set_description("enable lockdep lock dependency analyzer")
 459     .set_flag(Option::FLAG_NO_MON_UPDATE)
 460     .set_flag(Option::FLAG_STARTUP)
 461     .add_service("common"),
 462
 463     Option("lockdep_force_backtrace", Option::TYPE_BOOL, Option::LEVEL_DEV)
 464     .set_description("always gather current backtrace at every lock")
 465     .set_flag(Option::FLAG_STARTUP)
 466     .add_service("common")
 467     .add_see_also("lockdep"),
 468
 469     Option("run_dir", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 470     .set_default("/var/run/ceph")
 471     .set_flag(Option::FLAG_STARTUP)
 472     .set_description("path for the 'run' directory for storing pid and socket files")
 473     .add_service("common")
 474     .add_see_also("admin_socket"),
 475
 476     Option("admin_socket", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 477     .set_default("")
 478     .set_daemon_default("$run_dir/$cluster-$name.asok")
 479     .set_flag(Option::FLAG_STARTUP)
 480     .set_description("path for the runtime control socket file, used by the 'ceph daemon' command")
 481     .add_service("common"),
 482
 483     Option("admin_socket_mode", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 484     .set_description("file mode to set for the admin socket file, e.g, '0755'")
 485     .set_flag(Option::FLAG_STARTUP)
 486     .add_service("common")
 487     .add_see_also("admin_socket"),
 488
 489     // daemon
 490     Option("daemonize", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
 491     .set_default(false)
 492     .set_daemon_default(true)
 493     .set_description("whether to daemonize (background) after startup")
 494     .set_flag(Option::FLAG_STARTUP)
 495     .set_flag(Option::FLAG_NO_MON_UPDATE)
 496     .add_service({"mon", "mgr", "osd", "mds"})
 497     .add_tag("service")
 498     .add_see_also({"pid_file", "chdir"}),
 499
 500     Option("setuser", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 501     .set_flag(Option::FLAG_STARTUP)
 502     .set_description("uid or user name to switch to on startup")
 503     .set_long_description("This is normally specified by the systemd unit file.")
 504     .add_service({"mon", "mgr", "osd", "mds"})
 505     .add_tag("service")
 506     .add_see_also("setgroup"),
 507
 508     Option("setgroup", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 509     .set_flag(Option::FLAG_STARTUP)
 510     .set_description("gid or group name to switch to on startup")
 511     .set_long_description("This is normally specified by the systemd unit file.")
 512     .add_service({"mon", "mgr", "osd", "mds"})
 513     .add_tag("service")
 514     .add_see_also("setuser"),
 515
 516     Option("setuser_match_path", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 517     .set_flag(Option::FLAG_STARTUP)
 518     .set_description("if set, setuser/setgroup is condition on this path matching ownership")
 519     .set_long_description("If setuser or setgroup are specified, and this option is non-empty, then the uid/gid of the daemon will only be changed if the file or directory specified by this option has a matching uid and/or gid.  This exists primarily to allow switching to user ceph for OSDs to be conditional on whether the osd data contents have also been chowned after an upgrade.  This is normally specified by the systemd unit file.")
 520     .add_service({"mon", "mgr", "osd", "mds"})
 521     .add_tag("service")
 522     .add_see_also({"setuser", "setgroup"}),
 523
 524     Option("pid_file", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 525     .set_flag(Option::FLAG_STARTUP)
 526     .set_description("path to write a pid file (if any)")
 527     .add_service({"mon", "mgr", "osd", "mds"})
 528     .add_tag("service"),
 529
 530     Option("chdir", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 531     .set_description("path to chdir(2) to after daemonizing")
 532     .set_flag(Option::FLAG_STARTUP)
 533     .set_flag(Option::FLAG_NO_MON_UPDATE)
 534     .add_service({"mon", "mgr", "osd", "mds"})
 535     .add_tag("service")
 536     .add_see_also("daemonize"),
 537
 538     Option("fatal_signal_handlers", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
 539     .set_default(true)
 540     .set_flag(Option::FLAG_STARTUP)
 541     .set_description("whether to register signal handlers for SIGABRT etc that dump a stack trace")
 542     .set_long_description("This is normally true for daemons and values for libraries.")
 543     .add_service({"mon", "mgr", "osd", "mds"})
 544     .add_tag("service"),
 545
 546     Option("crash_dir", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 547     .set_flag(Option::FLAG_STARTUP)
 548     .set_default("/var/lib/ceph/crash")
 549     .set_description("Directory where crash reports are archived"),
 550
 551     // restapi
 552     Option("restapi_log_level", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 553     .set_description("default set by python code"),
 554
 555     Option("restapi_base_url", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 556     .set_description("default set by python code"),
 557
 558     Option("erasure_code_dir", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 559     .set_default(CEPH_PKGLIBDIR"/erasure-code")
 560     .set_flag(Option::FLAG_STARTUP)
 561     .set_description("directory where erasure-code plugins can be found")
 562     .add_service({"mon", "osd"}),
 563
 564     // logging
 565     Option("log_file", Option::TYPE_STR, Option::LEVEL_BASIC)
 566     .set_default("")
 567     .set_daemon_default("/var/log/ceph/$cluster-$name.log")
 568     .set_description("path to log file")
 569     .add_see_also({"log_to_file",
 570                    "log_to_stderr",
 571                    "err_to_stderr",
 572                    "log_to_syslog",
 573                    "err_to_syslog"}),
 574
 575     Option("log_max_new", Option::TYPE_INT, Option::LEVEL_ADVANCED)
 576     .set_default(1000)
 577     .set_description("max unwritten log entries to allow before waiting to flush to the log")
 578     .add_see_also("log_max_recent"),
 579
 580     Option("log_max_recent", Option::TYPE_INT, Option::LEVEL_ADVANCED)
 581     .set_default(500)
 582     .set_daemon_default(10000)
 583     .set_description("recent log entries to keep in memory to dump in the event of a crash")
 584     .set_long_description("The purpose of this option is to log at a higher debug level only to the in-memory buffer, and write out the detailed log messages only if there is a crash.  Only log entries below the lower log level will be written unconditionally to the log.  For example, debug_osd=1/5 will write everything <= 1 to the log unconditionally but keep entries at levels 2-5 in memory.  If there is a seg fault or assertion failure, all entries will be dumped to the log."),
 585
 586     Option("log_to_file", Option::TYPE_BOOL, Option::LEVEL_BASIC)
 587     .set_default(true)
 588     .set_description("send log lines to a file")
 589     .add_see_also("log_file"),
 590
 591     Option("log_to_stderr", Option::TYPE_BOOL, Option::LEVEL_BASIC)
 592     .set_default(true)
 593     .set_daemon_default(false)
 594     .set_description("send log lines to stderr"),
 595
 596     Option("err_to_stderr", Option::TYPE_BOOL, Option::LEVEL_BASIC)
 597     .set_default(false)
 598     .set_daemon_default(true)
 599     .set_description("send critical error log lines to stderr"),
 600
 601     Option("log_stderr_prefix", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 602     .set_description("String to prefix log messages with when sent to stderr")
 603     .set_long_description("This is useful in container environments when combined with mon_cluster_log_to_stderr.  The mon log prefixes each line with the channel name (e.g., 'default', 'audit'), while log_stderr_prefix can be set to 'debug '.")
 604     .add_see_also("mon_cluster_log_to_stderr"),
 605
 606     Option("log_to_syslog", Option::TYPE_BOOL, Option::LEVEL_BASIC)
 607     .set_default(false)
 608     .set_description("send log lines to syslog facility"),
 609
 610     Option("err_to_syslog", Option::TYPE_BOOL, Option::LEVEL_BASIC)
 611     .set_default(false)
 612     .set_description("send critical error log lines to syslog facility"),
 613
 614     Option("log_flush_on_exit", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
 615     .set_default(false)
 616     .set_description("set a process exit handler to ensure the log is flushed on exit"),
 617
 618     Option("log_stop_at_utilization", Option::TYPE_FLOAT, Option::LEVEL_BASIC)
 619     .set_default(.97)
 620     .set_min_max(0.0, 1.0)
 621     .set_description("stop writing to the log file when device utilization reaches this ratio")
 622     .add_see_also("log_file"),
 623
 624     Option("log_to_graylog", Option::TYPE_BOOL, Option::LEVEL_BASIC)
 625     .set_default(false)
 626     .set_description("send log lines to remote graylog server")
 627     .add_see_also({"err_to_graylog",
 628                    "log_graylog_host",
 629                    "log_graylog_port"}),
 630
 631     Option("err_to_graylog", Option::TYPE_BOOL, Option::LEVEL_BASIC)
 632     .set_default(false)
 633     .set_description("send critical error log lines to remote graylog server")
 634     .add_see_also({"log_to_graylog",
 635                    "log_graylog_host",
 636                    "log_graylog_port"}),
 637
 638     Option("log_graylog_host", Option::TYPE_STR, Option::LEVEL_BASIC)
 639     .set_default("127.0.0.1")
 640     .set_description("address or hostname of graylog server to log to")
 641     .add_see_also({"log_to_graylog",
 642                    "err_to_graylog",
 643                    "log_graylog_port"}),
 644
 645     Option("log_graylog_port", Option::TYPE_INT, Option::LEVEL_BASIC)
 646     .set_default(12201)
 647     .set_description("port number for the remote graylog server")
 648     .add_see_also("log_graylog_host"),
 649
 650     Option("log_coarse_timestamps", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
 651     .set_default(true)
 652     .set_description("timestamp log entries from coarse system clock "
 653                      "to improve performance")
 654     .add_service("common")
 655     .add_tag("performance")
 656     .add_tag("service"),
 657
 658
 659     // unmodified
 660     Option("clog_to_monitors", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 661     .set_default("default=true")
 662     .set_flag(Option::FLAG_RUNTIME)
 663     .set_description("Make daemons send cluster log messages to monitors"),
 664
 665     Option("clog_to_syslog", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 666     .set_default("false")
 667     .set_flag(Option::FLAG_RUNTIME)
 668     .set_description("Make daemons send cluster log messages to syslog"),
 669
 670     Option("clog_to_syslog_level", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 671     .set_default("info")
 672     .set_flag(Option::FLAG_RUNTIME)
 673     .set_description("Syslog level for cluster log messages")
 674     .add_see_also("clog_to_syslog"),
 675
 676     Option("clog_to_syslog_facility", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 677     .set_default("default=daemon audit=local0")
 678     .set_flag(Option::FLAG_RUNTIME)
 679     .set_description("Syslog facility for cluster log messages")
 680     .add_see_also("clog_to_syslog"),
 681
 682     Option("clog_to_graylog", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 683     .set_default("false")
 684     .set_flag(Option::FLAG_RUNTIME)
 685     .set_description("Make daemons send cluster log to graylog"),
 686
 687     Option("clog_to_graylog_host", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 688     .set_default("127.0.0.1")
 689     .set_flag(Option::FLAG_RUNTIME)
 690     .set_description("Graylog host to cluster log messages")
 691     .add_see_also("clog_to_graylog"),
 692
 693     Option("clog_to_graylog_port", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 694     .set_default("12201")
 695     .set_flag(Option::FLAG_RUNTIME)
 696     .set_description("Graylog port number for cluster log messages")
 697     .add_see_also("clog_to_graylog"),
 698
 699     Option("mon_cluster_log_to_stderr", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
 700     .set_default(false)
 701     .add_service("mon")
 702     .set_flag(Option::FLAG_RUNTIME)
 703     .set_description("Make monitor send cluster log messages to stderr (prefixed by channel)")
 704     .add_see_also("log_stderr_prefix"),
 705
 706     Option("mon_cluster_log_to_syslog", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 707     .set_default("default=false")
 708     .set_flag(Option::FLAG_RUNTIME)
 709     .add_service("mon")
 710     .set_description("Make monitor send cluster log messages to syslog"),
 711
 712     Option("mon_cluster_log_to_syslog_level", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 713     .set_default("info")
 714     .add_service("mon")
 715     .set_flag(Option::FLAG_RUNTIME)
 716     .set_description("Syslog level for cluster log messages")
 717     .add_see_also("mon_cluster_log_to_syslog"),
 718
 719     Option("mon_cluster_log_to_syslog_facility", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 720     .set_default("daemon")
 721     .add_service("mon")
 722     .set_flag(Option::FLAG_RUNTIME)
 723     .set_description("Syslog facility for cluster log messages")
 724     .add_see_also("mon_cluster_log_to_syslog"),
 725
 726     Option("mon_cluster_log_to_file", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
 727     .set_default(true)
 728     .set_flag(Option::FLAG_RUNTIME)
 729     .add_service("mon")
 730     .set_description("Make monitor send cluster log messages to file")
 731     .add_see_also("mon_cluster_log_file"),
 732
 733     Option("mon_cluster_log_file", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 734     .set_default("default=/var/log/ceph/$cluster.$channel.log cluster=/var/log/ceph/$cluster.log")
 735     .set_flag(Option::FLAG_RUNTIME)
 736     .add_service("mon")
 737     .set_description("File(s) to write cluster log to")
 738     .set_long_description("This can either be a simple file name to receive all messages, or a list of key/value pairs where the key is the log channel and the value is the filename, which may include $cluster and $channel metavariables")
 739     .add_see_also("mon_cluster_log_to_file"),
 740
 741     Option("mon_cluster_log_file_level", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 742     .set_default("debug")
 743     .set_flag(Option::FLAG_RUNTIME)
 744     .add_service("mon")
 745     .set_description("Lowest level to include is cluster log file")
 746     .add_see_also("mon_cluster_log_file"),
 747
 748     Option("mon_cluster_log_to_graylog", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 749     .set_default("false")
 750     .set_flag(Option::FLAG_RUNTIME)
 751     .add_service("mon")
 752     .set_description("Make monitor send cluster log to graylog"),
 753
 754     Option("mon_cluster_log_to_graylog_host", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 755     .set_default("127.0.0.1")
 756     .set_flag(Option::FLAG_RUNTIME)
 757     .add_service("mon")
 758     .set_description("Graylog host for cluster log messages")
 759     .add_see_also("mon_cluster_log_to_graylog"),
 760
 761     Option("mon_cluster_log_to_graylog_port", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 762     .set_default("12201")
 763     .set_flag(Option::FLAG_RUNTIME)
 764     .add_service("mon")
 765     .set_description("Graylog port for cluster log messages")
 766     .add_see_also("mon_cluster_log_to_graylog"),
 767
 768     Option("enable_experimental_unrecoverable_data_corrupting_features", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 769     .set_flag(Option::FLAG_RUNTIME)
 770     .set_default("")
 771     .set_description("Enable named (or all with '*') experimental features that may be untested, dangerous, and/or cause permanent data loss"),
 772
 773     Option("plugin_dir", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 774     .set_default(CEPH_PKGLIBDIR)
 775     .set_flag(Option::FLAG_STARTUP)
 776     .add_service({"mon", "osd"})
 777     .set_description("Base directory for dynamically loaded plugins"),
 778
 779     // Compressor
 780     Option("compressor_zlib_isal", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
 781     .set_default(false)
 782     .set_description("Use Intel ISA-L accelerated zlib implementation if available"),
 783
 784     Option("compressor_zlib_level", Option::TYPE_INT, Option::LEVEL_ADVANCED)
 785     .set_default(5)
 786     .set_description("Zlib compression level to use"),
 787
 788     Option("qat_compressor_enabled", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
 789     .set_default(false)
 790     .set_description("Enable Intel QAT acceleration support for compression if available"),
 791
 792     Option("plugin_crypto_accelerator", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 793     .set_default("crypto_isal")
 794     .set_description("Crypto accelerator library to use"),
 795
 796     Option("mempool_debug", Option::TYPE_BOOL, Option::LEVEL_DEV)
 797     .set_default(false)
 798     .set_flag(Option::FLAG_NO_MON_UPDATE)
 799     .set_description(""),
 800
 801     Option("thp", Option::TYPE_BOOL, Option::LEVEL_DEV)
 802     .set_default(false)
 803     .set_flag(Option::FLAG_STARTUP)
 804     .set_description("enable transparent huge page (THP) support")
 805     .set_long_description("Ceph is known to suffer from memory fragmentation due to THP use. This is indicated by RSS usage above configured memory targets. Enabling THP is currently discouraged until selective use of THP by Ceph is implemented."),
 806
 807     Option("key", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 808     .set_default("")
 809     .set_description("Authentication key")
 810     .set_long_description("A CephX authentication key, base64 encoded.  It normally looks something like 'AQAtut9ZdMbNJBAAHz6yBAWyJyz2yYRyeMWDag=='.")
 811     .set_flag(Option::FLAG_STARTUP)
 812     .set_flag(Option::FLAG_NO_MON_UPDATE)
 813     .add_see_also("keyfile")
 814     .add_see_also("keyring"),
 815
 816     Option("keyfile", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 817     .set_default("")
 818     .set_description("Path to a file containing a key")
 819     .set_long_description("The file should contain a CephX authentication key and optionally a trailing newline, but nothing else.")
 820     .set_flag(Option::FLAG_STARTUP)
 821     .set_flag(Option::FLAG_NO_MON_UPDATE)
 822     .add_see_also("key"),
 823
 824     Option("keyring", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 825     .set_default(
 826       "/etc/ceph/$cluster.$name.keyring,/etc/ceph/$cluster.keyring,"
 827       "/etc/ceph/keyring,/etc/ceph/keyring.bin,"
 828   #if defined(__FreeBSD)
 829       "/usr/local/etc/ceph/$cluster.$name.keyring,"
 830       "/usr/local/etc/ceph/$cluster.keyring,"
 831       "/usr/local/etc/ceph/keyring,/usr/local/etc/ceph/keyring.bin,"
 832   #endif
 833     )
 834     .set_description("Path to a keyring file.")
 835     .set_long_description("A keyring file is an INI-style formatted file where the section names are client or daemon names (e.g., 'osd.0') and each section contains a 'key' property with CephX authentication key as the value.")
 836     .set_flag(Option::FLAG_STARTUP)
 837     .set_flag(Option::FLAG_NO_MON_UPDATE)
 838     .add_see_also("key")
 839     .add_see_also("keyfile"),
 840
 841     Option("heartbeat_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
 842     .set_default(5)
 843     .set_flag(Option::FLAG_STARTUP)
 844     .set_description("Frequency of internal heartbeat checks (seconds)"),
 845
 846     Option("heartbeat_file", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 847     .set_default("")
 848     .set_flag(Option::FLAG_STARTUP)
 849     .set_description("File to touch on successful internal heartbeat")
 850     .set_long_description("If set, this file will be touched every time an internal heartbeat check succeeds.")
 851     .add_see_also("heartbeat_interval"),
 852
 853     Option("heartbeat_inject_failure", Option::TYPE_INT, Option::LEVEL_DEV)
 854     .set_default(0)
 855     .set_description(""),
 856
 857     Option("perf", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
 858     .set_default(true)
 859     .set_description("Enable internal performance metrics")
 860     .set_long_description("If enabled, collect and expose internal health metrics"),
 861
 862     Option("ms_type", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 863     .set_flag(Option::FLAG_STARTUP)
 864     .set_default("async+posix")
 865     .set_description("Messenger implementation to use for network communication"),
 866
 867     Option("ms_public_type", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 868     .set_default("")
 869     .set_flag(Option::FLAG_STARTUP)
 870     .set_description("Messenger implementation to use for the public network")
 871     .set_long_description("If not specified, use ms_type")
 872     .add_see_also("ms_type"),
 873
 874     Option("ms_cluster_type", Option::TYPE_STR, Option::LEVEL_ADVANCED)
 875     .set_default("")
 876     .set_flag(Option::FLAG_STARTUP)
 877     .set_description("Messenger implementation to use for the internal cluster network")
 878     .set_long_description("If not specified, use ms_type")
 879     .add_see_also("ms_type"),
 880
 881     Option("ms_mon_cluster_mode", Option::TYPE_STR, Option::LEVEL_BASIC)
 882     .set_default("secure crc")
 883     .set_flag(Option::FLAG_STARTUP)
 884     .set_description("Connection modes (crc, secure) for intra-mon connections in order of preference")
 885     .add_see_also("ms_mon_service_mode")
 886     .add_see_also("ms_mon_client_mode")
 887     .add_see_also("ms_service_mode")
 888     .add_see_also("ms_cluster_mode")
 889     .add_see_also("ms_client_mode"),
 890
 891     Option("ms_mon_service_mode", Option::TYPE_STR, Option::LEVEL_BASIC)
 892     .set_default("secure crc")
 893     .set_flag(Option::FLAG_STARTUP)
 894     .set_description("Allowed connection modes (crc, secure) for connections to mons")
 895     .add_see_also("ms_service_mode")
 896     .add_see_also("ms_mon_cluster_mode")
 897     .add_see_also("ms_mon_client_mode")
 898     .add_see_also("ms_cluster_mode")
 899     .add_see_also("ms_client_mode"),
 900
 901     Option("ms_mon_client_mode", Option::TYPE_STR, Option::LEVEL_BASIC)
 902     .set_default("secure crc")
 903     .set_flag(Option::FLAG_STARTUP)
 904     .set_description("Connection modes (crc, secure) for connections from clients to monitors in order of preference")
 905     .add_see_also("ms_mon_service_mode")
 906     .add_see_also("ms_mon_cluster_mode")
 907     .add_see_also("ms_service_mode")
 908     .add_see_also("ms_cluster_mode")
 909     .add_see_also("ms_client_mode"),
 910
 911     Option("ms_cluster_mode", Option::TYPE_STR, Option::LEVEL_BASIC)
 912     .set_default("crc secure")
 913     .set_flag(Option::FLAG_STARTUP)
 914     .set_description("Connection modes (crc, secure) for intra-cluster connections in order of preference")
 915     .add_see_also("ms_service_mode")
 916     .add_see_also("ms_client_mode"),
 917
 918     Option("ms_service_mode", Option::TYPE_STR, Option::LEVEL_BASIC)
 919     .set_default("crc secure")
 920     .set_flag(Option::FLAG_STARTUP)
 921     .set_description("Allowed connection modes (crc, secure) for connections to daemons")
 922     .add_see_also("ms_cluster_mode")
 923     .add_see_also("ms_client_mode"),
 924
 925     Option("ms_client_mode", Option::TYPE_STR, Option::LEVEL_BASIC)
 926     .set_default("crc secure")
 927     .set_flag(Option::FLAG_STARTUP)
 928     .set_description("Connection modes (crc, secure) for connections from clients in order of preference")
 929     .add_see_also("ms_cluster_mode")
 930     .add_see_also("ms_service_mode"),
 931
 932     Option("ms_learn_addr_from_peer", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
 933     .set_default(true)
 934     .set_description("Learn address from what IP our first peer thinks we connect from")
 935     .set_long_description("Use the IP address our first peer (usually a monitor) sees that we are connecting from.  This is useful if a client is behind some sort of NAT and we want to see it identified by its local (not NATed) address."),
 936
 937     Option("ms_tcp_nodelay", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
 938     .set_default(true)
 939     .set_description("Disable Nagle's algorithm and send queued network traffic immediately"),
 940
 941     Option("ms_tcp_rcvbuf", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
 942     .set_default(0)
 943     .set_description("Size of TCP socket receive buffer"),
 944
 945     Option("ms_tcp_prefetch_max_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
 946     .set_default(4_K)
 947     .set_description("Maximum amount of data to prefetch out of the socket receive buffer"),
 948
 949     Option("ms_initial_backoff", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
 950     .set_default(.2)
 951     .set_description("Initial backoff after a network error is detected (seconds)"),
 952
 953     Option("ms_max_backoff", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
 954     .set_default(15.0)
 955     .set_description("Maximum backoff after a network error before retrying (seconds)")
 956     .add_see_also("ms_initial_backoff"),
 957
 958     Option("ms_crc_data", Option::TYPE_BOOL, Option::LEVEL_DEV)
 959     .set_default(true)
 960     .set_description("Set and/or verify crc32c checksum on data payload sent over network"),
 961
 962     Option("ms_crc_header", Option::TYPE_BOOL, Option::LEVEL_DEV)
 963     .set_default(true)
 964     .set_description("Set and/or verify crc32c checksum on header payload sent over network"),
 965
 966     Option("ms_die_on_bad_msg", Option::TYPE_BOOL, Option::LEVEL_DEV)
 967     .set_default(false)
 968     .set_description("Induce a daemon crash/exit when a bad network message is received"),
 969
 970     Option("ms_die_on_unhandled_msg", Option::TYPE_BOOL, Option::LEVEL_DEV)
 971     .set_default(false)
 972     .set_description("Induce a daemon crash/exit when an unrecognized message is received"),
 973
 974     Option("ms_die_on_old_message", Option::TYPE_BOOL, Option::LEVEL_DEV)
 975     .set_default(false)
 976     .set_description("Induce a daemon crash/exit when a old, undecodable message is received"),
 977
 978     Option("ms_die_on_skipped_message", Option::TYPE_BOOL, Option::LEVEL_DEV)
 979     .set_default(false)
 980     .set_description("Induce a daemon crash/exit if sender skips a message sequence number"),
 981
 982     Option("ms_die_on_bug", Option::TYPE_BOOL, Option::LEVEL_DEV)
 983     .set_default(false)
 984     .set_description("Induce a crash/exit on various bugs (for testing purposes)"),
 985
 986     Option("ms_dispatch_throttle_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
 987     .set_default(100_M)
 988     .set_description("Limit messages that are read off the network but still being processed"),
 989
 990     Option("ms_bind_ipv4", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
 991     .set_default(true)
 992     .set_description("Bind servers to IPv4 address(es)")
 993     .add_see_also("ms_bind_ipv6"),
 994
 995     Option("ms_bind_ipv6", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
 996     .set_default(false)
 997     .set_description("Bind servers to IPv6 address(es)")
 998     .add_see_also("ms_bind_ipv4"),
 999
1000     Option("ms_bind_prefer_ipv4", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1001     .set_default(false)
1002     .set_description("Prefer IPV4 over IPV6 address(es)"),
1003
1004     Option("ms_bind_msgr1", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1005     .set_default(true)
1006     .set_description("Bind servers to msgr1 (legacy) protocol address(es)")
1007     .add_see_also("ms_bind_msgr2"),
1008
1009     Option("ms_bind_msgr2", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1010     .set_default(true)
1011     .set_description("Bind servers to msgr2 (nautilus+) protocol address(es)")
1012     .add_see_also("ms_bind_msgr1"),
1013
1014     Option("ms_bind_port_min", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1015     .set_default(6800)
1016     .set_description("Lowest port number to bind daemon(s) to"),
1017
1018     Option("ms_bind_port_max", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1019     .set_default(7300)
1020     .set_description("Highest port number to bind daemon(s) to"),
1021
1022     Option("ms_bind_retry_count", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1023   #if !defined(__FreeBSD__)
1024     .set_default(3)
1025   #else
1026     // FreeBSD does not use SO_REAUSEADDR so allow for a bit more time per default
1027     .set_default(6)
1028   #endif
1029     .set_description("Number of attempts to make while bind(2)ing to a port"),
1030
1031     Option("ms_bind_retry_delay", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1032   #if !defined(__FreeBSD__)
1033     .set_default(5)
1034   #else
1035     // FreeBSD does not use SO_REAUSEADDR so allow for a bit more time per default
1036     .set_default(6)
1037   #endif
1038     .set_description("Delay between bind(2) attempts (seconds)"),
1039
1040     Option("ms_bind_before_connect", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1041     .set_default(false)
1042     .set_description("Call bind(2) on client sockets"),
1043
1044     Option("ms_tcp_listen_backlog", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1045     .set_default(512)
1046     .set_description("Size of queue of incoming connections for accept(2)"),
1047
1048
1049     Option("ms_connection_ready_timeout", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1050     .set_default(10)
1051     .set_description("Time before we declare a not yet ready connection as dead (seconds)"),
1052
1053     Option("ms_connection_idle_timeout", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1054     .set_default(900)
1055     .set_description("Time before an idle connection is closed (seconds)"),
1056
1057     Option("ms_pq_max_tokens_per_priority", Option::TYPE_UINT, Option::LEVEL_DEV)
1058     .set_default(16777216)
1059     .set_description(""),
1060
1061     Option("ms_pq_min_cost", Option::TYPE_SIZE, Option::LEVEL_DEV)
1062     .set_default(65536)
1063     .set_description(""),
1064
1065     Option("ms_inject_socket_failures", Option::TYPE_UINT, Option::LEVEL_DEV)
1066     .set_default(0)
1067     .set_description("Inject a socket failure every Nth socket operation"),
1068
1069     Option("ms_inject_delay_type", Option::TYPE_STR, Option::LEVEL_DEV)
1070     .set_default("")
1071     .set_description("Entity type to inject delays for")
1072     .set_flag(Option::FLAG_RUNTIME),
1073
1074     Option("ms_inject_delay_msg_type", Option::TYPE_STR, Option::LEVEL_DEV)
1075     .set_default("")
1076     .set_description("Message type to inject delays for"),
1077
1078     Option("ms_inject_delay_max", Option::TYPE_FLOAT, Option::LEVEL_DEV)
1079     .set_default(1)
1080     .set_description("Max delay to inject"),
1081
1082     Option("ms_inject_delay_probability", Option::TYPE_FLOAT, Option::LEVEL_DEV)
1083     .set_default(0)
1084     .set_description(""),
1085
1086     Option("ms_inject_internal_delays", Option::TYPE_FLOAT, Option::LEVEL_DEV)
1087     .set_default(0)
1088     .set_description("Inject various internal delays to induce races (seconds)"),
1089
1090     Option("ms_blackhole_osd", Option::TYPE_BOOL, Option::LEVEL_DEV)
1091     .set_default(false)
1092     .set_description(""),
1093
1094     Option("ms_blackhole_mon", Option::TYPE_BOOL, Option::LEVEL_DEV)
1095     .set_default(false)
1096     .set_description(""),
1097
1098     Option("ms_blackhole_mds", Option::TYPE_BOOL, Option::LEVEL_DEV)
1099     .set_default(false)
1100     .set_description(""),
1101
1102     Option("ms_blackhole_mgr", Option::TYPE_BOOL, Option::LEVEL_DEV)
1103     .set_default(false)
1104     .set_description(""),
1105
1106     Option("ms_blackhole_client", Option::TYPE_BOOL, Option::LEVEL_DEV)
1107     .set_default(false)
1108     .set_description(""),
1109
1110     Option("ms_dump_on_send", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1111     .set_default(false)
1112     .set_description("Hexdump message to debug log on message send"),
1113
1114     Option("ms_dump_corrupt_message_level", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1115     .set_default(1)
1116     .set_description("Log level at which to hexdump corrupt messages we receive"),
1117
1118     Option("ms_async_op_threads", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1119     .set_default(3)
1120     .set_min_max(1, 24)
1121     .set_description("Threadpool size for AsyncMessenger (ms_type=async)"),
1122
1123     Option("ms_async_max_op_threads", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1124     .set_default(5)
1125     .set_description("Maximum threadpool size of AsyncMessenger")
1126     .add_see_also("ms_async_op_threads"),
1127
1128     Option("ms_async_rdma_device_name", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1129     .set_default("")
1130     .set_description(""),
1131
1132     Option("ms_async_rdma_enable_hugepage", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1133     .set_default(false)
1134     .set_description(""),
1135
1136     Option("ms_async_rdma_buffer_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
1137     .set_default(128_K)
1138     .set_description(""),
1139
1140     Option("ms_async_rdma_send_buffers", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1141     .set_default(1_K)
1142     .set_description(""),
1143
1144     Option("ms_async_rdma_receive_buffers", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1145     .set_default(32768)
1146     .set_description(""),
1147
1148     Option("ms_async_rdma_receive_queue_len", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1149     .set_default(4096)
1150     .set_description(""),
1151
1152     Option("ms_async_rdma_support_srq", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1153     .set_default(true)
1154     .set_description(""),
1155
1156     Option("ms_async_rdma_port_num", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1157     .set_default(1)
1158     .set_description(""),
1159
1160     Option("ms_async_rdma_polling_us", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1161     .set_default(1000)
1162     .set_description(""),
1163
1164     Option("ms_async_rdma_gid_idx", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1165     .set_default(0)
1166     .set_description("use gid_idx to select GID for choosing RoCEv1 or RoCEv2"),
1167
1168     Option("ms_async_rdma_local_gid", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1169     .set_default("")
1170     .set_description(""),
1171
1172     Option("ms_async_rdma_roce_ver", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1173     .set_default(1)
1174     .set_description(""),
1175
1176     Option("ms_async_rdma_sl", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1177     .set_default(3)
1178     .set_description(""),
1179
1180     Option("ms_async_rdma_dscp", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1181     .set_default(96)
1182     .set_description(""),
1183
1184     Option("ms_max_accept_failures", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1185     .set_default(4)
1186     .set_description("The maximum number of consecutive failed accept() calls before "
1187                      "considering the daemon is misconfigured and abort it."),
1188
1189     Option("ms_async_rdma_cm", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1190     .set_default(false)
1191     .set_description(""),
1192
1193     Option("ms_async_rdma_type", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1194     .set_default("ib")
1195     .set_description(""),
1196
1197     Option("ms_dpdk_port_id", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1198     .set_default(0)
1199     .set_description(""),
1200
1201     Option("ms_dpdk_coremask", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1202     .set_default("0xF")         //begin with 0x for the string
1203     .set_description("")
1204     .add_see_also("ms_async_op_threads"),
1205
1206     Option("ms_dpdk_memory_channel", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1207     .set_default("4")
1208     .set_description(""),
1209
1210     Option("ms_dpdk_hugepages", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1211     .set_default("")
1212     .set_description(""),
1213
1214     Option("ms_dpdk_pmd", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1215     .set_default("")
1216     .set_description(""),
1217
1218     Option("ms_dpdk_host_ipv4_addr", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1219     .set_default("")
1220     .set_description(""),
1221
1222     Option("ms_dpdk_gateway_ipv4_addr", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1223     .set_default("")
1224     .set_description(""),
1225
1226     Option("ms_dpdk_netmask_ipv4_addr", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1227     .set_default("")
1228     .set_description(""),
1229
1230     Option("ms_dpdk_lro", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1231     .set_default(true)
1232     .set_description(""),
1233
1234     Option("ms_dpdk_hw_flow_control", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1235     .set_default(true)
1236     .set_description(""),
1237
1238     Option("ms_dpdk_hw_queue_weight", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1239     .set_default(1)
1240     .set_description(""),
1241
1242     Option("ms_dpdk_debug_allow_loopback", Option::TYPE_BOOL, Option::LEVEL_DEV)
1243     .set_default(false)
1244     .set_description(""),
1245
1246     Option("ms_dpdk_rx_buffer_count_per_core", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1247     .set_default(8192)
1248     .set_description(""),
1249
1250     Option("inject_early_sigterm", Option::TYPE_BOOL, Option::LEVEL_DEV)
1251     .set_default(false)
1252     .set_description("send ourselves a SIGTERM early during startup"),
1253
1254     // MON
1255     Option("mon_enable_op_tracker", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1256     .set_default(true)
1257     .add_service("mon")
1258     .set_description("enable/disable MON op tracking"),
1259
1260     Option("mon_op_complaint_time", Option::TYPE_SECS, Option::LEVEL_ADVANCED)
1261     .set_default(30)
1262     .add_service("mon")
1263     .set_description("time after which to consider a monitor operation blocked "
1264                      "after no updates"),
1265
1266     Option("mon_op_log_threshold", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1267     .set_default(5)
1268     .add_service("mon")
1269     .set_description("max number of slow ops to display"),
1270
1271     Option("mon_op_history_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1272     .set_default(20)
1273     .add_service("mon")
1274     .set_description("max number of completed ops to track"),
1275
1276     Option("mon_op_history_duration", Option::TYPE_SECS, Option::LEVEL_ADVANCED)
1277     .set_default(600)
1278     .add_service("mon")
1279     .set_description("expiration time in seconds of historical MON OPS"),
1280
1281     Option("mon_op_history_slow_op_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1282     .set_default(20)
1283     .add_service("mon")
1284     .set_description("max number of slow historical MON OPS to keep"),
1285
1286     Option("mon_op_history_slow_op_threshold", Option::TYPE_SECS, Option::LEVEL_ADVANCED)
1287     .set_default(10)
1288     .add_service("mon")
1289     .set_description("duration of an op to be considered as a historical slow op"),
1290
1291     Option("mon_data", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1292     .set_flag(Option::FLAG_NO_MON_UPDATE)
1293     .set_default("/var/lib/ceph/mon/$cluster-$id")
1294     .add_service("mon")
1295     .set_description("path to mon database"),
1296
1297     Option("mon_initial_members", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1298     .set_default("")
1299     .add_service("mon")
1300     .set_flag(Option::FLAG_NO_MON_UPDATE)
1301     .set_flag(Option::FLAG_CLUSTER_CREATE)
1302     .set_description(""),
1303
1304     Option("mon_compact_on_start", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1305     .set_default(false)
1306     .add_service("mon")
1307     .set_description(""),
1308
1309     Option("mon_compact_on_bootstrap", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1310     .set_default(false)
1311     .add_service("mon")
1312     .set_description(""),
1313
1314     Option("mon_compact_on_trim", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1315     .set_default(true)
1316     .add_service("mon")
1317     .set_description(""),
1318
1319     /* -- mon: osdmap prune (begin) -- */
1320     Option("mon_osdmap_full_prune_enabled", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1321     .set_default(true)
1322     .add_service("mon")
1323     .set_description("enables pruning full osdmap versions when we go over a given number of maps")
1324     .add_see_also("mon_osdmap_full_prune_min")
1325     .add_see_also("mon_osdmap_full_prune_interval")
1326     .add_see_also("mon_osdmap_full_prune_txsize"),
1327
1328     Option("mon_osdmap_full_prune_min", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1329     .set_default(10000)
1330     .add_service("mon")
1331     .set_description("minimum number of versions in the store to trigger full map pruning")
1332     .add_see_also("mon_osdmap_full_prune_enabled")
1333     .add_see_also("mon_osdmap_full_prune_interval")
1334     .add_see_also("mon_osdmap_full_prune_txsize"),
1335
1336     Option("mon_osdmap_full_prune_interval", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1337     .set_default(10)
1338     .add_service("mon")
1339     .set_description("interval between maps that will not be pruned; maps in the middle will be pruned.")
1340     .add_see_also("mon_osdmap_full_prune_enabled")
1341     .add_see_also("mon_osdmap_full_prune_interval")
1342     .add_see_also("mon_osdmap_full_prune_txsize"),
1343
1344     Option("mon_osdmap_full_prune_txsize", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1345     .set_default(100)
1346     .add_service("mon")
1347     .set_description("number of maps we will prune per iteration")
1348     .add_see_also("mon_osdmap_full_prune_enabled")
1349     .add_see_also("mon_osdmap_full_prune_interval")
1350     .add_see_also("mon_osdmap_full_prune_txsize"),
1351     /* -- mon: osdmap prune (end) -- */
1352
1353     Option("mon_osd_cache_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1354     .set_default(500)
1355     .add_service("mon")
1356     .set_description("maximum number of OSDMaps to cache in memory"),
1357
1358     Option("mon_osd_cache_size_min", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
1359     .set_default(128_M)
1360     .add_service("mon")
1361     .set_description("The minimum amount of bytes to be kept mapped in memory for osd monitor caches."),
1362
1363     Option("mon_memory_target", Option::TYPE_SIZE, Option::LEVEL_BASIC)
1364     .set_default(2_G)
1365     .set_flag(Option::FLAG_RUNTIME)
1366     .add_service("mon")
1367     .set_description("The amount of bytes pertaining to osd monitor caches and kv cache to be kept mapped in memory with cache auto-tuning enabled"),
1368
1369     Option("mon_memory_autotune", Option::TYPE_BOOL, Option::LEVEL_BASIC)
1370     .set_default(true)
1371     .set_flag(Option::FLAG_RUNTIME)
1372     .add_service("mon")
1373     .set_description("Autotune the cache memory being used for osd monitors and kv database"),
1374
1375     Option("mon_cpu_threads", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1376     .set_default(4)
1377     .add_service("mon")
1378     .set_description("worker threads for CPU intensive background work"),
1379
1380     Option("mon_osd_mapping_pgs_per_chunk", Option::TYPE_INT, Option::LEVEL_DEV)
1381     .set_default(4096)
1382     .add_service("mon")
1383     .set_description("granularity of PG placement calculation background work"),
1384
1385     Option("mon_clean_pg_upmaps_per_chunk", Option::TYPE_UINT, Option::LEVEL_DEV)
1386     .set_default(256)
1387     .add_service("mon")
1388     .set_description("granularity of PG upmap validation background work"),
1389
1390     Option("mon_osd_max_creating_pgs", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1391     .set_default(1024)
1392     .add_service("mon")
1393     .set_description("maximum number of PGs the mon will create at once"),
1394
1395     Option("mon_osd_max_initial_pgs", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1396     .set_default(1024)
1397     .add_service("mon")
1398     .set_description("maximum number of PGs a pool will created with")
1399     .set_long_description("If the user specifies more PGs than this, the cluster will subsequently split PGs after the pool is created in order to reach the target."),
1400
1401     Option("mon_tick_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1402     .set_default(5)
1403     .add_service("mon")
1404     .set_description("interval for internal mon background checks"),
1405
1406     Option("mon_session_timeout", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1407     .set_default(300)
1408     .add_service("mon")
1409     .set_description("close inactive mon client connections after this many seconds"),
1410
1411     Option("mon_subscribe_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
1412     .set_default(1_day)
1413     .add_service("mon")
1414     .set_description("subscribe interval for pre-jewel clients"),
1415
1416     Option("mon_delta_reset_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1417     .set_default(10)
1418     .add_service("mon")
1419     .add_service("mon")
1420     .set_description("window duration for rate calculations in 'ceph status'"),
1421
1422     Option("mon_osd_laggy_halflife", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1423     .set_default(1_hr)
1424     .add_service("mon")
1425     .set_description("halflife of OSD 'lagginess' factor"),
1426
1427     Option("mon_osd_laggy_weight", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1428     .set_default(.3)
1429     .set_min_max(0.0, 1.0)
1430     .add_service("mon")
1431     .set_description("how heavily to weight OSD marking itself back up in overall laggy_probability")
1432     .set_long_description("1.0 means that an OSD marking itself back up (because it was marked down but not actually dead) means a 100% laggy_probability; 0.0 effectively disables tracking of laggy_probability."),
1433
1434     Option("mon_osd_laggy_max_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1435     .set_default(300)
1436     .add_service("mon")
1437     .set_description("cap value for period for OSD to be marked for laggy_interval calculation"),
1438
1439     Option("mon_osd_adjust_heartbeat_grace", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1440     .set_default(true)
1441     .add_service("mon")
1442     .set_description("increase OSD heartbeat grace if peers appear to be laggy")
1443     .set_long_description("If an OSD is marked down but then marks itself back up, it implies it wasn't actually down but was unable to respond to heartbeats.  If this option is true, we can use the laggy_probability and laggy_interval values calculated to model this situation to increase the heartbeat grace period for this OSD so that it isn't marked down again.  laggy_probability is an estimated probability that the given OSD is down because it is laggy (not actually down), and laggy_interval is an estiate on how long it stays down when it is laggy.")
1444     .add_see_also("mon_osd_laggy_halflife")
1445     .add_see_also("mon_osd_laggy_weight")
1446     .add_see_also("mon_osd_laggy_max_interval"),
1447
1448     Option("mon_osd_adjust_down_out_interval", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1449     .set_default(true)
1450     .add_service("mon")
1451     .set_description("increase the mon_osd_down_out_interval if an OSD appears to be laggy")
1452     .add_see_also("mon_osd_adjust_heartbeat_grace"),
1453
1454     Option("mon_osd_auto_mark_in", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1455     .set_default(false)
1456     .add_service("mon")
1457     .set_description("mark any OSD that comes up 'in'"),
1458
1459     Option("mon_osd_auto_mark_auto_out_in", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1460     .set_default(true)
1461     .add_service("mon")
1462     .set_description("mark any OSD that comes up that was automatically marked 'out' back 'in'")
1463     .add_see_also("mon_osd_down_out_interval"),
1464
1465     Option("mon_osd_auto_mark_new_in", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1466     .set_default(true)
1467     .add_service("mon")
1468     .set_description("mark any new OSD that comes up 'in'"),
1469
1470     Option("mon_osd_destroyed_out_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1471     .set_default(600)
1472     .add_service("mon")
1473     .set_description("mark any OSD 'out' that has been 'destroy'ed for this long (seconds)"),
1474
1475     Option("mon_osd_down_out_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1476     .set_default(600)
1477     .add_service("mon")
1478     .set_description("mark any OSD 'out' that has been 'down' for this long (seconds)"),
1479
1480     Option("mon_osd_down_out_subtree_limit", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1481     .set_default("rack")
1482     .set_flag(Option::FLAG_RUNTIME)
1483     .add_service("mon")
1484     .set_description("do not automatically mark OSDs 'out' if an entire subtree of this size is down")
1485     .add_see_also("mon_osd_down_out_interval"),
1486
1487     Option("mon_osd_min_up_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1488     .set_default(.3)
1489     .add_service("mon")
1490     .set_description("do not automatically mark OSDs 'out' if fewer than this many OSDs are 'up'")
1491     .add_see_also("mon_osd_down_out_interval"),
1492
1493     Option("mon_osd_min_in_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1494     .set_default(.75)
1495     .add_service("mon")
1496     .set_description("do not automatically mark OSDs 'out' if fewer than this many OSDs are 'in'")
1497     .add_see_also("mon_osd_down_out_interval"),
1498
1499     Option("mon_osd_warn_op_age", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1500     .set_default(32)
1501     .add_service("mgr")
1502     .set_description("issue REQUEST_SLOW health warning if OSD ops are slower than this age (seconds)"),
1503
1504     Option("mon_osd_warn_num_repaired", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1505     .set_default(10)
1506     .add_service("mon")
1507     .set_description("issue OSD_TOO_MANY_REPAIRS health warning if an OSD has more than this many read repairs"),
1508
1509     Option("mon_osd_err_op_age_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1510     .set_default(128)
1511     .add_service("mgr")
1512     .set_description("issue REQUEST_STUCK health error if OSD ops are slower than is age (seconds)"),
1513
1514     Option("mon_osd_prime_pg_temp", Option::TYPE_BOOL, Option::LEVEL_DEV)
1515     .set_default(true)
1516     .add_service("mon")
1517     .set_description("minimize peering work by priming pg_temp values after a map change"),
1518
1519     Option("mon_osd_prime_pg_temp_max_time", Option::TYPE_FLOAT, Option::LEVEL_DEV)
1520     .set_default(.5)
1521     .add_service("mon")
1522     .set_description("maximum time to spend precalculating PG mappings on map change (seconds)"),
1523
1524     Option("mon_osd_prime_pg_temp_max_estimate", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1525     .set_default(.25)
1526     .add_service("mon")
1527     .set_description("calculate all PG mappings if estimated fraction of PGs that change is above this amount"),
1528
1529     Option("mon_stat_smooth_intervals", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1530     .set_default(6)
1531     .set_min(1)
1532     .add_service("mgr")
1533     .set_description("number of PGMaps stats over which we calc the average read/write throughput of the whole cluster"),
1534
1535     Option("mon_election_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1536     .set_default(5)
1537     .add_service("mon")
1538     .set_description("maximum time for a mon election (seconds)"),
1539
1540     Option("mon_lease", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1541     .set_default(5)
1542     .add_service("mon")
1543     .set_description("lease interval between quorum monitors (seconds)")
1544     .set_long_description("This setting controls how sensitive your mon quorum is to intermittent network issues or other failures."),
1545
1546     Option("mon_lease_renew_interval_factor", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1547     .set_default(.6)
1548     .set_min_max((double)0.0, (double).9999999)
1549     .add_service("mon")
1550     .set_description("multiple of mon_lease for the lease renewal interval")
1551     .set_long_description("Leases must be renewed before they time out.  A smaller value means frequent renewals, while a value close to 1 makes a lease expiration more likely.")
1552     .add_see_also("mon_lease"),
1553
1554     Option("mon_lease_ack_timeout_factor", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1555     .set_default(2.0)
1556     .set_min_max(1.0001, 100.0)
1557     .add_service("mon")
1558     .set_description("multiple of mon_lease for the lease ack interval before calling new election")
1559     .add_see_also("mon_lease"),
1560
1561     Option("mon_accept_timeout_factor", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1562     .set_default(2.0)
1563     .add_service("mon")
1564     .set_description("multiple of mon_lease for follower mons to accept proposed state changes before calling a new election")
1565     .add_see_also("mon_lease"),
1566
1567     Option("mon_clock_drift_allowed", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1568     .set_default(.050)
1569     .add_service("mon")
1570     .set_description("allowed clock drift (in seconds) between mons before issuing a health warning"),
1571
1572     Option("mon_clock_drift_warn_backoff", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1573     .set_default(5)
1574     .add_service("mon")
1575     .set_description("exponential backoff factor for logging clock drift warnings in the cluster log"),
1576
1577     Option("mon_timecheck_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1578     .set_default(300.0)
1579     .add_service("mon")
1580     .set_description("frequency of clock synchronization checks between monitors (seconds)"),
1581
1582     Option("mon_timecheck_skew_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1583     .set_default(30.0)
1584     .add_service("mon")
1585     .set_description("frequency of clock synchronization (re)checks between monitors while clocks are believed to be skewed (seconds)")
1586     .add_see_also("mon_timecheck_interval"),
1587
1588     Option("mon_pg_stuck_threshold", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1589     .set_default(60)
1590     .set_description("number of seconds after which pgs can be considered stuck inactive, unclean, etc")
1591     .set_long_description("see doc/control.rst under dump_stuck for more info")
1592     .add_service("mgr"),
1593
1594     Option("mon_pg_warn_min_per_osd", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1595     .set_default(0)
1596     .add_service("mgr")
1597     .set_description("minimal number PGs per (in) osd before we warn the admin"),
1598
1599     Option("mon_max_pg_per_osd", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1600     .set_min(1)
1601     .set_default(250)
1602     .add_service("mgr")
1603     .set_description("Max number of PGs per OSD the cluster will allow")
1604     .set_long_description("If the number of PGs per OSD exceeds this, a "
1605         "health warning will be visible in `ceph status`.  This is also used "
1606         "in automated PG management, as the threshold at which some pools' "
1607         "pg_num may be shrunk in order to enable increasing the pg_num of "
1608         "others."),
1609
1610     Option("mon_target_pg_per_osd", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1611     .set_min(1)
1612     .set_default(100)
1613     .set_description("Automated PG management creates this many PGs per OSD")
1614     .set_long_description("When creating pools, the automated PG management "
1615         "logic will attempt to reach this target.  In some circumstances, it "
1616         "may exceed this target, up to the ``mon_max_pg_per_osd`` limit. "
1617         "Conversely, a lower number of PGs per OSD may be created if the "
1618         "cluster is not yet fully utilised"),
1619
1620     Option("mon_pg_warn_max_object_skew", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1621     .set_default(10.0)
1622     .set_description("max skew few average in objects per pg")
1623     .add_service("mgr"),
1624
1625     Option("mon_pg_warn_min_objects", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1626     .set_default(10000)
1627     .set_description("do not warn below this object #")
1628     .add_service("mgr"),
1629
1630     Option("mon_pg_warn_min_pool_objects", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1631     .set_default(1000)
1632     .set_description("do not warn on pools below this object #")
1633     .add_service("mgr"),
1634
1635     Option("mon_pg_check_down_all_threshold", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1636     .set_default(.5)
1637     .set_description("threshold of down osds after which we check all pgs")
1638     .add_service("mgr"),
1639
1640     Option("mon_cache_target_full_warn_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1641     .set_default(.66)
1642     .add_service("mgr")
1643     .set_flag(Option::FLAG_NO_MON_UPDATE)
1644     .set_flag(Option::FLAG_CLUSTER_CREATE)
1645     .set_description("issue CACHE_POOL_NEAR_FULL health warning when cache pool utilization exceeds this ratio of usable space"),
1646
1647     Option("mon_osd_full_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1648     .set_default(.95)
1649     .set_flag(Option::FLAG_NO_MON_UPDATE)
1650     .set_flag(Option::FLAG_CLUSTER_CREATE)
1651     .set_description("full ratio of OSDs to be set during initial creation of the cluster"),
1652
1653     Option("mon_osd_backfillfull_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1654     .set_default(.90)
1655     .set_flag(Option::FLAG_NO_MON_UPDATE)
1656     .set_flag(Option::FLAG_CLUSTER_CREATE)
1657     .set_description(""),
1658
1659     Option("mon_osd_nearfull_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1660     .set_default(.85)
1661     .set_flag(Option::FLAG_NO_MON_UPDATE)
1662     .set_flag(Option::FLAG_CLUSTER_CREATE)
1663     .set_description("nearfull ratio for OSDs to be set during initial creation of cluster"),
1664
1665     Option("mon_osd_initial_require_min_compat_client", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1666     .set_default("jewel")
1667     .set_flag(Option::FLAG_NO_MON_UPDATE)
1668     .set_flag(Option::FLAG_CLUSTER_CREATE)
1669     .set_description(""),
1670
1671     Option("mon_allow_pool_delete", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1672     .set_default(false)
1673     .add_service("mon")
1674     .set_description("allow pool deletions"),
1675
1676     Option("mon_fake_pool_delete", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1677     .set_default(false)
1678     .add_service("mon")
1679     .set_description("fake pool deletions by renaming the rados pool"),
1680
1681     Option("mon_globalid_prealloc", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1682     .set_default(10000)
1683     .add_service("mon")
1684     .set_description("number of globalid values to preallocate")
1685     .set_long_description("This setting caps how many new clients can authenticate with the cluster before the monitors have to perform a write to preallocate more.  Large values burn through the 64-bit ID space more quickly."),
1686
1687     Option("mon_osd_report_timeout", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1688     .set_default(900)
1689     .add_service("mon")
1690     .set_description("time before OSDs who do not report to the mons are marked down (seconds)"),
1691
1692     Option("mon_warn_on_msgr2_not_enabled", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1693     .set_default(true)
1694     .add_service("mon")
1695     .set_description("issue MON_MSGR2_NOT_ENABLED health warning if monitors are all running Nautilus but not all binding to a msgr2 port")
1696     .add_see_also("ms_bind_msgr2"),
1697
1698     Option("mon_warn_on_legacy_crush_tunables", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1699     .set_default(true)
1700     .add_service("mgr")
1701     .set_description("issue OLD_CRUSH_TUNABLES health warning if CRUSH tunables are older than mon_crush_min_required_version")
1702     .add_see_also("mon_crush_min_required_version"),
1703
1704     Option("mon_crush_min_required_version", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1705     .set_default("hammer")
1706     .add_service("mgr")
1707     .set_description("minimum ceph release to use for mon_warn_on_legacy_crush_tunables")
1708     .add_see_also("mon_warn_on_legacy_crush_tunables"),
1709
1710     Option("mon_warn_on_crush_straw_calc_version_zero", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1711     .set_default(true)
1712     .add_service("mgr")
1713     .set_description("issue OLD_CRUSH_STRAW_CALC_VERSION health warning if the CRUSH map's straw_calc_version is zero"),
1714
1715     Option("mon_warn_on_osd_down_out_interval_zero", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1716     .set_default(true)
1717     .add_service("mgr")
1718     .set_description("issue OSD_NO_DOWN_OUT_INTERVAL health warning if mon_osd_down_out_interval is zero")
1719     .set_long_description("Having mon_osd_down_out_interval set to 0 means that down OSDs are not marked out automatically and the cluster does not heal itself without administrator intervention.")
1720     .add_see_also("mon_osd_down_out_interval"),
1721
1722     Option("mon_warn_on_cache_pools_without_hit_sets", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1723     .set_default(true)
1724     .add_service("mgr")
1725     .set_description("issue CACHE_POOL_NO_HIT_SET health warning for cache pools that do not have hit sets configured"),
1726
1727     Option("mon_warn_on_pool_no_app", Option::TYPE_BOOL, Option::LEVEL_DEV)
1728     .set_default(true)
1729     .add_service("mgr")
1730     .set_description("issue POOL_APP_NOT_ENABLED health warning if pool has not application enabled"),
1731
1732     Option("mon_warn_on_pool_pg_num_not_power_of_two", Option::TYPE_BOOL, Option::LEVEL_DEV)
1733     .set_default(true)
1734     .add_service("mon")
1735     .set_description("issue POOL_PG_NUM_NOT_POWER_OF_TWO warning if pool has a non-power-of-two pg_num value"),
1736
1737     Option("mon_warn_on_pool_no_redundancy", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1738     .set_default(true)
1739     .add_service("mon")
1740     .set_description("Issue a health warning if any pool is configured with no replicas")
1741     .add_see_also("osd_pool_default_size")
1742     .add_see_also("osd_pool_default_min_size"),
1743
1744     Option("mon_warn_on_misplaced", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1745     .set_default(false)
1746     .add_service("mgr")
1747     .set_description("Issue a health warning if there are misplaced objects"),
1748
1749     Option("mon_warn_on_too_few_osds", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1750     .set_default(true)
1751     .add_service("mgr")
1752     .set_description("Issue a health warning if there are fewer OSDs than osd_pool_default_size"),
1753
1754     Option("mon_warn_on_slow_ping_time", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1755     .set_default(0)
1756     .add_service("mgr")
1757     .set_description("Override mon_warn_on_slow_ping_ratio with specified threshold in milliseconds")
1758     .add_see_also("mon_warn_on_slow_ping_ratio"),
1759
1760     Option("mon_warn_on_slow_ping_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1761     .set_default(.05)
1762     .add_service("mgr")
1763     .set_description("Issue a health warning if heartbeat ping longer than percentage of osd_heartbeat_grace")
1764     .add_see_also("osd_heartbeat_grace")
1765     .add_see_also("mon_warn_on_slow_ping_time"),
1766
1767     Option("mon_max_snap_prune_per_epoch", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1768     .set_default(100)
1769     .add_service("mon")
1770     .set_description("max number of pruned snaps we will process in a single OSDMap epoch"),
1771
1772     Option("mon_min_osdmap_epochs", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1773     .set_default(500)
1774     .add_service("mon")
1775     .set_description("min number of OSDMaps to store"),
1776
1777     Option("mon_max_log_epochs", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1778     .set_default(500)
1779     .add_service("mon")
1780     .set_description("max number of past cluster log epochs to store"),
1781
1782     Option("mon_max_mdsmap_epochs", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1783     .set_default(500)
1784     .add_service("mon")
1785     .set_description("max number of FSMaps/MDSMaps to store"),
1786
1787     Option("mon_max_mgrmap_epochs", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1788     .set_default(500)
1789     .add_service("mon")
1790     .set_description("max number of MgrMaps to store"),
1791
1792     Option("mon_max_osd", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1793     .set_default(10000)
1794     .add_service("mon")
1795     .set_description("max number of OSDs in a cluster"),
1796
1797     Option("mon_probe_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1798     .set_default(2.0)
1799     .add_service("mon")
1800     .set_description("timeout for querying other mons during bootstrap pre-election phase (seconds)"),
1801
1802     Option("mon_client_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
1803     .set_default(100ul << 20)
1804     .add_service("mon")
1805     .set_description("max bytes of outstanding client messages mon will read off the network"),
1806
1807     Option("mon_daemon_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
1808     .set_default(400ul << 20)
1809     .add_service("mon")
1810     .set_description("max bytes of outstanding mon messages mon will read off the network"),
1811
1812     Option("mon_mgr_proxy_client_bytes_ratio", Option::TYPE_FLOAT, Option::LEVEL_DEV)
1813     .set_default(.3)
1814     .add_service("mon")
1815     .set_description("ratio of mon_client_bytes that can be consumed by "
1816                      "proxied mgr commands before we error out to client"),
1817
1818     Option("mon_log_max_summary", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1819     .set_default(50)
1820     .add_service("mon")
1821     .set_description("number of recent cluster log messages to retain"),
1822
1823     Option("mon_max_log_entries_per_event", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1824     .set_default(4096)
1825     .add_service("mon")
1826     .set_description("max cluster log entries per paxos event"),
1827
1828     Option("mon_reweight_min_pgs_per_osd", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1829     .set_default(10)
1830     .add_service("mgr")
1831     .set_description(""),
1832
1833     Option("mon_reweight_min_bytes_per_osd", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
1834     .set_default(100_M)
1835     .add_service("mgr")
1836     .set_description(""),
1837
1838     Option("mon_reweight_max_osds", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1839     .set_default(4)
1840     .add_service("mgr")
1841     .set_description(""),
1842
1843     Option("mon_reweight_max_change", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1844     .set_default(0.05)
1845     .add_service("mgr")
1846     .set_description(""),
1847
1848     Option("mon_health_to_clog", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1849     .set_default(true)
1850     .add_service("mon")
1851     .set_description("log monitor health to cluster log"),
1852
1853     Option("mon_health_to_clog_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1854     .set_default(1_hr)
1855     .add_service("mon")
1856     .set_description("frequency to log monitor health to cluster log")
1857     .add_see_also("mon_health_to_clog"),
1858
1859     Option("mon_health_to_clog_tick_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
1860     .set_default(60.0)
1861     .add_service("mon")
1862     .set_description(""),
1863
1864     Option("mon_health_max_detail", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1865     .set_default(50)
1866     .add_service("mon")
1867     .set_description("max detailed pgs to report in health detail"),
1868
1869     Option("mon_health_log_update_period", Option::TYPE_INT, Option::LEVEL_DEV)
1870     .set_default(5)
1871     .add_service("mon")
1872     .set_description("minimum time in seconds between log messages about "
1873                      "each health check")
1874     .set_min(0),
1875
1876     Option("mon_data_avail_crit", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1877     .set_default(5)
1878     .add_service("mon")
1879     .set_description("issue MON_DISK_CRIT health error when mon available space below this percentage"),
1880
1881     Option("mon_data_avail_warn", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1882     .set_default(30)
1883     .add_service("mon")
1884     .set_description("issue MON_DISK_LOW health warning when mon available space below this percentage"),
1885
1886     Option("mon_data_size_warn", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
1887     .set_default(15_G)
1888     .add_service("mon")
1889     .set_description("issue MON_DISK_BIG health warning when mon database is above this size"),
1890
1891     Option("mon_warn_pg_not_scrubbed_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1892     .set_default(0.5)
1893     .set_min(0)
1894     .set_description("Percentage of the scrub max interval past the scrub max interval to warn")
1895     .set_long_description("")
1896     .add_see_also("osd_scrub_max_interval"),
1897
1898     Option("mon_warn_pg_not_deep_scrubbed_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1899     .set_default(0.75)
1900     .set_min(0)
1901     .set_description("Percentage of the deep scrub interval past the deep scrub interval to warn")
1902     .set_long_description("")
1903     .add_see_also("osd_deep_scrub_interval"),
1904
1905     Option("mon_scrub_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1906     .set_default(1_day)
1907     .add_service("mon")
1908     .set_description("frequency for scrubbing mon database"),
1909
1910     Option("mon_scrub_timeout", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1911     .set_default(5_min)
1912     .add_service("mon")
1913     .set_description("timeout to restart scrub of mon quorum participant does not respond for the latest chunk"),
1914
1915     Option("mon_scrub_max_keys", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1916     .set_default(100)
1917     .add_service("mon")
1918     .set_description("max keys per on scrub chunk/step"),
1919
1920     Option("mon_scrub_inject_crc_mismatch", Option::TYPE_FLOAT, Option::LEVEL_DEV)
1921     .set_default(0.0)
1922     .add_service("mon")
1923     .set_description("probability for injecting crc mismatches into mon scrub"),
1924
1925     Option("mon_scrub_inject_missing_keys", Option::TYPE_FLOAT, Option::LEVEL_DEV)
1926     .set_default(0.0)
1927     .add_service("mon")
1928     .set_description("probability for injecting missing keys into mon scrub"),
1929
1930     Option("mon_config_key_max_entry_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
1931     .set_default(64_K)
1932     .add_service("mon")
1933     .set_description("Defines the number of bytes allowed to be held in a "
1934                      "single config-key entry"),
1935
1936     Option("mon_sync_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
1937     .set_default(60.0)
1938     .add_service("mon")
1939     .set_description("timeout before canceling sync if syncing mon does not respond"),
1940
1941     Option("mon_sync_max_payload_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
1942     .set_default(1_M)
1943     .add_service("mon")
1944     .set_description("target max message payload for mon sync"),
1945
1946     Option("mon_sync_max_payload_keys", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1947     .set_default(2000)
1948     .add_service("mon")
1949     .set_description("target max keys in message payload for mon sync"),
1950
1951     Option("mon_sync_debug", Option::TYPE_BOOL, Option::LEVEL_DEV)
1952     .set_default(false)
1953     .add_service("mon")
1954     .set_description("enable extra debugging during mon sync"),
1955
1956     Option("mon_inject_sync_get_chunk_delay", Option::TYPE_FLOAT, Option::LEVEL_DEV)
1957     .set_default(0)
1958     .add_service("mon")
1959     .set_description("inject delay during sync (seconds)"),
1960
1961     Option("mon_osd_min_down_reporters", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
1962     .set_default(2)
1963     .add_service("mon")
1964     .set_description("number of OSDs from different subtrees who need to report a down OSD for it to count")
1965     .add_see_also("mon_osd_reporter_subtree_level"),
1966
1967     Option("mon_osd_reporter_subtree_level", Option::TYPE_STR, Option::LEVEL_ADVANCED)
1968     .set_default("host")
1969     .add_service("mon")
1970     .set_flag(Option::FLAG_RUNTIME)
1971     .set_description("in which level of parent bucket the reporters are counted"),
1972
1973     Option("mon_osd_snap_trim_queue_warn_on", Option::TYPE_INT, Option::LEVEL_ADVANCED)
1974     .set_default(32768)
1975     .add_service("mon")
1976     .set_description("Warn when snap trim queue is that large (or larger).")
1977     .set_long_description("Warn when snap trim queue length for at least one PG crosses this value, as this is indicator of snap trimmer not keeping up, wasting disk space"),
1978
1979     Option("mon_osd_force_trim_to", Option::TYPE_INT, Option::LEVEL_DEV)
1980     .set_default(0)
1981     .add_service("mon")
1982     .set_description("force mons to trim osdmaps through this epoch"),
1983
1984     Option("mon_mds_force_trim_to", Option::TYPE_INT, Option::LEVEL_DEV)
1985     .set_default(0)
1986     .add_service("mon")
1987     .set_description("force mons to trim mdsmaps/fsmaps through this epoch"),
1988
1989     Option("mon_mds_skip_sanity", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
1990     .set_default(false)
1991     .add_service("mon")
1992     .set_description("skip sanity checks on fsmap/mdsmap"),
1993
1994     Option("mon_debug_extra_checks", Option::TYPE_BOOL, Option::LEVEL_DEV)
1995     .set_default(false)
1996     .add_service("mon")
1997     .set_description("Enable some additional monitor checks")
1998     .set_long_description(
1999         "Enable some additional monitor checks that would be too expensive "
2000         "to run on production systems, or would only be relevant while "
2001         "testing or debugging."),
2002
2003     Option("mon_debug_block_osdmap_trim", Option::TYPE_BOOL, Option::LEVEL_DEV)
2004     .set_default(false)
2005     .add_service("mon")
2006     .set_description("Block OSDMap trimming while the option is enabled.")
2007     .set_long_description(
2008         "Blocking OSDMap trimming may be quite helpful to easily reproduce "
2009         "states in which the monitor keeps (hundreds of) thousands of "
2010         "osdmaps."),
2011
2012     Option("mon_debug_deprecated_as_obsolete", Option::TYPE_BOOL, Option::LEVEL_DEV)
2013     .set_default(false)
2014     .add_service("mon")
2015     .set_description("treat deprecated mon commands as obsolete"),
2016
2017     Option("mon_debug_dump_transactions", Option::TYPE_BOOL, Option::LEVEL_DEV)
2018     .set_default(false)
2019     .add_service("mon")
2020     .set_description("dump paxos transactions to log")
2021     .add_see_also("mon_debug_dump_location"),
2022
2023     Option("mon_debug_dump_json", Option::TYPE_BOOL, Option::LEVEL_DEV)
2024     .set_default(false)
2025     .add_service("mon")
2026     .set_description("dump paxos transasctions to log as json")
2027     .add_see_also("mon_debug_dump_transactions"),
2028
2029     Option("mon_debug_dump_location", Option::TYPE_STR, Option::LEVEL_DEV)
2030     .set_default("/var/log/ceph/$cluster-$name.tdump")
2031     .add_service("mon")
2032     .set_description("file to dump paxos transactions to")
2033     .add_see_also("mon_debug_dump_transactions"),
2034
2035     Option("mon_debug_no_require_nautilus", Option::TYPE_BOOL, Option::LEVEL_DEV)
2036     .set_default(false)
2037     .add_service("mon")
2038     .set_flag(Option::FLAG_CLUSTER_CREATE)
2039     .set_description("do not set nautilus feature for new mon clusters"),
2040
2041     Option("mon_debug_no_require_octopus", Option::TYPE_BOOL, Option::LEVEL_DEV)
2042     .set_default(false)
2043     .add_service("mon")
2044     .set_flag(Option::FLAG_CLUSTER_CREATE)
2045     .set_description("do not set octopus feature for new mon clusters"),
2046
2047     Option("mon_debug_no_require_bluestore_for_ec_overwrites", Option::TYPE_BOOL, Option::LEVEL_DEV)
2048     .set_default(false)
2049     .add_service("mon")
2050     .set_description("do not require bluestore OSDs to enable EC overwrites on a rados pool"),
2051
2052     Option("mon_debug_no_initial_persistent_features", Option::TYPE_BOOL, Option::LEVEL_DEV)
2053     .set_default(false)
2054     .add_service("mon")
2055     .set_flag(Option::FLAG_CLUSTER_CREATE)
2056     .set_description("do not set any monmap features for new mon clusters"),
2057
2058     Option("mon_inject_transaction_delay_max", Option::TYPE_FLOAT, Option::LEVEL_DEV)
2059     .set_default(10.0)
2060     .add_service("mon")
2061     .set_description("max duration of injected delay in paxos"),
2062
2063     Option("mon_inject_transaction_delay_probability", Option::TYPE_FLOAT, Option::LEVEL_DEV)
2064     .set_default(0)
2065     .add_service("mon")
2066     .set_description("probability of injecting a delay in paxos"),
2067
2068     Option("mon_inject_pg_merge_bounce_probability", Option::TYPE_FLOAT, Option::LEVEL_DEV)
2069     .set_default(0)
2070     .add_service("mon")
2071     .set_description("probability of failing and reverting a pg_num decrement"),
2072
2073     Option("mon_sync_provider_kill_at", Option::TYPE_INT, Option::LEVEL_DEV)
2074     .set_default(0)
2075     .add_service("mon")
2076     .set_description("kill mon sync requester at specific point"),
2077
2078     Option("mon_sync_requester_kill_at", Option::TYPE_INT, Option::LEVEL_DEV)
2079     .set_default(0)
2080     .add_service("mon")
2081     .set_description("kill mon sync requestor at specific point"),
2082
2083     Option("mon_force_quorum_join", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2084     .set_default(false)
2085     .add_service("mon")
2086     .set_description("force mon to rejoin quorum even though it was just removed"),
2087
2088     Option("mon_keyvaluedb", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2089     .set_default("rocksdb")
2090     .set_enum_allowed({"leveldb", "rocksdb"})
2091     .set_flag(Option::FLAG_CREATE)
2092     .add_service("mon")
2093     .set_description("database backend to use for the mon database"),
2094
2095     Option("mon_debug_unsafe_allow_tier_with_nonempty_snaps", Option::TYPE_BOOL, Option::LEVEL_DEV)
2096     .set_default(false)
2097     .add_service("mon")
2098     .set_description(""),
2099
2100     Option("mon_osd_blacklist_default_expire", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2101     .set_default(1_hr)
2102     .add_service("mon")
2103     .set_description("Duration in seconds that blacklist entries for clients "
2104                      "remain in the OSD map"),
2105
2106     Option("mon_mds_blacklist_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
2107     .set_default(1_day)
2108     .set_min(1_hr)
2109     .add_service("mon")
2110     .set_description("Duration in seconds that blacklist entries for MDS "
2111                      "daemons remain in the OSD map")
2112     .set_flag(Option::FLAG_RUNTIME),
2113
2114     Option("mon_mgr_blacklist_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
2115     .set_default(1_day)
2116     .set_min(1_hr)
2117     .add_service("mon")
2118     .set_description("Duration in seconds that blacklist entries for mgr "
2119                      "daemons remain in the OSD map")
2120     .set_flag(Option::FLAG_RUNTIME),
2121
2122     Option("mon_osd_crush_smoke_test", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2123     .set_default(true)
2124     .add_service("mon")
2125     .set_description("perform a smoke test on any new CRUSH map before accepting changes"),
2126
2127     Option("mon_smart_report_timeout", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2128     .set_default(5)
2129     .add_service("mon")
2130     .set_description("Timeout (in seconds) for smarctl to run, default is set to 5"),
2131
2132
2133     // PAXOS
2134
2135     Option("paxos_stash_full_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2136     .set_default(25)
2137     .add_service("mon")
2138     .set_description(""),
2139
2140     Option("paxos_max_join_drift", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2141     .set_default(10)
2142     .add_service("mon")
2143     .set_description(""),
2144
2145     Option("paxos_propose_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2146     .set_default(1.0)
2147     .add_service("mon")
2148     .set_description(""),
2149
2150     Option("paxos_min_wait", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2151     .set_default(0.05)
2152     .add_service("mon")
2153     .set_description(""),
2154
2155     Option("paxos_min", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2156     .set_default(500)
2157     .add_service("mon")
2158     .set_description(""),
2159
2160     Option("paxos_trim_min", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2161     .set_default(250)
2162     .add_service("mon")
2163     .set_description(""),
2164
2165     Option("paxos_trim_max", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2166     .set_default(500)
2167     .add_service("mon")
2168     .set_description(""),
2169
2170     Option("paxos_service_trim_min", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2171     .set_default(250)
2172     .add_service("mon")
2173     .set_description(""),
2174
2175     Option("paxos_service_trim_max", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2176     .set_default(500)
2177     .add_service("mon")
2178     .set_description(""),
2179
2180     Option("paxos_kill_at", Option::TYPE_INT, Option::LEVEL_DEV)
2181     .set_default(0)
2182     .add_service("mon")
2183     .set_description(""),
2184
2185
2186     // AUTH
2187
2188     Option("auth_cluster_required", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2189     .set_default("cephx")
2190     .set_description("authentication methods required by the cluster"),
2191
2192     Option("auth_service_required", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2193     .set_default("cephx")
2194     .set_description("authentication methods required by service daemons"),
2195
2196     Option("auth_client_required", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2197     .set_default("cephx, none")
2198     .set_flag(Option::FLAG_MINIMAL_CONF)
2199     .set_description("authentication methods allowed by clients"),
2200
2201     Option("auth_supported", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2202     .set_default("")
2203     .set_description("authentication methods required (deprecated)"),
2204
2205     Option("max_rotating_auth_attempts", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2206     .set_default(10)
2207     .set_description("number of attempts to initialize rotating keys before giving up"),
2208
2209     Option("rotating_keys_bootstrap_timeout", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2210     .set_default(30)
2211     .set_description("timeout for obtaining rotating keys during bootstrap phase (seconds)"),
2212
2213     Option("rotating_keys_renewal_timeout", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2214     .set_default(10)
2215     .set_description("timeout for updating rotating keys (seconds)"),
2216
2217     Option("cephx_require_signatures", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2218     .set_default(false)
2219     .set_description(""),
2220
2221     Option("cephx_require_version", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2222     .set_default(1)
2223     .set_description("Cephx version required (1 = pre-mimic, 2 = mimic+)"),
2224
2225     Option("cephx_cluster_require_signatures", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2226     .set_default(false)
2227     .set_description(""),
2228
2229     Option("cephx_cluster_require_version", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2230     .set_default(1)
2231     .set_description("Cephx version required by the cluster from clients (1 = pre-mimic, 2 = mimic+)"),
2232
2233     Option("cephx_service_require_signatures", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2234     .set_default(false)
2235     .set_description(""),
2236
2237     Option("cephx_service_require_version", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2238     .set_default(1)
2239     .set_description("Cephx version required from ceph services (1 = pre-mimic, 2 = mimic+)"),
2240
2241     Option("cephx_sign_messages", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2242     .set_default(true)
2243     .set_description(""),
2244
2245     Option("auth_mon_ticket_ttl", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2246     .set_default(12_hr)
2247     .set_description(""),
2248
2249     Option("auth_service_ticket_ttl", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2250     .set_default(1_hr)
2251     .set_description(""),
2252
2253     Option("auth_debug", Option::TYPE_BOOL, Option::LEVEL_DEV)
2254     .set_default(false)
2255     .set_description(""),
2256
2257     Option("mon_client_hunt_parallel", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2258     .set_default(3)
2259     .set_description(""),
2260
2261     Option("mon_client_hunt_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2262     .set_default(3.0)
2263     .set_description(""),
2264
2265     Option("mon_client_log_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2266     .set_default(1.0)
2267     .set_description("How frequently we send queued cluster log messages to mon"),
2268
2269     Option("mon_client_ping_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2270     .set_default(10.0)
2271     .set_description(""),
2272
2273     Option("mon_client_ping_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2274     .set_default(30.0)
2275     .set_description(""),
2276
2277     Option("mon_client_hunt_interval_backoff", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2278     .set_default(1.5)
2279     .set_description(""),
2280
2281     Option("mon_client_hunt_interval_min_multiple", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2282     .set_default(1.0)
2283     .set_description(""),
2284
2285     Option("mon_client_hunt_interval_max_multiple", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2286     .set_default(10.0)
2287     .set_description(""),
2288
2289     Option("mon_client_max_log_entries_per_message", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2290     .set_default(1000)
2291     .set_description(""),
2292
2293     Option("mon_client_directed_command_retry", Option::TYPE_INT, Option::LEVEL_DEV)
2294     .set_default(2)
2295     .set_description("Number of times to try sending a comamnd directed at a specific monitor"),
2296
2297     Option("mon_max_pool_pg_num", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2298     .set_default(65536)
2299     .set_description(""),
2300
2301     Option("mon_pool_quota_warn_threshold", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2302     .set_default(0)
2303     .set_description("percent of quota at which to issue warnings")
2304     .add_service("mgr"),
2305
2306     Option("mon_pool_quota_crit_threshold", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2307     .set_default(0)
2308     .set_description("percent of quota at which to issue errors")
2309     .add_service("mgr"),
2310
2311     Option("crush_location", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2312     .set_default("")
2313     .set_description(""),
2314
2315     Option("crush_location_hook", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2316     .set_default("")
2317     .set_description(""),
2318
2319     Option("crush_location_hook_timeout", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2320     .set_default(10)
2321     .set_description(""),
2322
2323     Option("objecter_tick_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
2324     .set_default(5.0)
2325     .set_description(""),
2326
2327     Option("objecter_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2328     .set_default(10.0)
2329     .set_description("Seconds before in-flight op is considered 'laggy' and we query mon for the latest OSDMap"),
2330
2331     Option("objecter_inflight_op_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
2332     .set_default(100_M)
2333     .set_description("Max in-flight data in bytes (both directions)"),
2334
2335     Option("objecter_inflight_ops", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2336     .set_default(1024)
2337     .set_description("Max in-flight operations"),
2338
2339     Option("objecter_completion_locks_per_session", Option::TYPE_UINT, Option::LEVEL_DEV)
2340     .set_default(32)
2341     .set_description(""),
2342
2343     Option("objecter_inject_no_watch_ping", Option::TYPE_BOOL, Option::LEVEL_DEV)
2344     .set_default(false)
2345     .set_description(""),
2346
2347     Option("objecter_retry_writes_after_first_reply", Option::TYPE_BOOL, Option::LEVEL_DEV)
2348     .set_default(false)
2349     .set_description(""),
2350
2351     Option("objecter_debug_inject_relock_delay", Option::TYPE_BOOL, Option::LEVEL_DEV)
2352     .set_default(false)
2353     .set_description(""),
2354
2355     Option("filer_max_purge_ops", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2356     .set_default(10)
2357     .set_description("Max in-flight operations for purging a striped range (e.g., MDS journal)"),
2358
2359     Option("filer_max_truncate_ops", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2360     .set_default(128)
2361     .set_description("Max in-flight operations for truncating/deleting a striped sequence (e.g., MDS journal)"),
2362
2363     Option("journaler_write_head_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2364     .set_default(15)
2365     .set_description("Interval in seconds between journal header updates (to help bound replay time)"),
2366
2367     // * journal object size
2368     Option("journaler_prefetch_periods", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2369     .set_default(10)
2370     .set_min(2)                 // we need at least 2 periods to make progress.
2371     .set_description("Number of striping periods to prefetch while reading MDS journal"),
2372
2373     // * journal object size
2374     Option("journaler_prezero_periods", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2375     .set_default(5)
2376     // we need to zero at least two periods, minimum, to ensure that we
2377     // have a full empty object/period in front of us.
2378     .set_min(2)
2379     .set_description("Number of striping periods to zero head of MDS journal write position"),
2380
2381     // -- OSD --
2382     Option("osd_calc_pg_upmaps_aggressively", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2383     .set_default(true)
2384     .set_flag(Option::FLAG_RUNTIME)
2385     .set_description("try to calculate PG upmaps more aggressively, e.g., "
2386                      "by doing a fairly exhaustive search of existing PGs "
2387                      "that can be unmapped or upmapped"),
2388
2389     Option("osd_calc_pg_upmaps_local_fallback_retries", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2390     .set_default(100)
2391     .set_flag(Option::FLAG_RUNTIME)
2392     .set_description("Maximum number of PGs we can attempt to unmap or upmap "
2393                      "for a specific overfull or underfull osd per iteration "),
2394
2395     Option("osd_numa_prefer_iface", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2396     .set_default(true)
2397     .set_flag(Option::FLAG_STARTUP)
2398     .set_description("prefer IP on network interface on same numa node as storage")
2399     .add_see_also("osd_numa_auto_affinity"),
2400
2401     Option("osd_numa_auto_affinity", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2402     .set_default(true)
2403     .set_flag(Option::FLAG_STARTUP)
2404     .set_description("automatically set affinity to numa node when storage and network match"),
2405
2406     Option("osd_numa_node", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2407     .set_default(-1)
2408     .set_flag(Option::FLAG_STARTUP)
2409     .set_description("set affinity to a numa node (-1 for none)")
2410     .add_see_also("osd_numa_auto_affinity"),
2411
2412     Option("osd_smart_report_timeout", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2413     .set_default(5)
2414     .set_description("Timeout (in seconds) for smarctl to run, default is set to 5"),
2415
2416     Option("osd_check_max_object_name_len_on_startup", Option::TYPE_BOOL, Option::LEVEL_DEV)
2417     .set_default(true)
2418     .set_description(""),
2419
2420     Option("osd_max_backfills", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2421     .set_default(1)
2422     .set_description("Maximum number of concurrent local and remote backfills or recoveries per OSD ")
2423     .set_long_description("There can be osd_max_backfills local reservations AND the same remote reservations per OSD. So a value of 1 lets this OSD participate as 1 PG primary in recovery and 1 shard of another recovering PG."),
2424
2425     Option("osd_min_recovery_priority", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2426     .set_default(0)
2427     .set_description("Minimum priority below which recovery is not performed")
2428     .set_long_description("The purpose here is to prevent the cluster from doing *any* lower priority work (e.g., rebalancing) below this threshold and focus solely on higher priority work (e.g., replicating degraded objects)."),
2429
2430     Option("osd_backfill_retry_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2431     .set_default(30.0)
2432     .set_description("how frequently to retry backfill reservations after being denied (e.g., due to a full OSD)"),
2433
2434     Option("osd_recovery_retry_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2435     .set_default(30.0)
2436     .set_description("how frequently to retry recovery reservations after being denied (e.g., due to a full OSD)"),
2437
2438     Option("osd_agent_max_ops", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2439     .set_default(4)
2440     .set_description("maximum concurrent tiering operations for tiering agent"),
2441
2442     Option("osd_agent_max_low_ops", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2443     .set_default(2)
2444     .set_description("maximum concurrent low-priority tiering operations for tiering agent"),
2445
2446     Option("osd_agent_min_evict_effort", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2447     .set_default(.1)
2448     .set_min_max(0.0, .99)
2449     .set_description("minimum effort to expend evicting clean objects"),
2450
2451     Option("osd_agent_quantize_effort", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2452     .set_default(.1)
2453     .set_description("size of quantize unit for eviction effort"),
2454
2455     Option("osd_agent_delay_time", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2456     .set_default(5.0)
2457     .set_description("how long agent should sleep if it has no work to do"),
2458
2459     Option("osd_find_best_info_ignore_history_les", Option::TYPE_BOOL, Option::LEVEL_DEV)
2460     .set_default(false)
2461     .set_description("ignore last_epoch_started value when peering AND PROBABLY LOSE DATA")
2462     .set_long_description("THIS IS AN EXTREMELY DANGEROUS OPTION THAT SHOULD ONLY BE USED AT THE DIRECTION OF A DEVELOPER.  It makes peering ignore the last_epoch_started value when peering, which can allow the OSD to believe an OSD has an authoritative view of a PG's contents even when it is in fact old and stale, typically leading to data loss (by believing a stale PG is up to date)."),
2463
2464     Option("osd_agent_hist_halflife", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2465     .set_default(1000)
2466     .set_description("halflife of agent atime and temp histograms"),
2467
2468     Option("osd_agent_slop", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2469     .set_default(.02)
2470     .set_description("slop factor to avoid switching tiering flush and eviction mode"),
2471
2472     Option("osd_uuid", Option::TYPE_UUID, Option::LEVEL_ADVANCED)
2473     .set_default(uuid_d())
2474     .set_flag(Option::FLAG_CREATE)
2475     .set_description("uuid label for a new OSD"),
2476
2477     Option("osd_data", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2478     .set_default("/var/lib/ceph/osd/$cluster-$id")
2479     .set_flag(Option::FLAG_NO_MON_UPDATE)
2480     .set_description("path to OSD data"),
2481
2482     Option("osd_journal", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2483     .set_default("/var/lib/ceph/osd/$cluster-$id/journal")
2484     .set_flag(Option::FLAG_NO_MON_UPDATE)
2485     .set_description("path to OSD journal (when FileStore backend is in use)"),
2486
2487     Option("osd_journal_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
2488     .set_default(5120)
2489     .set_flag(Option::FLAG_CREATE)
2490     .set_description("size of FileStore journal (in MiB)"),
2491
2492     Option("osd_journal_flush_on_shutdown", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2493     .set_default(true)
2494     .set_description("flush FileStore journal contents during clean OSD shutdown"),
2495
2496     Option("osd_os_flags", Option::TYPE_UINT, Option::LEVEL_DEV)
2497     .set_default(0)
2498     .set_description("flags to skip filestore omap or journal initialization"),
2499
2500     Option("osd_max_write_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
2501     .set_min(4)
2502     .set_default(90)
2503     .set_description("Maximum size of a RADOS write operation in megabytes")
2504     .set_long_description("This setting prevents clients from doing "
2505         "very large writes to RADOS.  If you set this to a value "
2506         "below what clients expect, they will receive an error "
2507         "when attempting to write to the cluster."),
2508
2509     Option("osd_max_pgls", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2510     .set_default(1024)
2511     .set_description("maximum number of results when listing objects in a pool"),
2512
2513     Option("osd_client_message_size_cap", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
2514     .set_default(500_M)
2515     .set_description("maximum memory to devote to in-flight client requests")
2516     .set_long_description("If this value is exceeded, the OSD will not read any new client data off of the network until memory is freed."),
2517
2518     Option("osd_client_message_cap", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2519     .set_default(0)
2520     .set_description("maximum number of in-flight client requests"),
2521
2522     Option("osd_crush_update_weight_set", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2523     .set_default(true)
2524     .set_description("update CRUSH weight-set weights when updating weights")
2525     .set_long_description("If this setting is true, we will update the weight-set weights when adjusting an item's weight, effectively making changes take effect immediately, and discarding any previous optimization in the weight-set value.  Setting this value to false will leave it to the balancer to (slowly, presumably) adjust weights to approach the new target value."),
2526
2527     Option("osd_crush_chooseleaf_type", Option::TYPE_INT, Option::LEVEL_DEV)
2528     .set_default(1)
2529     .set_flag(Option::FLAG_CLUSTER_CREATE)
2530     .set_description("default chooseleaf type for osdmaptool --create"),
2531
2532     Option("osd_pool_use_gmt_hitset", Option::TYPE_BOOL, Option::LEVEL_DEV)
2533     .set_default(true)
2534     .set_description("use UTC for hitset timestamps")
2535     .set_long_description("This setting only exists for compatibility with hammer (and older) clusters."),
2536
2537     Option("osd_crush_update_on_start", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2538     .set_default(true)
2539     .set_description("update OSD CRUSH location on startup"),
2540
2541     Option("osd_class_update_on_start", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2542     .set_default(true)
2543     .set_description("set OSD device class on startup"),
2544
2545     Option("osd_crush_initial_weight", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2546     .set_default(-1)
2547     .set_description("if >= 0, initial CRUSH weight for newly created OSDs")
2548     .set_long_description("If this value is negative, the size of the OSD in TiB is used."),
2549
2550     Option("osd_pool_default_ec_fast_read", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2551     .set_default(false)
2552     .set_description("set ec_fast_read for new erasure-coded pools")
2553     .add_service("mon"),
2554
2555     Option("osd_pool_default_crush_rule", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2556     .set_default(-1)
2557     .set_description("CRUSH rule for newly created pools")
2558     .add_service("mon"),
2559
2560     Option("osd_pool_erasure_code_stripe_unit", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
2561     .set_default(4_K)
2562     .set_description("the amount of data (in bytes) in a data chunk, per stripe")
2563     .add_service("mon"),
2564
2565     Option("osd_pool_default_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2566     .set_default(3)
2567     .set_min_max(0, 10)
2568     .set_flag(Option::FLAG_RUNTIME)
2569     .set_description("the number of copies of an object for new replicated pools")
2570     .add_service("mon"),
2571
2572     Option("osd_pool_default_min_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2573     .set_default(0)
2574     .set_min_max(0, 255)
2575     .set_flag(Option::FLAG_RUNTIME)
2576     .set_description("the minimal number of copies allowed to write to a degraded pool for new replicated pools")
2577     .set_long_description("0 means no specific default; ceph will use size-size/2")
2578     .add_see_also("osd_pool_default_size")
2579     .add_service("mon"),
2580
2581     Option("osd_pool_default_pg_num", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2582     .set_default(32)
2583     .set_description("number of PGs for new pools")
2584     .set_flag(Option::FLAG_RUNTIME)
2585     .add_service("mon"),
2586
2587     Option("osd_pool_default_pgp_num", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2588     .set_default(0)
2589     .set_description("number of PGs for placement purposes (0 to match pg_num)")
2590     .add_see_also("osd_pool_default_pg_num")
2591     .set_flag(Option::FLAG_RUNTIME)
2592     .add_service("mon"),
2593
2594     Option("osd_pool_default_type", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2595     .set_default("replicated")
2596     .set_enum_allowed({"replicated", "erasure"})
2597     .set_flag(Option::FLAG_RUNTIME)
2598     .set_description("default type of pool to create")
2599     .add_service("mon"),
2600
2601     Option("osd_pool_default_erasure_code_profile", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2602     .set_default("plugin=jerasure technique=reed_sol_van k=2 m=2")
2603     .set_flag(Option::FLAG_RUNTIME)
2604     .set_description("default erasure code profile for new erasure-coded pools")
2605     .add_service("mon"),
2606
2607     Option("osd_erasure_code_plugins", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2608     .set_default("jerasure lrc"
2609   #ifdef HAVE_BETTER_YASM_ELF64
2610          " isa"
2611   #endif
2612         )
2613     .set_flag(Option::FLAG_STARTUP)
2614     .set_description("erasure code plugins to load")
2615     .add_service("mon")
2616     .add_service("osd"),
2617
2618     Option("osd_allow_recovery_below_min_size", Option::TYPE_BOOL, Option::LEVEL_DEV)
2619     .set_default(true)
2620     .set_description("allow replicated pools to recover with < min_size active members")
2621     .add_service("osd"),
2622
2623     Option("osd_pool_default_flags", Option::TYPE_INT, Option::LEVEL_DEV)
2624     .set_default(0)
2625     .set_description("(integer) flags to set on new pools")
2626     .add_service("mon"),
2627
2628     Option("osd_pool_default_flag_hashpspool", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2629     .set_default(true)
2630     .set_description("set hashpspool (better hashing scheme) flag on new pools")
2631     .add_service("mon"),
2632
2633     Option("osd_pool_default_flag_nodelete", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2634     .set_default(false)
2635     .set_description("set nodelete flag on new pools")
2636     .add_service("mon"),
2637
2638     Option("osd_pool_default_flag_nopgchange", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2639     .set_default(false)
2640     .set_description("set nopgchange flag on new pools")
2641     .add_service("mon"),
2642
2643     Option("osd_pool_default_flag_nosizechange", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2644     .set_default(false)
2645     .set_description("set nosizechange flag on new pools")
2646     .add_service("mon"),
2647
2648     Option("osd_pool_default_hit_set_bloom_fpp", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2649     .set_default(.05)
2650     .set_description("")
2651     .add_see_also("osd_tier_default_cache_hit_set_type")
2652     .add_service("mon"),
2653
2654     Option("osd_pool_default_cache_target_dirty_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2655     .set_default(.4)
2656     .set_description(""),
2657
2658     Option("osd_pool_default_cache_target_dirty_high_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2659     .set_default(.6)
2660     .set_description(""),
2661
2662     Option("osd_pool_default_cache_target_full_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2663     .set_default(.8)
2664     .set_description(""),
2665
2666     Option("osd_pool_default_cache_min_flush_age", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2667     .set_default(0)
2668     .set_description(""),
2669
2670     Option("osd_pool_default_cache_min_evict_age", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2671     .set_default(0)
2672     .set_description(""),
2673
2674     Option("osd_pool_default_cache_max_evict_check_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2675     .set_default(10)
2676     .set_description(""),
2677
2678     Option("osd_pool_default_pg_autoscale_mode", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2679     .set_default("on")
2680     .set_flag(Option::FLAG_RUNTIME)
2681     .set_enum_allowed({"off", "warn", "on"})
2682     .set_description("Default PG autoscaling behavior for new pools"),
2683
2684     Option("osd_pool_default_read_lease_ratio", Option::TYPE_FLOAT, Option::LEVEL_DEV)
2685     .set_default(.8)
2686     .set_flag(Option::FLAG_RUNTIME)
2687     .set_description("Default read_lease_ratio for a pool, as a multiple of osd_heartbeat_grace")
2688     .set_long_description("This should be <= 1.0 so that the read lease will have expired by the time we decide to mark a peer OSD down.")
2689     .add_see_also("osd_heartbeat_grace"),
2690
2691     Option("osd_hit_set_min_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2692     .set_default(1000)
2693     .set_description(""),
2694
2695     Option("osd_hit_set_max_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2696     .set_default(100000)
2697     .set_description(""),
2698
2699     Option("osd_hit_set_namespace", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2700     .set_default(".ceph-internal")
2701     .set_description(""),
2702
2703     Option("osd_tier_promote_max_objects_sec", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2704     .set_default(25)
2705     .set_description(""),
2706
2707     Option("osd_tier_promote_max_bytes_sec", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
2708     .set_default(5_M)
2709     .set_description(""),
2710
2711     Option("osd_tier_default_cache_mode", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2712     .set_default("writeback")
2713     .set_enum_allowed({"none", "writeback", "forward",
2714                        "readonly", "readforward", "readproxy", "proxy"})
2715     .set_flag(Option::FLAG_RUNTIME)
2716     .set_description(""),
2717
2718     Option("osd_tier_default_cache_hit_set_count", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2719     .set_default(4)
2720     .set_description(""),
2721
2722     Option("osd_tier_default_cache_hit_set_period", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2723     .set_default(1200)
2724     .set_description(""),
2725
2726     Option("osd_tier_default_cache_hit_set_type", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2727     .set_default("bloom")
2728     .set_enum_allowed({"bloom", "explicit_hash", "explicit_object"})
2729     .set_flag(Option::FLAG_RUNTIME)
2730     .set_description(""),
2731
2732     Option("osd_tier_default_cache_min_read_recency_for_promote", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2733     .set_default(1)
2734     .set_description("number of recent HitSets the object must appear in to be promoted (on read)"),
2735
2736     Option("osd_tier_default_cache_min_write_recency_for_promote", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2737     .set_default(1)
2738     .set_description("number of recent HitSets the object must appear in to be promoted (on write)"),
2739
2740     Option("osd_tier_default_cache_hit_set_grade_decay_rate", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2741     .set_default(20)
2742     .set_description(""),
2743
2744     Option("osd_tier_default_cache_hit_set_search_last_n", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2745     .set_default(1)
2746     .set_description(""),
2747
2748     Option("osd_objecter_finishers", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2749     .set_default(1)
2750     .set_flag(Option::FLAG_STARTUP)
2751     .set_description(""),
2752
2753     Option("osd_map_dedup", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2754     .set_default(true)
2755     .set_description(""),
2756
2757     Option("osd_map_cache_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2758     .set_default(50)
2759     .set_description(""),
2760
2761     Option("osd_map_message_max", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2762     .set_default(40)
2763     .set_description("maximum number of OSDMaps to include in a single message"),
2764
2765     Option("osd_map_message_max_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
2766     .set_default(10_M)
2767     .set_description("maximum number of bytes worth of OSDMaps to include in a single message"),
2768
2769     Option("osd_map_share_max_epochs", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2770     .set_default(40)
2771     .set_description(""),
2772
2773     Option("osd_pg_epoch_max_lag_factor", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2774     .set_default(2.0)
2775     .set_description("Max multiple of the map cache that PGs can lag before we throttle map injest")
2776     .add_see_also("osd_map_cache_size"),
2777
2778     Option("osd_inject_bad_map_crc_probability", Option::TYPE_FLOAT, Option::LEVEL_DEV)
2779     .set_default(0)
2780     .set_description(""),
2781
2782     Option("osd_inject_failure_on_pg_removal", Option::TYPE_BOOL, Option::LEVEL_DEV)
2783     .set_default(false)
2784     .set_description(""),
2785
2786     Option("osd_max_markdown_period", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2787     .set_default(600)
2788     .set_description(""),
2789
2790     Option("osd_max_markdown_count", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2791     .set_default(5)
2792     .set_description(""),
2793
2794     Option("osd_op_pq_max_tokens_per_priority", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2795     .set_default(4194304)
2796     .set_description(""),
2797
2798     Option("osd_op_pq_min_cost", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
2799     .set_default(65536)
2800     .set_description(""),
2801
2802     Option("osd_recover_clone_overlap", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2803     .set_default(true)
2804     .set_description(""),
2805
2806     Option("osd_num_cache_shards", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
2807     .set_default(32)
2808     .set_flag(Option::FLAG_STARTUP)
2809     .set_description("The number of cache shards to use in the object store."),
2810
2811     Option("osd_op_num_threads_per_shard", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2812     .set_default(0)
2813     .set_flag(Option::FLAG_STARTUP)
2814     .set_description(""),
2815
2816     Option("osd_op_num_threads_per_shard_hdd", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2817     .set_default(1)
2818     .set_flag(Option::FLAG_STARTUP)
2819     .set_description("")
2820     .add_see_also("osd_op_num_threads_per_shard"),
2821
2822     Option("osd_op_num_threads_per_shard_ssd", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2823     .set_default(2)
2824     .set_flag(Option::FLAG_STARTUP)
2825     .set_description("")
2826     .add_see_also("osd_op_num_threads_per_shard"),
2827
2828     Option("osd_op_num_shards", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2829     .set_default(0)
2830     .set_flag(Option::FLAG_STARTUP)
2831     .set_description(""),
2832
2833     Option("osd_op_num_shards_hdd", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2834     .set_default(5)
2835     .set_flag(Option::FLAG_STARTUP)
2836     .set_description("")
2837     .add_see_also("osd_op_num_shards"),
2838
2839     Option("osd_op_num_shards_ssd", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2840     .set_default(8)
2841     .set_flag(Option::FLAG_STARTUP)
2842     .set_description("")
2843     .add_see_also("osd_op_num_shards"),
2844
2845     Option("osd_skip_data_digest", Option::TYPE_BOOL, Option::LEVEL_DEV)
2846     .set_default(false)
2847     .set_description("Do not store full-object checksums if the backend (bluestore) does its own checksums.  Only usable with all BlueStore OSDs."),
2848
2849     Option("osd_op_queue", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2850     .set_default("wpq")
2851     .set_enum_allowed( { "wpq", "prioritized",
2852           "mclock_opclass", "mclock_client", "mclock_scheduler",
2853           "debug_random" } )
2854     .set_description("which operation priority queue algorithm to use")
2855     .set_long_description("which operation priority queue algorithm to use; "
2856                           "mclock_opclass mclock_client, and "
2857                           "mclock_client_profile are currently experimental")
2858     .add_see_also("osd_op_queue_cut_off"),
2859
2860     Option("osd_op_queue_cut_off", Option::TYPE_STR, Option::LEVEL_ADVANCED)
2861     .set_default("high")
2862     .set_enum_allowed( { "low", "high", "debug_random" } )
2863     .set_description("the threshold between high priority ops and low priority ops")
2864     .set_long_description("the threshold between high priority ops that use strict priority ordering and low priority ops that use a fairness algorithm that may or may not incorporate priority")
2865     .add_see_also("osd_op_queue"),
2866
2867     Option("osd_mclock_scheduler_client_res", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2868     .set_default(1)
2869     .set_description("IO proportion reserved for each client (default)")
2870     .set_long_description("Only considered for osd_op_queue = mClockScheduler")
2871     .add_see_also("osd_op_queue"),
2872
2873     Option("osd_mclock_scheduler_client_wgt", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2874     .set_default(1)
2875     .set_description("IO share for each client (default) over reservation")
2876     .set_long_description("Only considered for osd_op_queue = mClockScheduler")
2877     .add_see_also("osd_op_queue"),
2878
2879     Option("osd_mclock_scheduler_client_lim", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2880     .set_default(999999)
2881     .set_description("IO limit for each client (default) over reservation")
2882     .set_long_description("Only considered for osd_op_queue = mClockScheduler")
2883     .add_see_also("osd_op_queue"),
2884
2885     Option("osd_mclock_scheduler_background_recovery_res", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2886     .set_default(1)
2887     .set_description("IO proportion reserved for background recovery (default)")
2888     .set_long_description("Only considered for osd_op_queue = mClockScheduler")
2889     .add_see_also("osd_op_queue"),
2890
2891     Option("osd_mclock_scheduler_background_recovery_wgt", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2892     .set_default(1)
2893     .set_description("IO share for each background recovery over reservation")
2894     .set_long_description("Only considered for osd_op_queue = mClockScheduler")
2895     .add_see_also("osd_op_queue"),
2896
2897     Option("osd_mclock_scheduler_background_recovery_lim", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2898     .set_default(999999)
2899     .set_description("IO limit for background recovery over reservation")
2900     .set_long_description("Only considered for osd_op_queue = mClockScheduler")
2901     .add_see_also("osd_op_queue"),
2902
2903     Option("osd_mclock_scheduler_background_best_effort_res", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2904     .set_default(1)
2905     .set_description("IO proportion reserved for background best_effort (default)")
2906     .set_long_description("Only considered for osd_op_queue = mClockScheduler")
2907     .add_see_also("osd_op_queue"),
2908
2909     Option("osd_mclock_scheduler_background_best_effort_wgt", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2910     .set_default(1)
2911     .set_description("IO share for each background best_effort over reservation")
2912     .set_long_description("Only considered for osd_op_queue = mClockScheduler")
2913     .add_see_also("osd_op_queue"),
2914
2915     Option("osd_mclock_scheduler_background_best_effort_lim", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2916     .set_default(999999)
2917     .set_description("IO limit for background best_effort over reservation")
2918     .set_long_description("Only considered for osd_op_queue = mClockScheduler")
2919     .add_see_also("osd_op_queue"),
2920
2921     Option("osd_mclock_scheduler_anticipation_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2922     .set_default(0.0)
2923     .set_description("mclock anticipation timeout in seconds")
2924     .set_long_description("the amount of time that mclock waits until the unused resource is forfeited"),
2925
2926     Option("osd_ignore_stale_divergent_priors", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2927     .set_default(false)
2928     .set_description(""),
2929
2930     Option("osd_read_ec_check_for_errors", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2931     .set_default(false)
2932     .set_description(""),
2933
2934     // Only use clone_overlap for recovery if there are fewer than
2935     // osd_recover_clone_overlap_limit entries in the overlap set
2936     Option("osd_recover_clone_overlap_limit", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
2937     .set_default(10)
2938     .set_description("")
2939     .set_flag(Option::FLAG_RUNTIME),
2940
2941     Option("osd_debug_feed_pullee", Option::TYPE_INT, Option::LEVEL_DEV)
2942     .set_default(-1)
2943     .set_description("Feed a pullee, and force primary to pull "
2944                      "a currently missing object from it"),
2945
2946     Option("osd_backfill_scan_min", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2947     .set_default(64)
2948     .set_description(""),
2949
2950     Option("osd_backfill_scan_max", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2951     .set_default(512)
2952     .set_description(""),
2953
2954     Option("osd_op_thread_timeout", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2955     .set_default(15)
2956     .set_description(""),
2957
2958     Option("osd_op_thread_suicide_timeout", Option::TYPE_INT, Option::LEVEL_ADVANCED)
2959     .set_default(150)
2960     .set_description(""),
2961
2962     Option("osd_recovery_sleep", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2963     .set_default(0)
2964     .set_description("Time in seconds to sleep before next recovery or backfill op"),
2965
2966     Option("osd_recovery_sleep_hdd", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2967     .set_default(0.1)
2968     .set_description("Time in seconds to sleep before next recovery or backfill op for HDDs"),
2969
2970     Option("osd_recovery_sleep_ssd", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2971     .set_default(0)
2972     .set_description("Time in seconds to sleep before next recovery or backfill op for SSDs")
2973     .add_see_also("osd_recovery_sleep"),
2974
2975     Option("osd_recovery_sleep_hybrid", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2976     .set_default(0.025)
2977     .set_description("Time in seconds to sleep before next recovery or backfill op when data is on HDD and journal is on SSD")
2978     .add_see_also("osd_recovery_sleep"),
2979
2980     Option("osd_snap_trim_sleep", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2981     .set_default(0)
2982     .set_description("Time in seconds to sleep before next snap trim (overrides values below)"),
2983
2984     Option("osd_snap_trim_sleep_hdd", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2985     .set_default(5)
2986     .set_description("Time in seconds to sleep before next snap trim for HDDs"),
2987
2988     Option("osd_snap_trim_sleep_ssd", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2989     .set_default(0)
2990     .set_description("Time in seconds to sleep before next snap trim for SSDs"),
2991
2992     Option("osd_snap_trim_sleep_hybrid", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
2993     .set_default(2)
2994     .set_description("Time in seconds to sleep before next snap trim when data is on HDD and journal is on SSD"),
2995
2996     Option("osd_scrub_invalid_stats", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
2997     .set_default(true)
2998     .set_description(""),
2999
3000     Option("osd_command_thread_timeout", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3001     .set_default(10_min)
3002     .set_description(""),
3003
3004     Option("osd_command_thread_suicide_timeout", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3005     .set_default(15_min)
3006     .set_description(""),
3007
3008     Option("osd_heartbeat_interval", Option::TYPE_INT, Option::LEVEL_DEV)
3009     .set_default(6)
3010     .set_min_max(1, 60)
3011     .set_description("Interval (in seconds) between peer pings"),
3012
3013     Option("osd_heartbeat_grace", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3014     .set_default(20)
3015     .set_description(""),
3016
3017     Option("osd_heartbeat_stale", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3018     .set_default(600)
3019     .set_description("Interval (in seconds) we mark an unresponsive heartbeat peer as stale.")
3020     .set_long_description("Automatically mark unresponsive heartbeat sessions as stale and tear them down. "
3021                           "The primary benefit is that OSD doesn't need to keep a flood of "
3022                           "blocked heartbeat messages around in memory."),
3023
3024     Option("osd_heartbeat_min_peers", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3025     .set_default(10)
3026     .set_description(""),
3027
3028     Option("osd_heartbeat_use_min_delay_socket", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3029     .set_default(false)
3030     .set_description(""),
3031
3032     Option("osd_heartbeat_min_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3033     .set_default(2000)
3034     .set_description("Minimum heartbeat packet size in bytes. Will add dummy payload if heartbeat packet is smaller than this."),
3035
3036     Option("osd_pg_max_concurrent_snap_trims", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3037     .set_default(2)
3038     .set_description(""),
3039
3040     Option("osd_max_trimming_pgs", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3041     .set_default(2)
3042     .set_description(""),
3043
3044     Option("osd_heartbeat_min_healthy_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3045     .set_default(.33)
3046     .set_description(""),
3047
3048     Option("osd_mon_heartbeat_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3049     .set_default(30)
3050     .set_description(""),
3051
3052     Option("osd_mon_heartbeat_stat_stale", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3053     .set_default(1_hr)
3054     .set_description("Stop reporting on heartbeat ping times not updated for this many seconds.")
3055     .set_long_description("Stop reporting on old heartbeat information unless this is set to zero"),
3056
3057     Option("osd_mon_report_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3058     .set_default(5)
3059     .set_description("Frequency of OSD reports to mon for peer failures, fullness status changes"),
3060
3061     Option("osd_mon_report_max_in_flight", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3062     .set_default(2)
3063     .set_description(""),
3064
3065     Option("osd_beacon_report_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3066     .set_default(300)
3067     .set_description(""),
3068
3069     Option("osd_pg_stat_report_interval_max", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3070     .set_default(500)
3071     .set_description(""),
3072
3073     Option("osd_mon_ack_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3074     .set_default(30.0)
3075     .set_description(""),
3076
3077     Option("osd_stats_ack_timeout_factor", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3078     .set_default(2.0)
3079     .set_description(""),
3080
3081     Option("osd_stats_ack_timeout_decay", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3082     .set_default(.9)
3083     .set_description(""),
3084
3085     Option("osd_max_snap_prune_intervals_per_epoch", Option::TYPE_UINT, Option::LEVEL_DEV)
3086     .set_default(512)
3087     .set_description("Max number of snap intervals to report to mgr in pg_stat_t"),
3088
3089     Option("osd_default_data_pool_replay_window", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3090     .set_default(45)
3091     .set_description(""),
3092
3093     Option("osd_auto_mark_unfound_lost", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3094     .set_default(false)
3095     .set_description(""),
3096
3097     Option("osd_recovery_delay_start", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3098     .set_default(0)
3099     .set_description(""),
3100
3101     Option("osd_recovery_max_active", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3102     .set_default(0)
3103     .set_description("Number of simultaneous active recovery operations per OSD (overrides _ssd and _hdd if non-zero)")
3104     .add_see_also("osd_recovery_max_active_hdd")
3105     .add_see_also("osd_recovery_max_active_ssd"),
3106
3107     Option("osd_recovery_max_active_hdd", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3108     .set_default(3)
3109     .set_description("Number of simultaneous active recovery oeprations per OSD (for rotational devices)")
3110     .add_see_also("osd_recovery_max_active")
3111     .add_see_also("osd_recovery_max_active_ssd"),
3112
3113     Option("osd_recovery_max_active_ssd", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3114     .set_default(10)
3115     .set_description("Number of simultaneous active recovery oeprations per OSD (for non-rotational solid state devices)")
3116     .add_see_also("osd_recovery_max_active")
3117     .add_see_also("osd_recovery_max_active_hdd"),
3118
3119     Option("osd_recovery_max_single_start", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3120     .set_default(1)
3121     .set_description(""),
3122
3123     Option("osd_recovery_max_chunk", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3124     .set_default(8_M)
3125     .set_description(""),
3126
3127     Option("osd_recovery_max_omap_entries_per_chunk", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3128     .set_default(8096)
3129     .set_description(""),
3130
3131     Option("osd_copyfrom_max_chunk", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3132     .set_default(8_M)
3133     .set_description(""),
3134
3135     Option("osd_push_per_object_cost", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3136     .set_default(1000)
3137     .set_description(""),
3138
3139     Option("osd_max_push_cost", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3140     .set_default(8<<20)
3141     .set_description(""),
3142
3143     Option("osd_max_push_objects", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3144     .set_default(10)
3145     .set_description(""),
3146
3147     Option("osd_max_scrubs", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3148     .set_default(1)
3149     .set_description("Maximum concurrent scrubs on a single OSD"),
3150
3151     Option("osd_scrub_during_recovery", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3152     .set_default(false)
3153     .set_description("Allow scrubbing when PGs on the OSD are undergoing recovery"),
3154
3155     Option("osd_repair_during_recovery", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3156     .set_default(false)
3157     .set_description("Allow requested repairing when PGs on the OSD are undergoing recovery"),
3158
3159     Option("osd_scrub_begin_hour", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3160     .set_default(0)
3161     .set_description("Restrict scrubbing to this hour of the day or later")
3162     .add_see_also("osd_scrub_end_hour"),
3163
3164     Option("osd_scrub_end_hour", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3165     .set_default(24)
3166     .set_description("Restrict scrubbing to hours of the day earlier than this")
3167     .add_see_also("osd_scrub_begin_hour"),
3168
3169     Option("osd_scrub_begin_week_day", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3170     .set_default(0)
3171     .set_description("Restrict scrubbing to this day of the week or later")
3172     .set_long_description("0 or 7 = Sunday, 1 = Monday, etc.")
3173     .add_see_also("osd_scrub_end_week_day"),
3174
3175     Option("osd_scrub_end_week_day", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3176     .set_default(7)
3177     .set_description("Restrict scrubbing to days of the week earlier than this")
3178     .set_long_description("0 or 7 = Sunday, 1 = Monday, etc.")
3179     .add_see_also("osd_scrub_begin_week_day"),
3180
3181     Option("osd_scrub_load_threshold", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3182     .set_default(0.5)
3183     .set_description("Allow scrubbing when system load divided by number of CPUs is below this value"),
3184
3185     Option("osd_scrub_min_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3186     .set_default(1_day)
3187     .set_description("Scrub each PG no more often than this interval")
3188     .add_see_also("osd_scrub_max_interval"),
3189
3190     Option("osd_scrub_max_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3191     .set_default(7_day)
3192     .set_description("Scrub each PG no less often than this interval")
3193     .add_see_also("osd_scrub_min_interval"),
3194
3195     Option("osd_scrub_interval_randomize_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3196     .set_default(0.5)
3197     .set_description("Ratio of scrub interval to randomly vary")
3198     .set_long_description("This prevents a scrub 'stampede' by randomly varying the scrub intervals so that they are soon uniformly distributed over the week")
3199     .add_see_also("osd_scrub_min_interval"),
3200
3201     Option("osd_scrub_backoff_ratio", Option::TYPE_FLOAT, Option::LEVEL_DEV)
3202     .set_default(.66)
3203     .set_long_description("This is the precentage of ticks that do NOT schedule scrubs, 66% means that 1 out of 3 ticks will schedule scrubs")
3204     .set_description("Backoff ratio for scheduling scrubs"),
3205
3206     Option("osd_scrub_chunk_min", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3207     .set_default(5)
3208     .set_description("Minimum number of objects to scrub in a single chunk")
3209     .add_see_also("osd_scrub_chunk_max"),
3210
3211     Option("osd_scrub_chunk_max", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3212     .set_default(25)
3213     .set_description("Maximum number of objects to scrub in a single chunk")
3214     .add_see_also("osd_scrub_chunk_min"),
3215
3216     Option("osd_scrub_sleep", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3217     .set_default(0)
3218     .set_description("Duration to inject a delay during scrubbing"),
3219
3220     Option("osd_scrub_extended_sleep", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3221     .set_default(0)
3222     .set_description("Duration to inject a delay during scrubbing out of scrubbing hours")
3223     .add_see_also("osd_scrub_begin_hour")
3224     .add_see_also("osd_scrub_end_hour")
3225     .add_see_also("osd_scrub_begin_week_day")
3226     .add_see_also("osd_scrub_end_week_day"),
3227
3228     Option("osd_scrub_auto_repair", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3229     .set_default(false)
3230     .set_description("Automatically repair damaged objects detected during scrub"),
3231
3232     Option("osd_scrub_auto_repair_num_errors", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3233     .set_default(5)
3234     .set_description("Maximum number of detected errors to automatically repair")
3235     .add_see_also("osd_scrub_auto_repair"),
3236
3237     Option("osd_scrub_max_preemptions", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3238     .set_default(5)
3239     .set_min_max(0, 30)
3240     .set_description("Set the maximum number of times we will preempt a deep scrub due to a client operation before blocking client IO to complete the scrub"),
3241
3242     Option("osd_deep_scrub_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3243     .set_default(7_day)
3244     .set_description("Deep scrub each PG (i.e., verify data checksums) at least this often"),
3245
3246     Option("osd_deep_scrub_randomize_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3247     .set_default(0.15)
3248     .set_description("Scrubs will randomly become deep scrubs at this rate (0.15 -> 15% of scrubs are deep)")
3249     .set_long_description("This prevents a deep scrub 'stampede' by spreading deep scrubs so they are uniformly distributed over the week"),
3250
3251     Option("osd_deep_scrub_stride", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3252     .set_default(512_K)
3253     .set_description("Number of bytes to read from an object at a time during deep scrub"),
3254
3255     Option("osd_deep_scrub_keys", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3256     .set_default(1024)
3257     .set_description("Number of keys to read from an object at a time during deep scrub"),
3258
3259     Option("osd_deep_scrub_update_digest_min_age", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3260     .set_default(2_hr)
3261     .set_description("Update overall object digest only if object was last modified longer ago than this"),
3262
3263     Option("osd_deep_scrub_large_omap_object_key_threshold", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3264     .set_default(200000)
3265     .set_description("Warn when we encounter an object with more omap keys than this")
3266     .add_service("osd")
3267     .add_see_also("osd_deep_scrub_large_omap_object_value_sum_threshold"),
3268
3269     Option("osd_deep_scrub_large_omap_object_value_sum_threshold", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3270     .set_default(1_G)
3271     .set_description("Warn when we encounter an object with more omap key bytes than this")
3272     .add_service("osd")
3273     .add_see_also("osd_deep_scrub_large_omap_object_key_threshold"),
3274
3275     Option("osd_class_dir", Option::TYPE_STR, Option::LEVEL_ADVANCED)
3276     .set_default(CEPH_LIBDIR "/rados-classes")
3277     .set_description(""),
3278
3279     Option("osd_open_classes_on_start", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3280     .set_default(true)
3281     .set_description(""),
3282
3283     Option("osd_class_load_list", Option::TYPE_STR, Option::LEVEL_ADVANCED)
3284     .set_default("cephfs hello journal lock log numops " "otp rbd refcount rgw rgw_gc timeindex user version cas")
3285     .set_description(""),
3286
3287     Option("osd_class_default_list", Option::TYPE_STR, Option::LEVEL_ADVANCED)
3288     .set_default("cephfs hello journal lock log numops " "otp rbd refcount rgw rgw_gc timeindex user version cas")
3289     .set_description(""),
3290
3291     Option("osd_check_for_log_corruption", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3292     .set_default(false)
3293     .set_description(""),
3294
3295     Option("osd_use_stale_snap", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3296     .set_default(false)
3297     .set_description(""),
3298
3299     Option("osd_rollback_to_cluster_snap", Option::TYPE_STR, Option::LEVEL_ADVANCED)
3300     .set_default("")
3301     .set_description(""),
3302
3303     Option("osd_default_notify_timeout", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3304     .set_default(30)
3305     .set_description(""),
3306
3307     Option("osd_kill_backfill_at", Option::TYPE_INT, Option::LEVEL_DEV)
3308     .set_default(0)
3309     .set_description(""),
3310
3311     Option("osd_pg_epoch_persisted_max_stale", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3312     .set_default(40)
3313     .set_description(""),
3314
3315     Option("osd_target_pg_log_entries_per_osd", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3316     .set_default(3000 * 100)
3317     .set_description("target number of PG entries total on an OSD")
3318     .add_see_also("osd_max_pg_log_entries")
3319     .add_see_also("osd_min_pg_log_entries"),
3320
3321     Option("osd_min_pg_log_entries", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3322     .set_default(250)
3323     .set_description("minimum number of entries to maintain in the PG log")
3324     .add_service("osd")
3325     .add_see_also("osd_max_pg_log_entries")
3326     .add_see_also("osd_pg_log_dups_tracked"),
3327
3328     Option("osd_max_pg_log_entries", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3329     .set_default(10000)
3330     .set_description("maximum number of entries to maintain in the PG log when degraded before we trim")
3331     .add_service("osd")
3332     .add_see_also("osd_min_pg_log_entries")
3333     .add_see_also("osd_pg_log_dups_tracked"),
3334
3335     Option("osd_pg_log_dups_tracked", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3336     .set_default(3000)
3337     .set_description("how many versions back to track in order to detect duplicate ops; this is combined with both the regular pg log entries and additional minimal dup detection entries")
3338     .add_service("osd")
3339     .add_see_also("osd_min_pg_log_entries")
3340     .add_see_also("osd_max_pg_log_entries"),
3341
3342     Option("osd_object_clean_region_max_num_intervals", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3343     .set_default(10)
3344     .set_description("number of intervals in clean_offsets")
3345     .set_long_description("partial recovery uses multiple intervals to record the clean part of the object"
3346         "when the number of intervals is greater than osd_object_clean_region_max_num_intervals, minimum interval will be trimmed"
3347         "(0 will recovery the entire object data interval)")
3348     .add_service("osd"),
3349
3350     Option("osd_force_recovery_pg_log_entries_factor", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3351     .set_default(1.3)
3352     .set_description(""),
3353
3354     Option("osd_pg_log_trim_min", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3355     .set_default(100)
3356     .set_description(""),
3357
3358     Option("osd_force_auth_primary_missing_objects", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3359     .set_default(100)
3360     .set_description("Approximate missing objects above which to force auth_log_shard to be primary temporarily"),
3361
3362     Option("osd_async_recovery_min_cost", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3363     .set_default(100)
3364     .set_description("A mixture measure of number of current log entries difference "
3365                      "and historical missing objects,  above which we switch to use "
3366                      "asynchronous recovery when appropriate"),
3367
3368     Option("osd_max_pg_per_osd_hard_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3369     .set_default(3)
3370     .set_min(1)
3371     .set_description("Maximum number of PG per OSD, a factor of 'mon_max_pg_per_osd'")
3372     .set_long_description("OSD will refuse to instantiate PG if the number of PG it serves exceeds this number.")
3373     .add_see_also("mon_max_pg_per_osd"),
3374
3375     Option("osd_pg_log_trim_max", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3376     .set_default(10000)
3377     .set_description("maximum number of entries to remove at once from the PG log")
3378     .add_service("osd")
3379     .add_see_also("osd_min_pg_log_entries")
3380     .add_see_also("osd_max_pg_log_entries"),
3381
3382     Option("osd_op_complaint_time", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3383     .set_default(30)
3384     .set_description(""),
3385
3386     Option("osd_command_max_records", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3387     .set_default(256)
3388     .set_description(""),
3389
3390     Option("osd_max_pg_blocked_by", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3391     .set_default(16)
3392     .set_description(""),
3393
3394     Option("osd_op_log_threshold", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3395     .set_default(5)
3396     .set_description(""),
3397
3398     Option("osd_backoff_on_unfound", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3399     .set_default(true)
3400     .set_description(""),
3401
3402     Option("osd_backoff_on_degraded", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3403     .set_default(false)
3404     .set_description(""),
3405
3406     Option("osd_backoff_on_peering", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3407     .set_default(false)
3408     .set_description(""),
3409
3410     Option("osd_debug_shutdown", Option::TYPE_BOOL, Option::LEVEL_DEV)
3411     .set_default(false)
3412     .set_description("Turn up debug levels during shutdown"),
3413
3414     Option("osd_debug_crash_on_ignored_backoff", Option::TYPE_BOOL, Option::LEVEL_DEV)
3415     .set_default(false)
3416     .set_description(""),
3417
3418     Option("osd_debug_inject_dispatch_delay_probability", Option::TYPE_FLOAT, Option::LEVEL_DEV)
3419     .set_default(0)
3420     .set_description(""),
3421
3422     Option("osd_debug_inject_dispatch_delay_duration", Option::TYPE_FLOAT, Option::LEVEL_DEV)
3423     .set_default(.1)
3424     .set_description(""),
3425
3426     Option("osd_debug_drop_ping_probability", Option::TYPE_FLOAT, Option::LEVEL_DEV)
3427     .set_default(0)
3428     .set_description(""),
3429
3430     Option("osd_debug_drop_ping_duration", Option::TYPE_INT, Option::LEVEL_DEV)
3431     .set_default(0)
3432     .set_description(""),
3433
3434     Option("osd_debug_op_order", Option::TYPE_BOOL, Option::LEVEL_DEV)
3435     .set_default(false)
3436     .set_description(""),
3437
3438     Option("osd_debug_verify_missing_on_start", Option::TYPE_BOOL, Option::LEVEL_DEV)
3439     .set_default(false)
3440     .set_description(""),
3441
3442     Option("osd_debug_verify_snaps", Option::TYPE_BOOL, Option::LEVEL_DEV)
3443     .set_default(false)
3444     .set_description(""),
3445
3446     Option("osd_debug_verify_stray_on_activate", Option::TYPE_BOOL, Option::LEVEL_DEV)
3447     .set_default(false)
3448     .set_description(""),
3449
3450     Option("osd_debug_skip_full_check_in_backfill_reservation", Option::TYPE_BOOL, Option::LEVEL_DEV)
3451     .set_default(false)
3452     .set_description(""),
3453
3454     Option("osd_debug_reject_backfill_probability", Option::TYPE_FLOAT, Option::LEVEL_DEV)
3455     .set_default(0)
3456     .set_description(""),
3457
3458     Option("osd_debug_inject_copyfrom_error", Option::TYPE_BOOL, Option::LEVEL_DEV)
3459     .set_default(false)
3460     .set_description(""),
3461
3462     Option("osd_debug_misdirected_ops", Option::TYPE_BOOL, Option::LEVEL_DEV)
3463     .set_default(false)
3464     .set_description(""),
3465
3466     Option("osd_debug_skip_full_check_in_recovery", Option::TYPE_BOOL, Option::LEVEL_DEV)
3467     .set_default(false)
3468     .set_description(""),
3469
3470     Option("osd_debug_random_push_read_error", Option::TYPE_FLOAT, Option::LEVEL_DEV)
3471     .set_default(0)
3472     .set_description(""),
3473
3474     Option("osd_debug_verify_cached_snaps", Option::TYPE_BOOL, Option::LEVEL_DEV)
3475     .set_default(false)
3476     .set_description(""),
3477
3478     Option("osd_debug_deep_scrub_sleep", Option::TYPE_FLOAT, Option::LEVEL_DEV)
3479     .set_default(0)
3480     .set_description("Inject an expensive sleep during deep scrub IO to make it easier to induce preemption"),
3481
3482     Option("osd_debug_no_acting_change", Option::TYPE_BOOL, Option::LEVEL_DEV)
3483     .set_default(false),
3484     Option("osd_debug_no_purge_strays", Option::TYPE_BOOL, Option::LEVEL_DEV)
3485     .set_default(false),
3486
3487     Option("osd_debug_pretend_recovery_active", Option::TYPE_BOOL, Option::LEVEL_DEV)
3488     .set_default(false)
3489     .set_description(""),
3490
3491     Option("osd_enable_op_tracker", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3492     .set_default(true)
3493     .set_description(""),
3494
3495     Option("osd_num_op_tracker_shard", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3496     .set_default(32)
3497     .set_description(""),
3498
3499     Option("osd_op_history_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3500     .set_default(20)
3501     .set_description(""),
3502
3503     Option("osd_op_history_duration", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3504     .set_default(600)
3505     .set_description(""),
3506
3507     Option("osd_op_history_slow_op_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3508     .set_default(20)
3509     .set_description(""),
3510
3511     Option("osd_op_history_slow_op_threshold", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3512     .set_default(10.0)
3513     .set_description(""),
3514
3515     Option("osd_target_transaction_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3516     .set_default(30)
3517     .set_description(""),
3518
3519     Option("osd_delete_sleep", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3520     .set_default(0)
3521     .set_description("Time in seconds to sleep before next removal transaction (overrides values below)"),
3522
3523     Option("osd_delete_sleep_hdd", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3524     .set_default(5)
3525     .set_description("Time in seconds to sleep before next removal transaction for HDDs"),
3526
3527     Option("osd_delete_sleep_ssd", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3528     .set_default(0)
3529     .set_description("Time in seconds to sleep before next removal transaction for SSDs"),
3530
3531     Option("osd_delete_sleep_hybrid", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3532     .set_default(2)
3533     .set_description("Time in seconds to sleep before next removal transaction when data is on HDD and journal is on SSD"),
3534
3535     Option("osd_failsafe_full_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3536     .set_default(.97)
3537     .set_description(""),
3538
3539     Option("osd_fast_shutdown", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3540     .set_default(true)
3541     .set_description("Fast, immediate shutdown")
3542     .set_long_description("Setting this to false makes the OSD do a slower teardown of all state when it receives a SIGINT or SIGTERM or when shutting down for any other reason.  That slow shutdown is primarilyy useful for doing memory leak checking with valgrind."),
3543
3544     Option("osd_fast_fail_on_connection_refused", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3545     .set_default(true)
3546     .set_description(""),
3547
3548     Option("osd_pg_object_context_cache_count", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3549     .set_default(64)
3550     .set_description(""),
3551
3552     Option("osd_tracing", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3553     .set_default(false)
3554     .set_description(""),
3555
3556     Option("osd_function_tracing", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3557     .set_default(false)
3558     .set_description(""),
3559
3560     Option("osd_fast_info", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3561     .set_default(true)
3562     .set_description(""),
3563
3564     Option("osd_debug_pg_log_writeout", Option::TYPE_BOOL, Option::LEVEL_DEV)
3565     .set_default(false)
3566     .set_description(""),
3567
3568     Option("osd_loop_before_reset_tphandle", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3569     .set_default(64)
3570     .set_description(""),
3571
3572     Option("threadpool_default_timeout", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3573     .set_default(60)
3574     .set_description(""),
3575
3576     Option("threadpool_empty_queue_max_wait", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3577     .set_default(2)
3578     .set_description(""),
3579
3580     Option("leveldb_log_to_ceph_log", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3581     .set_default(true)
3582     .set_description(""),
3583
3584     Option("leveldb_write_buffer_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3585     .set_default(8_M)
3586     .set_description(""),
3587
3588     Option("leveldb_cache_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3589     .set_default(128_M)
3590     .set_description(""),
3591
3592     Option("leveldb_block_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3593     .set_default(0)
3594     .set_description(""),
3595
3596     Option("leveldb_bloom_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3597     .set_default(0)
3598     .set_description(""),
3599
3600     Option("leveldb_max_open_files", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3601     .set_default(0)
3602     .set_description(""),
3603
3604     Option("leveldb_compression", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3605     .set_default(true)
3606     .set_description(""),
3607
3608     Option("leveldb_paranoid", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3609     .set_default(false)
3610     .set_description(""),
3611
3612     Option("leveldb_log", Option::TYPE_STR, Option::LEVEL_ADVANCED)
3613     .set_default("/dev/null")
3614     .set_description(""),
3615
3616     Option("leveldb_compact_on_mount", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3617     .set_default(false)
3618     .set_description(""),
3619
3620     Option("rocksdb_log_to_ceph_log", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3621     .set_default(true)
3622     .set_description(""),
3623
3624     Option("rocksdb_cache_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3625     .set_default(512_M)
3626     .set_flag(Option::FLAG_RUNTIME)
3627     .set_description(""),
3628
3629     Option("rocksdb_cache_row_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3630     .set_default(0)
3631     .set_description(""),
3632
3633     Option("rocksdb_cache_shard_bits", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3634     .set_default(4)
3635     .set_description(""),
3636
3637     Option("rocksdb_cache_type", Option::TYPE_STR, Option::LEVEL_ADVANCED)
3638     .set_default("binned_lru")
3639     .set_description(""),
3640
3641     Option("rocksdb_block_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3642     .set_default(4_K)
3643     .set_description(""),
3644
3645     Option("rocksdb_perf", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3646     .set_default(false)
3647     .set_description(""),
3648
3649     Option("rocksdb_collect_compaction_stats", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3650     .set_default(false)
3651     .set_description(""),
3652
3653     Option("rocksdb_collect_extended_stats", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3654     .set_default(false)
3655     .set_description(""),
3656
3657     Option("rocksdb_collect_memory_stats", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3658     .set_default(false)
3659     .set_description(""),
3660
3661     Option("rocksdb_delete_range_threshold", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3662     .set_default(1048576)
3663     .set_description("The number of keys required to invoke DeleteRange when deleting muliple keys."),
3664
3665     Option("rocksdb_bloom_bits_per_key", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3666     .set_default(20)
3667     .set_description("Number of bits per key to use for RocksDB's bloom filters.")
3668     .set_long_description("RocksDB bloom filters can be used to quickly answer the question of whether or not a key may exist or definitely does not exist in a given RocksDB SST file without having to read all keys into memory.  Using a higher bit value decreases the likelihood of false positives at the expense of additional disk space and memory consumption when the filter is loaded into RAM.  The current default value of 20 was found to provide significant performance gains when getattr calls are made (such as during new object creation in bluestore) without significant memory overhead or cache pollution when combined with rocksdb partitioned index filters.  See: https://github.com/facebook/rocksdb/wiki/Partitioned-Index-Filters for more information."),
3669
3670     Option("rocksdb_cache_index_and_filter_blocks", Option::TYPE_BOOL, Option::LEVEL_DEV)
3671     .set_default(true)
3672     .set_description("Whether to cache indices and filters in block cache")
3673     .set_long_description("By default RocksDB will load an SST file's index and bloom filters into memory when it is opened and remove them from memory when an SST file is closed.  Thus, memory consumption by indices and bloom filters is directly tied to the number of concurrent SST files allowed to be kept open.  This option instead stores cached indicies and filters in the block cache where they directly compete with other cached data.  By default we set this option to true to better account for and bound rocksdb memory usage and keep filters in memory even when an SST file is closed."),
3674
3675     Option("rocksdb_cache_index_and_filter_blocks_with_high_priority", Option::TYPE_BOOL, Option::LEVEL_DEV)
3676     .set_default(true)
3677     .set_description("Whether to cache indices and filters in the block cache with high priority")
3678     .set_long_description("A downside of setting rocksdb_cache_index_and_filter_blocks to true is that regular data can push indices and filters out of memory.  Setting this option to true means they are cached with higher priority than other data and should typically stay in the block cache."),
3679
3680     Option("rocksdb_pin_l0_filter_and_index_blocks_in_cache", Option::TYPE_BOOL, Option::LEVEL_DEV)
3681     .set_default(false)
3682     .set_description("Whether to pin Level 0 indices and bloom filters in the block cache")
3683     .set_long_description("A downside of setting rocksdb_cache_index_and_filter_blocks to true is that regular data can push indices and filters out of memory.  Setting this option to true means that level 0 SST files will always have their indices and filters pinned in the block cache."),
3684
3685     Option("rocksdb_index_type", Option::TYPE_STR, Option::LEVEL_DEV)
3686     .set_default("binary_search")
3687     .set_description("Type of index for SST files: binary_search, hash_search, two_level")
3688     .set_long_description("This option controls the table index type.  binary_search is a space efficient index block that is optimized for block-search-based index. hash_search may improve prefix lookup performance at the expense of higher disk and memory usage and potentially slower compactions.  two_level is an experimental index type that uses two binary search indexes and works in conjunction with partition filters.  See: http://rocksdb.org/blog/2017/05/12/partitioned-index-filter.html"),
3689
3690     Option("rocksdb_partition_filters", Option::TYPE_BOOL, Option::LEVEL_DEV)
3691     .set_default(false)
3692     .set_description("(experimental) partition SST index/filters into smaller blocks")
3693     .set_long_description("This is an experimental option for rocksdb that works in conjunction with two_level indices to avoid having to keep the entire filter/index in cache when cache_index_and_filter_blocks is true.  The idea is to keep a much smaller top-level index in heap/cache and then opportunistically cache the lower level indices.  See: https://github.com/facebook/rocksdb/wiki/Partitioned-Index-Filters"),
3694
3695     Option("rocksdb_metadata_block_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
3696     .set_default(4_K)
3697     .set_description("The block size for index partitions. (0 = rocksdb default)"),
3698
3699     Option("mon_rocksdb_options", Option::TYPE_STR, Option::LEVEL_ADVANCED)
3700     .set_default("write_buffer_size=33554432,"
3701                  "compression=kNoCompression,"
3702                  "level_compaction_dynamic_level_bytes=true")
3703     .set_description(""),
3704
3705     Option("osd_client_op_priority", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3706     .set_default(63)
3707     .set_description(""),
3708
3709     Option("osd_recovery_op_priority", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3710     .set_default(3)
3711     .set_description("Priority to use for recovery operations if not specified for the pool"),
3712
3713     Option("osd_peering_op_priority", Option::TYPE_UINT, Option::LEVEL_DEV)
3714     .set_default(255)
3715     .set_description(""),
3716
3717     Option("osd_snap_trim_priority", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3718     .set_default(5)
3719     .set_description(""),
3720
3721     Option("osd_snap_trim_cost", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3722     .set_default(1<<20)
3723     .set_description(""),
3724
3725     Option("osd_pg_delete_priority", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3726     .set_default(5)
3727     .set_description(""),
3728
3729     Option("osd_pg_delete_cost", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3730     .set_default(1<<20)
3731     .set_description(""),
3732
3733     Option("osd_scrub_priority", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3734     .set_default(5)
3735     .set_description("Priority for scrub operations in work queue"),
3736
3737     Option("osd_scrub_cost", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3738     .set_default(50<<20)
3739     .set_description("Cost for scrub operations in work queue"),
3740
3741     Option("osd_requested_scrub_priority", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3742     .set_default(120)
3743     .set_description(""),
3744
3745     Option("osd_recovery_priority", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3746     .set_default(5)
3747     .set_description("Priority of recovery in the work queue")
3748     .set_long_description("Not related to a pool's recovery_priority"),
3749
3750     Option("osd_recovery_cost", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3751     .set_default(20<<20)
3752     .set_description(""),
3753
3754     Option("osd_recovery_op_warn_multiple", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3755     .set_default(16)
3756     .set_description(""),
3757
3758     Option("osd_mon_shutdown_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3759     .set_default(5)
3760     .set_description(""),
3761
3762     Option("osd_shutdown_pgref_assert", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3763     .set_default(false)
3764     .set_description(""),
3765
3766     Option("osd_max_object_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3767     .set_default(128_M)
3768     .set_description(""),
3769
3770     Option("osd_max_object_name_len", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3771     .set_default(2048)
3772     .set_description(""),
3773
3774     Option("osd_max_object_namespace_len", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3775     .set_default(256)
3776     .set_description(""),
3777
3778     Option("osd_max_attr_name_len", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3779     .set_default(100)
3780     .set_description(""),
3781
3782     Option("osd_max_attr_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3783     .set_default(0)
3784     .set_description(""),
3785
3786     Option("osd_max_omap_entries_per_request", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3787     .set_default(1024)
3788     .set_description(""),
3789
3790     Option("osd_max_omap_bytes_per_request", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3791     .set_default(1_G)
3792     .set_description(""),
3793
3794     Option("osd_max_write_op_reply_len", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3795     .set_default(32)
3796     .set_description("Max size of the per-op payload for requests with the RETURNVEC flag set")
3797     .set_long_description("This value caps the amount of data (per op; a request may have many ops) that will be sent back to the client and recorded in the PG log."),
3798
3799     Option("osd_objectstore", Option::TYPE_STR, Option::LEVEL_ADVANCED)
3800     .set_default("bluestore")
3801     .set_enum_allowed({"bluestore", "filestore", "memstore", "kstore"})
3802     .set_flag(Option::FLAG_CREATE)
3803     .set_description("backend type for an OSD (like filestore or bluestore)"),
3804
3805     Option("osd_objectstore_tracing", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3806     .set_default(false)
3807     .set_description(""),
3808
3809     Option("osd_objectstore_fuse", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3810     .set_default(false)
3811     .set_description(""),
3812
3813     Option("osd_bench_small_size_max_iops", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3814     .set_default(100)
3815     .set_description(""),
3816
3817     Option("osd_bench_large_size_max_throughput", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3818     .set_default(100_M)
3819     .set_description(""),
3820
3821     Option("osd_bench_max_block_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3822     .set_default(64_M)
3823     .set_description(""),
3824
3825     Option("osd_bench_duration", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
3826     .set_default(30)
3827     .set_description(""),
3828
3829     Option("osd_blkin_trace_all", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3830     .set_default(false)
3831     .set_description(""),
3832
3833     Option("osdc_blkin_trace_all", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3834     .set_default(false)
3835     .set_description(""),
3836
3837     Option("osd_discard_disconnected_ops", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3838     .set_default(true)
3839     .set_description(""),
3840
3841     Option("osd_memory_target", Option::TYPE_SIZE, Option::LEVEL_BASIC)
3842     .set_default(4_G)
3843     .set_min(896_M)
3844     .set_flag(Option::FLAG_RUNTIME)
3845     .add_see_also("bluestore_cache_autotune")
3846     .add_see_also("osd_memory_cache_min")
3847     .add_see_also("osd_memory_base")
3848     .set_description("When tcmalloc and cache autotuning is enabled, try to keep this many bytes mapped in memory.")
3849     .set_long_description("The minimum value must be at least equal to osd_memory_base + osd_memory_cache_min."),
3850
3851     Option("osd_memory_target_cgroup_limit_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3852     .set_default(0.8)
3853     .set_min_max(0.0, 1.0)
3854     .add_see_also("osd_memory_target")
3855     .set_description("Set the default value for osd_memory_target to the cgroup memory limit (if set) times this value")
3856     .set_long_description("A value of 0 disables this feature."),
3857
3858     Option("osd_memory_base", Option::TYPE_SIZE, Option::LEVEL_DEV)
3859     .set_default(768_M)
3860     .set_flag(Option::FLAG_RUNTIME)
3861     .add_see_also("bluestore_cache_autotune")
3862     .set_description("When tcmalloc and cache autotuning is enabled, estimate the minimum amount of memory in bytes the OSD will need."),
3863
3864     Option("osd_memory_expected_fragmentation", Option::TYPE_FLOAT, Option::LEVEL_DEV)
3865     .set_default(0.15)
3866     .set_min_max(0.0, 1.0)
3867     .set_flag(Option::FLAG_RUNTIME)
3868     .add_see_also("bluestore_cache_autotune")
3869     .set_description("When tcmalloc and cache autotuning is enabled, estimate the percent of memory fragmentation."),
3870
3871     Option("osd_memory_cache_min", Option::TYPE_SIZE, Option::LEVEL_DEV)
3872     .set_default(128_M)
3873     .set_min(128_M)
3874     .set_flag(Option::FLAG_RUNTIME)
3875     .add_see_also("bluestore_cache_autotune")
3876     .set_description("When tcmalloc and cache autotuning is enabled, set the minimum amount of memory used for caches."),
3877
3878     Option("osd_memory_cache_resize_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
3879     .set_default(1)
3880     .add_see_also("bluestore_cache_autotune")
3881     .set_description("When tcmalloc and cache autotuning is enabled, wait this many seconds between resizing caches."),
3882
3883     Option("memstore_device_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3884     .set_default(1_G)
3885     .set_description(""),
3886
3887     Option("memstore_page_set", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3888     .set_default(false)
3889     .set_description(""),
3890
3891     Option("memstore_page_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3892     .set_default(64_K)
3893     .set_description(""),
3894
3895     Option("memstore_debug_omit_block_device_write", Option::TYPE_BOOL, Option::LEVEL_DEV)
3896     .set_default(false)
3897     .add_see_also("bluestore_debug_omit_block_device_write")
3898     .set_description("write metadata only"),
3899
3900     Option("objectstore_blackhole", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3901     .set_default(false)
3902     .set_description(""),
3903
3904     // --------------------------
3905     // bluestore
3906
3907     Option("bdev_debug_inflight_ios", Option::TYPE_BOOL, Option::LEVEL_DEV)
3908     .set_default(false)
3909     .set_description(""),
3910
3911     Option("bdev_inject_crash", Option::TYPE_INT, Option::LEVEL_DEV)
3912     .set_default(0)
3913     .set_description(""),
3914
3915     Option("bdev_inject_crash_flush_delay", Option::TYPE_INT, Option::LEVEL_DEV)
3916     .set_default(2)
3917     .set_description(""),
3918
3919     Option("bdev_aio", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3920     .set_default(true)
3921     .set_description(""),
3922
3923     Option("bdev_aio_poll_ms", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3924     .set_default(250)
3925     .set_description(""),
3926
3927     Option("bdev_aio_max_queue_depth", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3928     .set_default(1024)
3929     .set_description(""),
3930
3931     Option("bdev_aio_reap_max", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3932     .set_default(16)
3933     .set_description(""),
3934
3935     Option("bdev_block_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3936     .set_default(4_K)
3937     .set_description(""),
3938
3939     Option("bdev_debug_aio", Option::TYPE_BOOL, Option::LEVEL_DEV)
3940     .set_default(false)
3941     .set_description(""),
3942
3943     Option("bdev_debug_aio_suicide_timeout", Option::TYPE_FLOAT, Option::LEVEL_DEV)
3944     .set_default(60.0)
3945     .set_description(""),
3946
3947     Option("bdev_debug_aio_log_age", Option::TYPE_FLOAT, Option::LEVEL_DEV)
3948     .set_default(5.0)
3949     .set_description(""),
3950
3951     Option("bdev_nvme_unbind_from_kernel", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3952     .set_default(false)
3953     .set_description(""),
3954
3955     Option("bdev_nvme_retry_count", Option::TYPE_INT, Option::LEVEL_ADVANCED)
3956     .set_default(-1)
3957     .set_description(""),
3958
3959     Option("bdev_enable_discard", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3960     .set_default(false)
3961     .set_description(""),
3962
3963     Option("bdev_async_discard", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
3964     .set_default(false)
3965     .set_description(""),
3966
3967     Option("bluefs_alloc_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3968     .set_default(1_M)
3969     .set_description("Allocation unit size for DB and WAL devices"),
3970
3971     Option("bluefs_shared_alloc_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3972     .set_default(64_K)
3973     .set_description("Allocation unit size for primary/shared device"),
3974
3975     Option("bluefs_max_prefetch", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3976     .set_default(1_M)
3977     .set_description(""),
3978
3979     Option("bluefs_min_log_runway", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3980     .set_default(1_M)
3981     .set_description(""),
3982
3983     Option("bluefs_max_log_runway", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3984     .set_default(4194304)
3985     .set_description(""),
3986
3987     Option("bluefs_log_compact_min_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
3988     .set_default(5.0)
3989     .set_description(""),
3990
3991     Option("bluefs_log_compact_min_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3992     .set_default(16_M)
3993     .set_description(""),
3994
3995     Option("bluefs_min_flush_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
3996     .set_default(512_K)
3997     .set_description(""),
3998
3999     Option("bluefs_compact_log_sync", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4000     .set_default(false)
4001     .set_description(""),
4002
4003     Option("bluefs_buffered_io", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4004     .set_default(false)
4005     .set_description("Enabled buffered IO for bluefs reads.")
4006     .set_long_description("When this option is enabled, bluefs will in some cases perform buffered reads.  This allows the kernel page cache to act as a secondary cache for things like RocksDB compaction.  For example, if the rocksdb block cache isn't large enough to hold blocks from the compressed SST files itself, they can be read from page cache instead of from the disk.  This option previously was enabled by default, however in some test cases it appears to cause excessive swap utilization by the linux kernel and a large negative performance impact after several hours of run time.  Please exercise caution when enabling."),
4007
4008     Option("bluefs_sync_write", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4009     .set_default(false)
4010     .set_description(""),
4011
4012     Option("bluefs_allocator", Option::TYPE_STR, Option::LEVEL_DEV)
4013     .set_default("hybrid")
4014     .set_enum_allowed({"bitmap", "stupid", "avl", "hybrid"})
4015     .set_description(""),
4016
4017     Option("bluefs_preextend_wal_files", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4018     .set_default(false)
4019     .set_description("Preextent rocksdb wal files on mkfs to avoid performance penalty"),
4020
4021     Option("bluefs_log_replay_check_allocations", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4022       .set_default(true)
4023       .set_description("Enables checks for allocations consistency during log replay"),
4024
4025     Option("bluefs_replay_recovery", Option::TYPE_BOOL, Option::LEVEL_DEV)
4026     .set_default(false)
4027     .set_description("Attempt to read bluefs log so large that it became unreadable.")
4028     .set_long_description("If BlueFS log grows to extreme sizes (200GB+) it is likely that it becames unreadable. "
4029                           "This options enables heuristics that scans devices for missing data. "
4030                           "DO NOT ENABLE BY DEFAULT"),
4031
4032     Option("bluefs_replay_recovery_disable_compact", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4033     .set_default(false)
4034     .set_description(""),
4035
4036     Option("bluestore_bluefs", Option::TYPE_BOOL, Option::LEVEL_DEV)
4037     .set_default(true)
4038     .set_flag(Option::FLAG_CREATE)
4039     .set_description("Use BlueFS to back rocksdb")
4040     .set_long_description("BlueFS allows rocksdb to share the same physical device(s) as the rest of BlueStore.  It should be used in all cases unless testing/developing an alternative metadata database for BlueStore."),
4041
4042     Option("bluestore_bluefs_env_mirror", Option::TYPE_BOOL, Option::LEVEL_DEV)
4043     .set_default(false)
4044     .set_flag(Option::FLAG_CREATE)
4045     .set_description("Mirror bluefs data to file system for testing/validation"),
4046
4047     Option("bluestore_bluefs_min", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4048     .set_default(1_G)
4049     .set_description("minimum disk space allocated to BlueFS (e.g., at mkfs)"),
4050
4051     Option("bluestore_bluefs_min_free", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4052     .set_default(1_G)
4053     .set_description("minimum free space allocated to BlueFS"),
4054
4055     Option("bluestore_bluefs_max_free", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4056     .set_default(10_G)
4057     .set_description("Maximum free space allocated to BlueFS"),
4058
4059     Option("bluestore_bluefs_min_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4060     .set_default(.02)
4061     .set_description("Minimum fraction of free space devoted to BlueFS"),
4062
4063     Option("bluestore_bluefs_max_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4064     .set_default(.90)
4065     .set_description("Maximum fraction of free storage devoted to BlueFS"),
4066
4067     Option("bluestore_bluefs_gift_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4068     .set_default(.02)
4069     .set_description("Maximum fraction of free space to give to BlueFS at once"),
4070
4071     Option("bluestore_bluefs_reclaim_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4072     .set_default(.20)
4073     .set_description("Maximum fraction of free space to reclaim from BlueFS at once"),
4074
4075     Option("bluestore_bluefs_balance_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4076     .set_default(1)
4077     .set_description("How frequently (in seconds) to balance free space between BlueFS and BlueStore"),
4078
4079     Option("bluestore_bluefs_alloc_failure_dump_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4080     .set_default(0)
4081     .set_description("How frequently (in seconds) to dump allocator on"
4082       "BlueFS space allocation failure"),
4083
4084     Option("bluestore_bluefs_db_compatibility", Option::TYPE_BOOL, Option::LEVEL_DEV)
4085     .set_default(true)
4086     .set_description("Sync db with legacy bluefs extents info")
4087     .set_long_description("Enforces db sync with legacy bluefs extents information on close."
4088                           " Enables downgrades to pre-nautilus releases"),
4089
4090     Option("bluestore_spdk_mem", Option::TYPE_SIZE, Option::LEVEL_DEV)
4091     .set_default(512)
4092     .set_description("Amount of dpdk memory size in MB")
4093     .set_long_description("If running multiple SPDK instances per node, you must specify the amount of dpdk memory size in MB each instance will use, to make sure each instance uses its own dpdk memory"),
4094
4095     Option("bluestore_spdk_coremask", Option::TYPE_STR, Option::LEVEL_DEV)
4096     .set_default("0x1")
4097     .set_description("A hexadecimal bit mask of the cores to run on. Note the core numbering can change between platforms and should be determined beforehand"),
4098
4099     Option("bluestore_spdk_max_io_completion", Option::TYPE_UINT, Option::LEVEL_DEV)
4100     .set_default(0)
4101     .set_description("Maximal I/Os to be batched completed while checking queue pair completions, 0 means let spdk library determine it"),
4102
4103     Option("bluestore_spdk_io_sleep", Option::TYPE_UINT, Option::LEVEL_DEV)
4104     .set_default(5)
4105     .set_description("Time period to wait if there is no completed I/O from polling"),
4106
4107     Option("bluestore_block_path", Option::TYPE_STR, Option::LEVEL_DEV)
4108     .set_default("")
4109     .set_flag(Option::FLAG_CREATE)
4110     .set_description("Path to block device/file"),
4111
4112     Option("bluestore_block_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
4113     .set_default(100_G)
4114     .set_flag(Option::FLAG_CREATE)
4115     .set_description("Size of file to create for backing bluestore"),
4116
4117     Option("bluestore_block_create", Option::TYPE_BOOL, Option::LEVEL_DEV)
4118     .set_default(true)
4119     .set_flag(Option::FLAG_CREATE)
4120     .set_description("Create bluestore_block_path if it doesn't exist")
4121     .add_see_also("bluestore_block_path").add_see_also("bluestore_block_size"),
4122
4123     Option("bluestore_block_db_path", Option::TYPE_STR, Option::LEVEL_DEV)
4124     .set_default("")
4125     .set_flag(Option::FLAG_CREATE)
4126     .set_description("Path for db block device"),
4127
4128     Option("bluestore_block_db_size", Option::TYPE_UINT, Option::LEVEL_DEV)
4129     .set_default(0)
4130     .set_flag(Option::FLAG_CREATE)
4131     .set_description("Size of file to create for bluestore_block_db_path"),
4132
4133     Option("bluestore_block_db_create", Option::TYPE_BOOL, Option::LEVEL_DEV)
4134     .set_default(false)
4135     .set_flag(Option::FLAG_CREATE)
4136     .set_description("Create bluestore_block_db_path if it doesn't exist")
4137     .add_see_also("bluestore_block_db_path")
4138     .add_see_also("bluestore_block_db_size"),
4139
4140     Option("bluestore_block_wal_path", Option::TYPE_STR, Option::LEVEL_DEV)
4141     .set_default("")
4142     .set_flag(Option::FLAG_CREATE)
4143     .set_description("Path to block device/file backing bluefs wal"),
4144
4145     Option("bluestore_block_wal_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
4146     .set_default(96_M)
4147     .set_flag(Option::FLAG_CREATE)
4148     .set_description("Size of file to create for bluestore_block_wal_path"),
4149
4150     Option("bluestore_block_wal_create", Option::TYPE_BOOL, Option::LEVEL_DEV)
4151     .set_default(false)
4152     .set_flag(Option::FLAG_CREATE)
4153     .set_description("Create bluestore_block_wal_path if it doesn't exist")
4154     .add_see_also("bluestore_block_wal_path")
4155     .add_see_also("bluestore_block_wal_size"),
4156
4157     Option("bluestore_block_preallocate_file", Option::TYPE_BOOL, Option::LEVEL_DEV)
4158     .set_default(false)
4159     .set_flag(Option::FLAG_CREATE)
4160     .set_description("Preallocate file created via bluestore_block*_create"),
4161
4162     Option("bluestore_ignore_data_csum", Option::TYPE_BOOL, Option::LEVEL_DEV)
4163     .set_default(false)
4164     .set_flag(Option::FLAG_RUNTIME)
4165     .set_description("Ignore checksum errors on read and do not generate an EIO error"),
4166
4167     Option("bluestore_csum_type", Option::TYPE_STR, Option::LEVEL_ADVANCED)
4168     .set_default("crc32c")
4169     .set_enum_allowed({"none", "crc32c", "crc32c_16", "crc32c_8", "xxhash32", "xxhash64"})
4170     .set_flag(Option::FLAG_RUNTIME)
4171     .set_description("Default checksum algorithm to use")
4172     .set_long_description("crc32c, xxhash32, and xxhash64 are available.  The _16 and _8 variants use only a subset of the bits for more compact (but less reliable) checksumming."),
4173
4174     Option("bluestore_retry_disk_reads", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4175     .set_default(3)
4176     .set_min_max(0, 255)
4177     .set_flag(Option::FLAG_RUNTIME)
4178     .set_description("Number of read retries on checksum validation error")
4179     .set_long_description("Retries to read data from the disk this many times when checksum validation fails to handle spurious read errors gracefully."),
4180
4181     Option("bluestore_min_alloc_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4182     .set_default(0)
4183     .set_flag(Option::FLAG_CREATE)
4184     .set_description("Minimum allocation size to allocate for an object")
4185     .set_long_description("A smaller allocation size generally means less data is read and then rewritten when a copy-on-write operation is triggered (e.g., when writing to something that was recently snapshotted).  Similarly, less data is journaled before performing an overwrite (writes smaller than min_alloc_size must first pass through the BlueStore journal).  Larger values of min_alloc_size reduce the amount of metadata required to describe the on-disk layout and reduce overall fragmentation."),
4186
4187     Option("bluestore_min_alloc_size_hdd", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4188     .set_default(64_K)
4189     .set_flag(Option::FLAG_CREATE)
4190     .set_description("Default min_alloc_size value for rotational media")
4191     .add_see_also("bluestore_min_alloc_size"),
4192
4193     Option("bluestore_min_alloc_size_ssd", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4194     .set_default(4_K)
4195     .set_flag(Option::FLAG_CREATE)
4196     .set_description("Default min_alloc_size value for non-rotational (solid state)  media")
4197     .add_see_also("bluestore_min_alloc_size"),
4198
4199     Option("bluestore_max_alloc_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4200     .set_default(0)
4201     .set_flag(Option::FLAG_CREATE)
4202     .set_description("Maximum size of a single allocation (0 for no max)"),
4203
4204     Option("bluestore_prefer_deferred_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4205     .set_default(0)
4206     .set_flag(Option::FLAG_RUNTIME)
4207     .set_description("Writes smaller than this size will be written to the journal and then asynchronously written to the device.  This can be beneficial when using rotational media where seeks are expensive, and is helpful both with and without solid state journal/wal devices."),
4208
4209     Option("bluestore_prefer_deferred_size_hdd", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4210     .set_default(32768)
4211     .set_flag(Option::FLAG_RUNTIME)
4212     .set_description("Default bluestore_prefer_deferred_size for rotational media")
4213     .add_see_also("bluestore_prefer_deferred_size"),
4214
4215     Option("bluestore_prefer_deferred_size_ssd", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4216     .set_default(0)
4217     .set_flag(Option::FLAG_RUNTIME)
4218     .set_description("Default bluestore_prefer_deferred_size for non-rotational (solid state) media")
4219     .add_see_also("bluestore_prefer_deferred_size"),
4220
4221     Option("bluestore_compression_mode", Option::TYPE_STR, Option::LEVEL_ADVANCED)
4222     .set_default("none")
4223     .set_enum_allowed({"none", "passive", "aggressive", "force"})
4224     .set_flag(Option::FLAG_RUNTIME)
4225     .set_description("Default policy for using compression when pool does not specify")
4226     .set_long_description("'none' means never use compression.  'passive' means use compression when clients hint that data is compressible.  'aggressive' means use compression unless clients hint that data is not compressible.  This option is used when the per-pool property for the compression mode is not present."),
4227
4228     Option("bluestore_compression_algorithm", Option::TYPE_STR, Option::LEVEL_ADVANCED)
4229     .set_default("snappy")
4230     .set_enum_allowed({"", "snappy", "zlib", "zstd", "lz4"})
4231     .set_flag(Option::FLAG_RUNTIME)
4232     .set_description("Default compression algorithm to use when writing object data")
4233     .set_long_description("This controls the default compressor to use (if any) if the per-pool property is not set.  Note that zstd is *not* recommended for bluestore due to high CPU overhead when compressing small amounts of data."),
4234
4235     Option("bluestore_compression_min_blob_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4236     .set_default(0)
4237     .set_flag(Option::FLAG_RUNTIME)
4238     .set_description("Maximum chunk size to apply compression to when random access is expected for an object.")
4239     .set_long_description("Chunks larger than this are broken into smaller chunks before being compressed"),
4240
4241     Option("bluestore_compression_min_blob_size_hdd", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4242     .set_default(128_K)
4243     .set_flag(Option::FLAG_RUNTIME)
4244     .set_description("Default value of bluestore_compression_min_blob_size for rotational media")
4245     .add_see_also("bluestore_compression_min_blob_size"),
4246
4247     Option("bluestore_compression_min_blob_size_ssd", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4248     .set_default(8_K)
4249     .set_flag(Option::FLAG_RUNTIME)
4250     .set_description("Default value of bluestore_compression_min_blob_size for non-rotational (solid state) media")
4251     .add_see_also("bluestore_compression_min_blob_size"),
4252
4253     Option("bluestore_compression_max_blob_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4254     .set_default(0)
4255     .set_flag(Option::FLAG_RUNTIME)
4256     .set_description("Maximum chunk size to apply compression to when non-random access is expected for an object.")
4257     .set_long_description("Chunks larger than this are broken into smaller chunks before being compressed"),
4258
4259     Option("bluestore_compression_max_blob_size_hdd", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4260     .set_default(512_K)
4261     .set_flag(Option::FLAG_RUNTIME)
4262     .set_description("Default value of bluestore_compression_max_blob_size for rotational media")
4263     .add_see_also("bluestore_compression_max_blob_size"),
4264
4265     Option("bluestore_compression_max_blob_size_ssd", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4266     .set_default(64_K)
4267     .set_flag(Option::FLAG_RUNTIME)
4268     .set_description("Default value of bluestore_compression_max_blob_size for non-rotational (solid state) media")
4269     .add_see_also("bluestore_compression_max_blob_size"),
4270
4271     Option("bluestore_gc_enable_blob_threshold", Option::TYPE_INT, Option::LEVEL_DEV)
4272     .set_default(0)
4273     .set_flag(Option::FLAG_RUNTIME)
4274     .set_description(""),
4275
4276     Option("bluestore_gc_enable_total_threshold", Option::TYPE_INT, Option::LEVEL_DEV)
4277     .set_default(0)
4278     .set_flag(Option::FLAG_RUNTIME)
4279     .set_description(""),
4280
4281     Option("bluestore_max_blob_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
4282     .set_default(0)
4283     .set_flag(Option::FLAG_RUNTIME)
4284     .set_description("")
4285     .set_long_description("Bluestore blobs are collections of extents (ie on-disk data) originating from one or more objects.  Blobs can be compressed, typically have checksum data, may be overwritten, may be shared (with an extent ref map), or split.  This setting controls the maximum size a blob is allowed to be."),
4286
4287     Option("bluestore_max_blob_size_hdd", Option::TYPE_SIZE, Option::LEVEL_DEV)
4288     .set_default(512_K)
4289     .set_flag(Option::FLAG_RUNTIME)
4290     .set_description("")
4291     .add_see_also("bluestore_max_blob_size"),
4292
4293     Option("bluestore_max_blob_size_ssd", Option::TYPE_SIZE, Option::LEVEL_DEV)
4294     .set_default(64_K)
4295     .set_flag(Option::FLAG_RUNTIME)
4296     .set_description("")
4297     .add_see_also("bluestore_max_blob_size"),
4298
4299     Option("bluestore_compression_required_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4300     .set_default(.875)
4301     .set_flag(Option::FLAG_RUNTIME)
4302     .set_description("Compression ratio required to store compressed data")
4303     .set_long_description("If we compress data and get less than this we discard the result and store the original uncompressed data."),
4304
4305     Option("bluestore_extent_map_shard_max_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
4306     .set_default(1200)
4307     .set_description("Max size (bytes) for a single extent map shard before splitting"),
4308
4309     Option("bluestore_extent_map_shard_target_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
4310     .set_default(500)
4311     .set_description("Target size (bytes) for a single extent map shard"),
4312
4313     Option("bluestore_extent_map_shard_min_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
4314     .set_default(150)
4315     .set_description("Min size (bytes) for a single extent map shard before merging"),
4316
4317     Option("bluestore_extent_map_shard_target_size_slop", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4318     .set_default(.2)
4319     .set_description("Ratio above/below target for a shard when trying to align to an existing extent or blob boundary"),
4320
4321     Option("bluestore_extent_map_inline_shard_prealloc_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
4322     .set_default(256)
4323     .set_description("Preallocated buffer for inline shards"),
4324
4325     Option("bluestore_cache_trim_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4326     .set_default(.05)
4327     .set_description("How frequently we trim the bluestore cache"),
4328
4329     Option("bluestore_cache_trim_max_skip_pinned", Option::TYPE_UINT, Option::LEVEL_DEV)
4330     .set_default(64)
4331     .set_description("Max pinned cache entries we consider before giving up"),
4332
4333     Option("bluestore_cache_type", Option::TYPE_STR, Option::LEVEL_DEV)
4334     .set_default("2q")
4335     .set_enum_allowed({"2q", "lru"})
4336     .set_description("Cache replacement algorithm"),
4337
4338     Option("bluestore_2q_cache_kin_ratio", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4339     .set_default(.5)
4340     .set_description("2Q paper suggests .5"),
4341
4342     Option("bluestore_2q_cache_kout_ratio", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4343     .set_default(.5)
4344     .set_description("2Q paper suggests .5"),
4345
4346     Option("bluestore_cache_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
4347     .set_default(0)
4348     .set_description("Cache size (in bytes) for BlueStore")
4349     .set_long_description("This includes data and metadata cached by BlueStore as well as memory devoted to rocksdb's cache(s)."),
4350
4351     Option("bluestore_cache_size_hdd", Option::TYPE_SIZE, Option::LEVEL_DEV)
4352     .set_default(1_G)
4353     .set_description("Default bluestore_cache_size for rotational media")
4354     .add_see_also("bluestore_cache_size"),
4355
4356     Option("bluestore_cache_size_ssd", Option::TYPE_SIZE, Option::LEVEL_DEV)
4357     .set_default(3_G)
4358     .set_description("Default bluestore_cache_size for non-rotational (solid state) media")
4359     .add_see_also("bluestore_cache_size"),
4360
4361     Option("bluestore_cache_meta_ratio", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4362     .set_default(.4)
4363     .add_see_also("bluestore_cache_size")
4364     .set_description("Ratio of bluestore cache to devote to metadata"),
4365
4366     Option("bluestore_cache_kv_ratio", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4367     .set_default(.4)
4368     .add_see_also("bluestore_cache_size")
4369     .set_description("Ratio of bluestore cache to devote to kv database (rocksdb)"),
4370
4371     Option("bluestore_cache_autotune", Option::TYPE_BOOL, Option::LEVEL_DEV)
4372     .set_default(true)
4373     .add_see_also("bluestore_cache_size")
4374     .add_see_also("bluestore_cache_meta_ratio")
4375     .set_description("Automatically tune the ratio of caches while respecting min values."),
4376
4377     Option("bluestore_cache_autotune_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4378     .set_default(5)
4379     .add_see_also("bluestore_cache_autotune")
4380     .set_description("The number of seconds to wait between rebalances when cache autotune is enabled."),
4381
4382     Option("bluestore_alloc_stats_dump_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4383       .set_default(3600 * 24)
4384       .set_description("The period (in second) for logging allocation statistics."),
4385
4386     Option("bluestore_kvbackend", Option::TYPE_STR, Option::LEVEL_DEV)
4387     .set_default("rocksdb")
4388     .set_flag(Option::FLAG_CREATE)
4389     .set_description("Key value database to use for bluestore"),
4390
4391     Option("bluestore_allocator", Option::TYPE_STR, Option::LEVEL_ADVANCED)
4392     .set_default("hybrid")
4393     .set_enum_allowed({"bitmap", "stupid", "avl", "hybrid"})
4394     .set_description("Allocator policy")
4395     .set_long_description("Allocator to use for bluestore.  Stupid should only be used for testing."),
4396
4397     Option("bluestore_freelist_blocks_per_key", Option::TYPE_SIZE, Option::LEVEL_DEV)
4398     .set_default(128)
4399     .set_description("Block (and bits) per database key"),
4400
4401     Option("bluestore_bitmapallocator_blocks_per_zone", Option::TYPE_SIZE, Option::LEVEL_DEV)
4402     .set_default(1024)
4403     .set_description(""),
4404
4405     Option("bluestore_bitmapallocator_span_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
4406     .set_default(1024)
4407     .set_description(""),
4408
4409     Option("bluestore_max_deferred_txc", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4410     .set_default(32)
4411     .set_description("Max transactions with deferred writes that can accumulate before we force flush deferred writes"),
4412
4413     Option("bluestore_max_defer_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4414     .set_default(3)
4415     .set_description("max duration to force deferred submit"),
4416
4417     Option("bluestore_rocksdb_options", Option::TYPE_STR, Option::LEVEL_ADVANCED)
4418     .set_default("compression=kNoCompression,max_write_buffer_number=4,min_write_buffer_number_to_merge=1,recycle_log_file_num=4,write_buffer_size=268435456,writable_file_max_buffer_size=0,compaction_readahead_size=2097152,max_background_compactions=2")
4419     .set_description("Rocksdb options"),
4420
4421     Option("bluestore_rocksdb_cf", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4422     .set_default(false)
4423     .set_description("Enable use of rocksdb column families for bluestore metadata"),
4424
4425     Option("bluestore_rocksdb_cfs", Option::TYPE_STR, Option::LEVEL_DEV)
4426     .set_default("M= P= L=")
4427     .set_description("List of whitespace-separate key/value pairs where key is CF name and value is CF options"),
4428
4429     Option("bluestore_fsck_on_mount", Option::TYPE_BOOL, Option::LEVEL_DEV)
4430     .set_default(false)
4431     .set_description("Run fsck at mount"),
4432
4433     Option("bluestore_fsck_on_mount_deep", Option::TYPE_BOOL, Option::LEVEL_DEV)
4434     .set_default(false)
4435     .set_description("Run deep fsck at mount when bluestore_fsck_on_mount is set to true"),
4436
4437     Option("bluestore_fsck_quick_fix_on_mount", Option::TYPE_BOOL, Option::LEVEL_DEV)
4438       .set_default(true)
4439       .set_description("Do quick-fix for the store at mount"),
4440
4441     Option("bluestore_fsck_on_umount", Option::TYPE_BOOL, Option::LEVEL_DEV)
4442     .set_default(false)
4443     .set_description("Run fsck at umount"),
4444
4445     Option("bluestore_fsck_on_umount_deep", Option::TYPE_BOOL, Option::LEVEL_DEV)
4446     .set_default(false)
4447     .set_description("Run deep fsck at umount when bluestore_fsck_on_umount is set to true"),
4448
4449     Option("bluestore_fsck_on_mkfs", Option::TYPE_BOOL, Option::LEVEL_DEV)
4450     .set_default(true)
4451     .set_description("Run fsck after mkfs"),
4452
4453     Option("bluestore_fsck_on_mkfs_deep", Option::TYPE_BOOL, Option::LEVEL_DEV)
4454     .set_default(false)
4455     .set_description("Run deep fsck after mkfs"),
4456
4457     Option("bluestore_sync_submit_transaction", Option::TYPE_BOOL, Option::LEVEL_DEV)
4458     .set_default(false)
4459     .set_description("Try to submit metadata transaction to rocksdb in queuing thread context"),
4460
4461     Option("bluestore_fsck_read_bytes_cap", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4462     .set_default(64_M)
4463     .set_flag(Option::FLAG_RUNTIME)
4464     .set_description("Maximum bytes read at once by deep fsck"),
4465
4466     Option("bluestore_fsck_quick_fix_threads", Option::TYPE_INT, Option::LEVEL_ADVANCED)
4467       .set_default(2)
4468       .set_description("Number of additional threads to perform quick-fix (shallow fsck) command"),
4469
4470     Option("bluestore_throttle_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4471     .set_default(64_M)
4472     .set_flag(Option::FLAG_RUNTIME)
4473     .set_description("Maximum bytes in flight before we throttle IO submission"),
4474
4475     Option("bluestore_throttle_deferred_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4476     .set_default(128_M)
4477     .set_flag(Option::FLAG_RUNTIME)
4478     .set_description("Maximum bytes for deferred writes before we throttle IO submission"),
4479
4480     Option("bluestore_throttle_cost_per_io", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4481     .set_default(0)
4482     .set_flag(Option::FLAG_RUNTIME)
4483     .set_description("Overhead added to transaction cost (in bytes) for each IO"),
4484
4485   Option("bluestore_throttle_cost_per_io_hdd", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4486     .set_default(670000)
4487     .set_flag(Option::FLAG_RUNTIME)
4488     .set_description("Default bluestore_throttle_cost_per_io for rotational media")
4489     .add_see_also("bluestore_throttle_cost_per_io"),
4490
4491     Option("bluestore_throttle_cost_per_io_ssd", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4492     .set_default(4000)
4493     .set_flag(Option::FLAG_RUNTIME)
4494     .set_description("Default bluestore_throttle_cost_per_io for non-rotation (solid state) media")
4495     .add_see_also("bluestore_throttle_cost_per_io"),
4496
4497     Option("bluestore_deferred_batch_ops", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4498     .set_default(0)
4499     .set_flag(Option::FLAG_RUNTIME)
4500     .set_description("Max number of deferred writes before we flush the deferred write queue"),
4501
4502     Option("bluestore_deferred_batch_ops_hdd", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4503     .set_default(64)
4504     .set_flag(Option::FLAG_RUNTIME)
4505     .set_description("Default bluestore_deferred_batch_ops for rotational media")
4506     .add_see_also("bluestore_deferred_batch_ops"),
4507
4508     Option("bluestore_deferred_batch_ops_ssd", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4509     .set_default(16)
4510     .set_flag(Option::FLAG_RUNTIME)
4511     .set_description("Default bluestore_deferred_batch_ops for non-rotational (solid state) media")
4512     .add_see_also("bluestore_deferred_batch_ops"),
4513
4514     Option("bluestore_nid_prealloc", Option::TYPE_INT, Option::LEVEL_DEV)
4515     .set_default(1024)
4516     .set_description("Number of unique object ids to preallocate at a time"),
4517
4518     Option("bluestore_blobid_prealloc", Option::TYPE_UINT, Option::LEVEL_DEV)
4519     .set_default(10240)
4520     .set_description("Number of unique blob ids to preallocate at a time"),
4521
4522     Option("bluestore_clone_cow", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4523     .set_default(true)
4524     .set_flag(Option::FLAG_RUNTIME)
4525     .set_description("Use copy-on-write when cloning objects (versus reading and rewriting them at clone time)"),
4526
4527     Option("bluestore_default_buffered_read", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4528     .set_default(true)
4529     .set_flag(Option::FLAG_RUNTIME)
4530     .set_description("Cache read results by default (unless hinted NOCACHE or WONTNEED)"),
4531
4532     Option("bluestore_default_buffered_write", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4533     .set_default(false)
4534     .set_flag(Option::FLAG_RUNTIME)
4535     .set_description("Cache writes by default (unless hinted NOCACHE or WONTNEED)"),
4536
4537     Option("bluestore_debug_misc", Option::TYPE_BOOL, Option::LEVEL_DEV)
4538     .set_default(false)
4539     .set_description(""),
4540
4541     Option("bluestore_debug_no_reuse_blocks", Option::TYPE_BOOL, Option::LEVEL_DEV)
4542     .set_default(false)
4543     .set_description(""),
4544
4545     Option("bluestore_debug_small_allocations", Option::TYPE_INT, Option::LEVEL_DEV)
4546     .set_default(0)
4547     .set_description(""),
4548
4549     Option("bluestore_debug_too_many_blobs_threshold", Option::TYPE_INT, Option::LEVEL_DEV)
4550     .set_default(24*1024)
4551     .set_description(""),
4552
4553     Option("bluestore_debug_freelist", Option::TYPE_BOOL, Option::LEVEL_DEV)
4554     .set_default(false)
4555     .set_description(""),
4556
4557     Option("bluestore_debug_prefill", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4558     .set_default(0)
4559     .set_description("simulate fragmentation"),
4560
4561     Option("bluestore_debug_prefragment_max", Option::TYPE_SIZE, Option::LEVEL_DEV)
4562     .set_default(1_M)
4563     .set_description(""),
4564
4565     Option("bluestore_debug_inject_read_err", Option::TYPE_BOOL, Option::LEVEL_DEV)
4566     .set_default(false)
4567     .set_description(""),
4568
4569     Option("bluestore_debug_randomize_serial_transaction", Option::TYPE_INT, Option::LEVEL_DEV)
4570     .set_default(0)
4571     .set_description(""),
4572
4573     Option("bluestore_debug_omit_block_device_write", Option::TYPE_BOOL, Option::LEVEL_DEV)
4574     .set_default(false)
4575     .set_description(""),
4576
4577     Option("bluestore_debug_fsck_abort", Option::TYPE_BOOL, Option::LEVEL_DEV)
4578     .set_default(false)
4579     .set_description(""),
4580
4581     Option("bluestore_debug_omit_kv_commit", Option::TYPE_BOOL, Option::LEVEL_DEV)
4582     .set_default(false)
4583     .set_description(""),
4584
4585     Option("bluestore_debug_permit_any_bdev_label", Option::TYPE_BOOL, Option::LEVEL_DEV)
4586     .set_default(false)
4587     .set_description(""),
4588
4589     Option("bluestore_debug_random_read_err", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4590     .set_default(0)
4591     .set_description(""),
4592
4593     Option("bluestore_debug_inject_bug21040", Option::TYPE_BOOL, Option::LEVEL_DEV)
4594     .set_default(false)
4595     .set_description(""),
4596
4597     Option("bluestore_debug_inject_csum_err_probability", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4598     .set_default(0.0)
4599     .set_description("inject crc verification errors into bluestore device reads"),
4600
4601     Option("bluestore_fsck_error_on_no_per_pool_stats", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4602     .set_default(false)
4603     .set_description("Make fsck error (instead of warn) when bluestore lacks per-pool stats, e.g., after an upgrade"),
4604
4605     Option("bluestore_warn_on_bluefs_spillover", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4606     .set_default(true)
4607     .set_description("Enable health indication on bluefs slow device usage"),
4608
4609     Option("bluestore_warn_on_legacy_statfs", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4610     .set_default(true)
4611     .set_description("Enable health indication on lack of per-pool statfs reporting from bluestore"),
4612
4613     Option("bluestore_fsck_error_on_no_per_pool_omap", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4614     .set_default(false)
4615     .set_description("Make fsck error (instead of warn) when objects without per-pool omap are found"),
4616
4617     Option("bluestore_warn_on_no_per_pool_omap", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4618     .set_default(true)
4619     .set_description("Enable health indication on lack of per-pool omap"),
4620
4621     Option("bluestore_log_op_age", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4622     .set_default(5)
4623     .set_description("log operation if it's slower than this age (seconds)"),
4624
4625     Option("bluestore_log_omap_iterator_age", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4626     .set_default(5)
4627     .set_description("log omap iteration operation if it's slower than this age (seconds)"),
4628
4629     Option("bluestore_log_collection_list_age", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4630     .set_default(60)
4631     .set_description("log collection list operation if it's slower than this age (seconds)"),
4632
4633     Option("bluestore_debug_enforce_settings", Option::TYPE_STR, Option::LEVEL_DEV)
4634     .set_default("default")
4635     .set_enum_allowed({"default", "hdd", "ssd"})
4636     .set_description("Enforces specific hw profile settings")
4637     .set_long_description("'hdd' enforces settings intended for BlueStore above a rotational drive. 'ssd' enforces settings intended for BlueStore above a solid drive. 'default' - using settings for the actual hardware."),
4638
4639     Option("bluestore_avl_alloc_bf_threshold", Option::TYPE_UINT, Option::LEVEL_DEV)
4640     .set_default(131072)
4641     .set_description(""),
4642
4643     Option("bluestore_avl_alloc_bf_free_pct", Option::TYPE_UINT, Option::LEVEL_DEV)
4644     .set_default(4)
4645     .set_description(""),
4646
4647     Option("bluestore_hybrid_alloc_mem_cap", Option::TYPE_UINT, Option::LEVEL_DEV)
4648     .set_default(64_M)
4649     .set_description("Maximum RAM hybrid allocator should use before enabling bitmap supplement"),
4650
4651     Option("bluestore_volume_selection_policy", Option::TYPE_STR, Option::LEVEL_DEV)
4652     .set_default("rocksdb_original")
4653     .set_enum_allowed({ "rocksdb_original", "use_some_extra" })
4654     .set_description("Determines bluefs volume selection policy")
4655     .set_long_description("Determines bluefs volume selection policy. 'use_some_extra' policy allows to override RocksDB level granularity and put high level's data to faster device even when the level doesn't completely fit there"),
4656
4657     Option("bluestore_volume_selection_reserved_factor", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4658       .set_flag(Option::FLAG_STARTUP)
4659       .set_default(2.0)
4660       .set_description("DB level size multiplier. Determines amount of space at DB device to bar from the usage when 'use some extra' policy is in action. Reserved size is determined as sum(L_max_size[0], L_max_size[L-1]) + L_max_size[L] * this_factor"),
4661
4662     Option("bluestore_volume_selection_reserved", Option::TYPE_INT, Option::LEVEL_ADVANCED)
4663       .set_flag(Option::FLAG_STARTUP)
4664       .set_default(0)
4665       .set_description("Space reserved at DB device and not allowed for 'use some extra' policy usage. Overrides 'bluestore_volume_selection_reserved_factor' setting and introduces straightforward limit."),
4666
4667     Option("bluestore_ioring", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4668     .set_default(false)
4669     .set_description("Enables Linux io_uring API instead of libaio"),
4670
4671     // -----------------------------------------
4672     // kstore
4673
4674     Option("kstore_max_ops", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4675     .set_default(512)
4676     .set_description(""),
4677
4678     Option("kstore_max_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4679     .set_default(64_M)
4680     .set_description(""),
4681
4682     Option("kstore_backend", Option::TYPE_STR, Option::LEVEL_ADVANCED)
4683     .set_default("rocksdb")
4684     .set_description(""),
4685
4686     Option("kstore_rocksdb_options", Option::TYPE_STR, Option::LEVEL_ADVANCED)
4687     .set_default("compression=kNoCompression")
4688     .set_description("Options to pass through when RocksDB is used as the KeyValueDB for kstore."),
4689
4690     Option("kstore_fsck_on_mount", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4691     .set_default(false)
4692     .set_description("Whether or not to run fsck on mount for kstore."),
4693
4694     Option("kstore_fsck_on_mount_deep", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4695     .set_default(true)
4696     .set_description("Whether or not to run deep fsck on mount for kstore"),
4697
4698     Option("kstore_nid_prealloc", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4699     .set_default(1024)
4700     .set_description(""),
4701
4702     Option("kstore_sync_transaction", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4703     .set_default(false)
4704     .set_description(""),
4705
4706     Option("kstore_sync_submit_transaction", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4707     .set_default(false)
4708     .set_description(""),
4709
4710     Option("kstore_onode_map_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4711     .set_default(1024)
4712     .set_description(""),
4713
4714     Option("kstore_default_stripe_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4715     .set_default(65536)
4716     .set_description(""),
4717
4718     // ---------------------
4719     // filestore
4720
4721     Option("filestore_rocksdb_options", Option::TYPE_STR, Option::LEVEL_DEV)
4722     .set_default("max_background_jobs=10,compaction_readahead_size=2097152,compression=kNoCompression")
4723     .set_description("Options to pass through when RocksDB is used as the KeyValueDB for filestore."),
4724
4725     Option("filestore_omap_backend", Option::TYPE_STR, Option::LEVEL_DEV)
4726     .set_default("rocksdb")
4727     .set_enum_allowed({"leveldb", "rocksdb"})
4728     .set_description("The KeyValueDB to use for filestore metadata (ie omap)."),
4729
4730     Option("filestore_omap_backend_path", Option::TYPE_STR, Option::LEVEL_DEV)
4731     .set_default("")
4732     .set_description("The path where the filestore KeyValueDB should store it's database(s)."),
4733
4734     Option("filestore_wbthrottle_enable", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4735     .set_default(true)
4736     .set_description("Enabling throttling of operations to backing file system"),
4737
4738     Option("filestore_wbthrottle_btrfs_bytes_start_flusher", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4739     .set_default(41943040)
4740     .set_description("Start flushing (fsyncing) when this many bytes are written(btrfs)"),
4741
4742     Option("filestore_wbthrottle_btrfs_bytes_hard_limit", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4743     .set_default(419430400)
4744     .set_description("Block writes when this many bytes haven't been flushed (fsynced) (btrfs)"),
4745
4746     Option("filestore_wbthrottle_btrfs_ios_start_flusher", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4747     .set_default(500)
4748     .set_description("Start flushing (fsyncing) when this many IOs are written (brtrfs)"),
4749
4750     Option("filestore_wbthrottle_btrfs_ios_hard_limit", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4751     .set_default(5000)
4752     .set_description("Block writes when this many IOs haven't been flushed (fsynced) (btrfs)"),
4753
4754     Option("filestore_wbthrottle_btrfs_inodes_start_flusher", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4755     .set_default(500)
4756     .set_description("Start flushing (fsyncing) when this many distinct inodes have been modified (btrfs)"),
4757
4758     Option("filestore_wbthrottle_xfs_bytes_start_flusher", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4759     .set_default(41943040)
4760     .set_description("Start flushing (fsyncing) when this many bytes are written(xfs)"),
4761
4762     Option("filestore_wbthrottle_xfs_bytes_hard_limit", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4763     .set_default(419430400)
4764     .set_description("Block writes when this many bytes haven't been flushed (fsynced) (xfs)"),
4765
4766     Option("filestore_wbthrottle_xfs_ios_start_flusher", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4767     .set_default(500)
4768     .set_description("Start flushing (fsyncing) when this many IOs are written (xfs)"),
4769
4770     Option("filestore_wbthrottle_xfs_ios_hard_limit", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4771     .set_default(5000)
4772     .set_description("Block writes when this many IOs haven't been flushed (fsynced) (xfs)"),
4773
4774     Option("filestore_wbthrottle_xfs_inodes_start_flusher", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4775     .set_default(500)
4776     .set_description("Start flushing (fsyncing) when this many distinct inodes have been modified (xfs)"),
4777
4778     Option("filestore_wbthrottle_btrfs_inodes_hard_limit", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4779     .set_default(5000)
4780     .set_description("Block writing when this many inodes have outstanding writes (btrfs)"),
4781
4782     Option("filestore_wbthrottle_xfs_inodes_hard_limit", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4783     .set_default(5000)
4784     .set_description("Block writing when this many inodes have outstanding writes (xfs)"),
4785
4786     Option("filestore_odsync_write", Option::TYPE_BOOL, Option::LEVEL_DEV)
4787     .set_default(false)
4788     .set_description("Write with O_DSYNC"),
4789
4790     Option("filestore_index_retry_probability", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4791     .set_default(0)
4792     .set_description(""),
4793
4794     Option("filestore_debug_inject_read_err", Option::TYPE_BOOL, Option::LEVEL_DEV)
4795     .set_default(false)
4796     .set_description(""),
4797
4798     Option("filestore_debug_random_read_err", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4799     .set_default(0)
4800     .set_description(""),
4801
4802     Option("filestore_debug_omap_check", Option::TYPE_BOOL, Option::LEVEL_DEV)
4803     .set_default(false)
4804     .set_description(""),
4805
4806     Option("filestore_omap_header_cache_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
4807     .set_default(1024)
4808     .set_description(""),
4809
4810     Option("filestore_max_inline_xattr_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
4811     .set_default(0)
4812     .set_description(""),
4813
4814     Option("filestore_max_inline_xattr_size_xfs", Option::TYPE_SIZE, Option::LEVEL_DEV)
4815     .set_default(65536)
4816     .set_description(""),
4817
4818     Option("filestore_max_inline_xattr_size_btrfs", Option::TYPE_SIZE, Option::LEVEL_DEV)
4819     .set_default(2048)
4820     .set_description(""),
4821
4822     Option("filestore_max_inline_xattr_size_other", Option::TYPE_SIZE, Option::LEVEL_DEV)
4823     .set_default(512)
4824     .set_description(""),
4825
4826     Option("filestore_max_inline_xattrs", Option::TYPE_UINT, Option::LEVEL_DEV)
4827     .set_default(0)
4828     .set_description(""),
4829
4830     Option("filestore_max_inline_xattrs_xfs", Option::TYPE_UINT, Option::LEVEL_DEV)
4831     .set_default(10)
4832     .set_description(""),
4833
4834     Option("filestore_max_inline_xattrs_btrfs", Option::TYPE_UINT, Option::LEVEL_DEV)
4835     .set_default(10)
4836     .set_description(""),
4837
4838     Option("filestore_max_inline_xattrs_other", Option::TYPE_UINT, Option::LEVEL_DEV)
4839     .set_default(2)
4840     .set_description(""),
4841
4842     Option("filestore_max_xattr_value_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
4843     .set_default(0)
4844     .set_description(""),
4845
4846     Option("filestore_max_xattr_value_size_xfs", Option::TYPE_SIZE, Option::LEVEL_DEV)
4847     .set_default(64_K)
4848     .set_description(""),
4849
4850     Option("filestore_max_xattr_value_size_btrfs", Option::TYPE_SIZE, Option::LEVEL_DEV)
4851     .set_default(64_K)
4852     .set_description(""),
4853
4854     Option("filestore_max_xattr_value_size_other", Option::TYPE_SIZE, Option::LEVEL_DEV)
4855     .set_default(1_K)
4856     .set_description(""),
4857
4858     Option("filestore_sloppy_crc", Option::TYPE_BOOL, Option::LEVEL_DEV)
4859     .set_default(false)
4860     .set_description(""),
4861
4862     Option("filestore_sloppy_crc_block_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
4863     .set_default(65536)
4864     .set_description(""),
4865
4866     Option("filestore_max_alloc_hint_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
4867     .set_default(1ULL << 20)
4868     .set_description(""),
4869
4870     Option("filestore_max_sync_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4871     .set_default(5)
4872     .set_description("Period between calls to syncfs(2) and journal trims (seconds)"),
4873
4874     Option("filestore_min_sync_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4875     .set_default(.01)
4876     .set_description("Minimum period between calls to syncfs(2)"),
4877
4878     Option("filestore_btrfs_snap", Option::TYPE_BOOL, Option::LEVEL_DEV)
4879     .set_default(true)
4880     .set_description(""),
4881
4882     Option("filestore_btrfs_clone_range", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4883     .set_default(true)
4884     .set_description("Use btrfs clone_range ioctl to efficiently duplicate objects"),
4885
4886     Option("filestore_zfs_snap", Option::TYPE_BOOL, Option::LEVEL_DEV)
4887     .set_default(false)
4888     .set_description(""),
4889
4890     Option("filestore_fsync_flushes_journal_data", Option::TYPE_BOOL, Option::LEVEL_DEV)
4891     .set_default(false)
4892     .set_description(""),
4893
4894     Option("filestore_fiemap", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4895     .set_default(false)
4896     .set_description("Use fiemap ioctl(2) to determine which parts of objects are sparse"),
4897
4898     Option("filestore_punch_hole", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4899     .set_default(false)
4900     .set_description("Use fallocate(2) FALLOC_FL_PUNCH_HOLE to efficiently zero ranges of objects"),
4901
4902     Option("filestore_seek_data_hole", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4903     .set_default(false)
4904     .set_description("Use lseek(2) SEEK_HOLE and SEEK_DATA to determine which parts of objects are sparse"),
4905
4906     Option("filestore_splice", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4907     .set_default(false)
4908     .set_description("Use splice(2) to more efficiently copy data between files"),
4909
4910     Option("filestore_fadvise", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4911     .set_default(true)
4912     .set_description("Use posix_fadvise(2) to pass hints to file system"),
4913
4914     Option("filestore_collect_device_partition_information", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4915     .set_default(true)
4916     .set_description("Collect metadata about the backing file system on OSD startup"),
4917
4918     Option("filestore_xfs_extsize", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
4919     .set_default(false)
4920     .set_description("Use XFS extsize ioctl(2) to hint allocator about expected write sizes"),
4921
4922     Option("filestore_journal_parallel", Option::TYPE_BOOL, Option::LEVEL_DEV)
4923     .set_default(false)
4924     .set_description(""),
4925
4926     Option("filestore_journal_writeahead", Option::TYPE_BOOL, Option::LEVEL_DEV)
4927     .set_default(false)
4928     .set_description(""),
4929
4930     Option("filestore_journal_trailing", Option::TYPE_BOOL, Option::LEVEL_DEV)
4931     .set_default(false)
4932     .set_description(""),
4933
4934     Option("filestore_queue_max_ops", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
4935     .set_default(50)
4936     .set_description("Max IO operations in flight"),
4937
4938     Option("filestore_queue_max_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
4939     .set_default(100_M)
4940     .set_description("Max (written) bytes in flight"),
4941
4942     Option("filestore_caller_concurrency", Option::TYPE_INT, Option::LEVEL_DEV)
4943     .set_default(10)
4944     .set_description(""),
4945
4946     Option("filestore_expected_throughput_bytes", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4947     .set_default(200_M)
4948     .set_description("Expected throughput of backend device (aids throttling calculations)"),
4949
4950     Option("filestore_expected_throughput_ops", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4951     .set_default(200)
4952     .set_description("Expected through of backend device in IOPS (aids throttling calculations)"),
4953
4954     Option("filestore_queue_max_delay_multiple", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4955     .set_default(0)
4956     .set_description(""),
4957
4958     Option("filestore_queue_high_delay_multiple", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4959     .set_default(0)
4960     .set_description(""),
4961
4962     Option("filestore_queue_max_delay_multiple_bytes", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4963     .set_default(0)
4964     .set_description(""),
4965
4966     Option("filestore_queue_high_delay_multiple_bytes", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4967     .set_default(0)
4968     .set_description(""),
4969
4970     Option("filestore_queue_max_delay_multiple_ops", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4971     .set_default(0)
4972     .set_description(""),
4973
4974     Option("filestore_queue_high_delay_multiple_ops", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4975     .set_default(0)
4976     .set_description(""),
4977
4978     Option("filestore_queue_low_threshhold", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4979     .set_default(0.3)
4980     .set_description(""),
4981
4982     Option("filestore_queue_high_threshhold", Option::TYPE_FLOAT, Option::LEVEL_DEV)
4983     .set_default(0.9)
4984     .set_description(""),
4985
4986     Option("filestore_op_threads", Option::TYPE_INT, Option::LEVEL_ADVANCED)
4987     .set_default(2)
4988     .set_description("Threads used to apply changes to backing file system"),
4989
4990     Option("filestore_op_thread_timeout", Option::TYPE_INT, Option::LEVEL_ADVANCED)
4991     .set_default(60)
4992     .set_description("Seconds before a worker thread is considered stalled"),
4993
4994     Option("filestore_op_thread_suicide_timeout", Option::TYPE_INT, Option::LEVEL_ADVANCED)
4995     .set_default(180)
4996     .set_description("Seconds before a worker thread is considered dead"),
4997
4998     Option("filestore_commit_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
4999     .set_default(600)
5000     .set_description("Seconds before backing file system is considered hung"),
5001
5002     Option("filestore_fiemap_threshold", Option::TYPE_SIZE, Option::LEVEL_DEV)
5003     .set_default(4_K)
5004     .set_description(""),
5005
5006     Option("filestore_merge_threshold", Option::TYPE_INT, Option::LEVEL_DEV)
5007     .set_default(-10)
5008     .set_description(""),
5009
5010     Option("filestore_split_multiple", Option::TYPE_INT, Option::LEVEL_DEV)
5011     .set_default(2)
5012     .set_description(""),
5013
5014     Option("filestore_split_rand_factor", Option::TYPE_UINT, Option::LEVEL_DEV)
5015     .set_default(20)
5016     .set_description(""),
5017
5018     Option("filestore_update_to", Option::TYPE_INT, Option::LEVEL_DEV)
5019     .set_default(1000)
5020     .set_description(""),
5021
5022     Option("filestore_blackhole", Option::TYPE_BOOL, Option::LEVEL_DEV)
5023     .set_default(false)
5024     .set_description(""),
5025
5026     Option("filestore_fd_cache_size", Option::TYPE_INT, Option::LEVEL_DEV)
5027     .set_default(128)
5028     .set_description(""),
5029
5030     Option("filestore_fd_cache_shards", Option::TYPE_INT, Option::LEVEL_DEV)
5031     .set_default(16)
5032     .set_description(""),
5033
5034     Option("filestore_ondisk_finisher_threads", Option::TYPE_INT, Option::LEVEL_DEV)
5035     .set_default(1)
5036     .set_description(""),
5037
5038     Option("filestore_apply_finisher_threads", Option::TYPE_INT, Option::LEVEL_DEV)
5039     .set_default(1)
5040     .set_description(""),
5041
5042     Option("filestore_dump_file", Option::TYPE_STR, Option::LEVEL_DEV)
5043     .set_default("")
5044     .set_description(""),
5045
5046     Option("filestore_kill_at", Option::TYPE_INT, Option::LEVEL_DEV)
5047     .set_default(0)
5048     .set_description(""),
5049
5050     Option("filestore_inject_stall", Option::TYPE_INT, Option::LEVEL_DEV)
5051     .set_default(0)
5052     .set_description(""),
5053
5054     Option("filestore_fail_eio", Option::TYPE_BOOL, Option::LEVEL_DEV)
5055     .set_default(true)
5056     .set_description(""),
5057
5058     Option("filestore_debug_verify_split", Option::TYPE_BOOL, Option::LEVEL_DEV)
5059     .set_default(false)
5060     .set_description(""),
5061
5062     Option("journal_dio", Option::TYPE_BOOL, Option::LEVEL_DEV)
5063     .set_default(true)
5064     .set_description(""),
5065
5066     Option("journal_aio", Option::TYPE_BOOL, Option::LEVEL_DEV)
5067     .set_default(true)
5068     .set_description(""),
5069
5070     Option("journal_force_aio", Option::TYPE_BOOL, Option::LEVEL_DEV)
5071     .set_default(false)
5072     .set_description(""),
5073
5074     Option("journal_block_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
5075     .set_default(4_K)
5076     .set_description(""),
5077
5078     Option("journal_block_align", Option::TYPE_BOOL, Option::LEVEL_DEV)
5079     .set_default(true)
5080     .set_description(""),
5081
5082     Option("journal_write_header_frequency", Option::TYPE_UINT, Option::LEVEL_DEV)
5083     .set_default(0)
5084     .set_description(""),
5085
5086     Option("journal_max_write_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
5087     .set_default(10_M)
5088     .set_description("Max bytes in flight to journal"),
5089
5090     Option("journal_max_write_entries", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5091     .set_default(100)
5092     .set_description("Max IOs in flight to journal"),
5093
5094     Option("journal_throttle_low_threshhold", Option::TYPE_FLOAT, Option::LEVEL_DEV)
5095     .set_default(0.6)
5096     .set_description(""),
5097
5098     Option("journal_throttle_high_threshhold", Option::TYPE_FLOAT, Option::LEVEL_DEV)
5099     .set_default(0.9)
5100     .set_description(""),
5101
5102     Option("journal_throttle_high_multiple", Option::TYPE_FLOAT, Option::LEVEL_DEV)
5103     .set_default(0)
5104     .set_description(""),
5105
5106     Option("journal_throttle_max_multiple", Option::TYPE_FLOAT, Option::LEVEL_DEV)
5107     .set_default(0)
5108     .set_description(""),
5109
5110     Option("journal_align_min_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
5111     .set_default(64_K)
5112     .set_description(""),
5113
5114     Option("journal_replay_from", Option::TYPE_INT, Option::LEVEL_DEV)
5115     .set_default(0)
5116     .set_description(""),
5117
5118     Option("mgr_stats_threshold", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5119     .set_default((int64_t)PerfCountersBuilder::PRIO_USEFUL)
5120     .set_description("Lowest perfcounter priority collected by mgr")
5121     .set_long_description("Daemons only set perf counter data to the manager "
5122                           "daemon if the counter has a priority higher than this.")
5123     .set_min_max((int64_t)PerfCountersBuilder::PRIO_DEBUGONLY,
5124                  (int64_t)PerfCountersBuilder::PRIO_CRITICAL + 1),
5125
5126     Option("journal_zero_on_create", Option::TYPE_BOOL, Option::LEVEL_DEV)
5127     .set_default(false)
5128     .set_description(""),
5129
5130     Option("journal_ignore_corruption", Option::TYPE_BOOL, Option::LEVEL_DEV)
5131     .set_default(false)
5132     .set_description(""),
5133
5134     Option("journal_discard", Option::TYPE_BOOL, Option::LEVEL_DEV)
5135     .set_default(false)
5136     .set_description(""),
5137
5138     Option("fio_dir", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5139     .set_default("/tmp/fio")
5140     .set_description(""),
5141
5142     Option("rados_mon_op_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
5143     .set_default(0)
5144     .set_description(""),
5145
5146     Option("rados_osd_op_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
5147     .set_default(0)
5148     .set_description(""),
5149
5150     Option("rados_tracing", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5151     .set_default(false)
5152     .set_description(""),
5153
5154     Option("cephadm_path", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5155     .set_default("/usr/sbin/cephadm")
5156     .add_service("mgr")
5157     .set_description("Path to cephadm utility"),
5158
5159     Option("mgr_module_path", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5160     .set_default(CEPH_DATADIR "/mgr")
5161     .add_service("mgr")
5162     .set_description("Filesystem path to manager modules."),
5163
5164     Option("mgr_initial_modules", Option::TYPE_STR, Option::LEVEL_BASIC)
5165     .set_default("restful iostat")
5166     .set_flag(Option::FLAG_NO_MON_UPDATE)
5167     .set_flag(Option::FLAG_CLUSTER_CREATE)
5168     .add_service("mon")
5169     .set_description("List of manager modules to enable when the cluster is "
5170                      "first started")
5171     .set_long_description("This list of module names is read by the monitor "
5172         "when the cluster is first started after installation, to populate "
5173         "the list of enabled manager modules.  Subsequent updates are done using "
5174         "the 'mgr module [enable|disable]' commands.  List may be comma "
5175         "or space separated."),
5176
5177     Option("mgr_data", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5178     .set_default("/var/lib/ceph/mgr/$cluster-$id")
5179     .set_flag(Option::FLAG_NO_MON_UPDATE)
5180     .add_service("mgr")
5181     .set_description("Filesystem path to the ceph-mgr data directory, used to "
5182                      "contain keyring."),
5183
5184     Option("mgr_tick_period", Option::TYPE_SECS, Option::LEVEL_ADVANCED)
5185     .set_default(2)
5186     .add_service("mgr")
5187     .set_description("Period in seconds of beacon messages to monitor"),
5188
5189     Option("mgr_stats_period", Option::TYPE_INT, Option::LEVEL_BASIC)
5190     .set_default(5)
5191     .add_service("mgr")
5192     .set_description("Period in seconds of OSD/MDS stats reports to manager")
5193     .set_long_description("Use this setting to control the granularity of "
5194                           "time series data collection from daemons.  Adjust "
5195                           "upwards if the manager CPU load is too high, or "
5196                           "if you simply do not require the most up to date "
5197                           "performance counter data."),
5198
5199     Option("mgr_client_bytes", Option::TYPE_SIZE, Option::LEVEL_DEV)
5200     .set_default(128_M)
5201     .add_service("mgr"),
5202
5203     Option("mgr_client_messages", Option::TYPE_UINT, Option::LEVEL_DEV)
5204     .set_default(512)
5205     .add_service("mgr"),
5206
5207     Option("mgr_osd_bytes", Option::TYPE_SIZE, Option::LEVEL_DEV)
5208     .set_default(512_M)
5209     .add_service("mgr"),
5210
5211     Option("mgr_osd_messages", Option::TYPE_UINT, Option::LEVEL_DEV)
5212     .set_default(8192)
5213     .add_service("mgr"),
5214
5215     Option("mgr_mds_bytes", Option::TYPE_SIZE, Option::LEVEL_DEV)
5216     .set_default(128_M)
5217     .add_service("mgr"),
5218
5219     Option("mgr_mds_messages", Option::TYPE_UINT, Option::LEVEL_DEV)
5220     .set_default(128)
5221     .add_service("mgr"),
5222
5223     Option("mgr_mon_bytes", Option::TYPE_SIZE, Option::LEVEL_DEV)
5224     .set_default(128_M)
5225     .add_service("mgr"),
5226
5227     Option("mgr_mon_messages", Option::TYPE_UINT, Option::LEVEL_DEV)
5228     .set_default(128)
5229     .add_service("mgr"),
5230
5231     Option("mgr_connect_retry_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
5232     .set_default(1.0)
5233     .add_service("common"),
5234
5235     Option("mgr_service_beacon_grace", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
5236     .set_default(60.0)
5237     .add_service("mgr")
5238     .set_description("Period in seconds from last beacon to manager dropping "
5239                      "state about a monitored service (RGW, rbd-mirror etc)"),
5240
5241     Option("mgr_client_service_daemon_unregister_timeout", Option::TYPE_FLOAT, Option::LEVEL_DEV)
5242     .set_default(1.0)
5243     .set_description("Time to wait during shutdown to deregister service with mgr"),
5244
5245     Option("mgr_debug_aggressive_pg_num_changes", Option::TYPE_BOOL, Option::LEVEL_DEV)
5246     .set_default(false)
5247     .set_description("Bypass most throttling and safety checks in pg[p]_num controller")
5248     .add_service("mgr"),
5249
5250     Option("mon_mgr_digest_period", Option::TYPE_INT, Option::LEVEL_DEV)
5251     .set_default(5)
5252     .add_service("mon")
5253     .set_description("Period in seconds between monitor-to-manager "
5254                      "health/status updates"),
5255
5256     Option("mon_mgr_beacon_grace", Option::TYPE_SECS, Option::LEVEL_ADVANCED)
5257     .set_default(30)
5258     .add_service("mon")
5259     .set_description("Period in seconds from last beacon to monitor marking "
5260                      "a manager daemon as failed"),
5261
5262     Option("mon_mgr_inactive_grace", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5263     .set_default(60)
5264     .add_service("mon")
5265     .set_description("Period in seconds after cluster creation during which "
5266                      "cluster may have no active manager")
5267     .set_long_description("This grace period enables the cluster to come "
5268                           "up cleanly without raising spurious health check "
5269                           "failures about managers that aren't online yet"),
5270
5271     Option("mon_mgr_mkfs_grace", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5272     .set_default(120)
5273     .add_service("mon")
5274     .set_description("Period in seconds that the cluster may have no active "
5275                      "manager before this is reported as an ERR rather than "
5276                      "a WARN"),
5277
5278     Option("throttler_perf_counter", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5279     .set_default(true)
5280     .set_description(""),
5281
5282     Option("event_tracing", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5283     .set_default(false)
5284     .set_description(""),
5285
5286     Option("bluestore_tracing", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5287     .set_default(false)
5288     .set_description("Enable bluestore event tracing."),
5289
5290     Option("bluestore_throttle_trace_rate", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
5291     .set_default(0)
5292     .set_description("Rate at which to sample bluestore transactions (per second)"),
5293
5294     Option("debug_deliberately_leak_memory", Option::TYPE_BOOL, Option::LEVEL_DEV)
5295     .set_default(false)
5296     .set_description(""),
5297
5298     Option("debug_asserts_on_shutdown", Option::TYPE_BOOL,Option::LEVEL_DEV)
5299     .set_default(false)
5300     .set_description("Enable certain asserts to check for refcounting bugs on shutdown; see http://tracker.ceph.com/issues/21738"),
5301
5302     Option("debug_asok_assert_abort", Option::TYPE_BOOL, Option::LEVEL_DEV)
5303     .set_default(false)
5304     .set_description("allow commands 'assert' and 'abort' via asok for testing crash dumps etc"),
5305
5306     Option("target_max_misplaced_ratio", Option::TYPE_FLOAT, Option::LEVEL_BASIC)
5307     .set_default(.05)
5308     .set_description("Max ratio of misplaced objects to target when throttling data rebalancing activity"),
5309
5310     Option("device_failure_prediction_mode", Option::TYPE_STR, Option::LEVEL_BASIC)
5311     .set_default("none")
5312     .set_flag(Option::FLAG_RUNTIME)
5313     .set_enum_allowed({"none", "local", "cloud"})
5314     .set_description("Method used to predict device failures")
5315     .set_long_description("To disable prediction, use 'none',  'local' uses a prediction model that runs inside the mgr daemon.  'cloud' will share metrics with a cloud service and query the service for devicelife expectancy."),
5316
5317     /*  KRB Authentication. */
5318     Option("gss_ktab_client_file", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5319     .set_default("/var/lib/ceph/$name/gss_client_$name.ktab")
5320     .set_description("GSS/KRB5 Keytab file for client authentication")
5321     .add_service({"mon", "osd"})
5322     .set_long_description("This sets the full path for the GSS/Kerberos client keytab file location."),
5323
5324     Option("gss_target_name", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5325     .set_default("ceph")
5326     .set_description("")
5327     .add_service({"mon", "osd"})
5328     .set_long_description("This sets the gss target service name."),
5329
5330     Option("debug_disable_randomized_ping", Option::TYPE_BOOL, Option::LEVEL_DEV)
5331     .set_default(false)
5332     .set_description("Disable heartbeat ping randomization for testing purposes"),
5333
5334     Option("debug_heartbeat_testing_span", Option::TYPE_INT, Option::LEVEL_DEV)
5335     .set_default(0)
5336     .set_description("Override 60 second periods for testing only"),
5337
5338     // ----------------------------
5339     // Crimson specific options
5340
5341     Option("crimson_osd_obc_lru_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
5342     .set_default(10)
5343     .set_description("Number of obcs to cache")
5344
5345   });
5346 }
5347
5348 std::vector<Option> get_rgw_options() {
5349   return std::vector<Option>({
5350     Option("rgw_acl_grants_max_num", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5351     .set_default(100)
5352     .set_description("Max number of ACL grants in a single request"),
5353
5354     Option("rgw_cors_rules_max_num", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5355     .set_default(100)
5356     .set_description("Max number of cors rules in a single request"),
5357
5358     Option("rgw_delete_multi_obj_max_num", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5359     .set_default(1000)
5360     .set_description("Max number of objects in a single multi-object delete request"),
5361
5362     Option("rgw_website_routing_rules_max_num", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5363     .set_default(50)
5364     .set_description("Max number of website routing rules in a single request"),
5365
5366     Option("rgw_rados_tracing", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5367     .set_default(false)
5368     .set_description("true if LTTng-UST tracepoints should be enabled"),
5369
5370     Option("rgw_op_tracing", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5371     .set_default(false)
5372     .set_description("true if LTTng-UST tracepoints should be enabled"),
5373
5374     Option("rgw_max_chunk_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
5375     .set_default(4_M)
5376     .set_description("Set RGW max chunk size")
5377     .set_long_description(
5378         "The chunk size is the size of RADOS I/O requests that RGW sends when accessing "
5379         "data objects. RGW read and write operation will never request more than this amount "
5380         "in a single request. This also defines the rgw object head size, as head operations "
5381         "need to be atomic, and anything larger than this would require more than a single "
5382         "operation."),
5383
5384     Option("rgw_put_obj_min_window_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
5385     .set_default(16_M)
5386     .set_description("The minimum RADOS write window size (in bytes).")
5387     .set_long_description(
5388         "The window size determines the total concurrent RADOS writes of a single rgw object. "
5389         "When writing an object RGW will send multiple chunks to RADOS. The total size of the "
5390         "writes does not exceed the window size. The window size can be automatically "
5391         "in order to better utilize the pipe.")
5392     .add_see_also({"rgw_put_obj_max_window_size", "rgw_max_chunk_size"}),
5393
5394     Option("rgw_put_obj_max_window_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
5395     .set_default(64_M)
5396     .set_description("The maximum RADOS write window size (in bytes).")
5397     .set_long_description("The window size may be dynamically adjusted, but will not surpass this value.")
5398     .add_see_also({"rgw_put_obj_min_window_size", "rgw_max_chunk_size"}),
5399
5400     Option("rgw_max_put_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
5401     .set_default(5_G)
5402     .set_description("Max size (in bytes) of regular (non multi-part) object upload.")
5403     .set_long_description(
5404         "Plain object upload is capped at this amount of data. In order to upload larger "
5405         "objects, a special upload mechanism is required. The S3 API provides the "
5406         "multi-part upload, and Swift provides DLO and SLO."),
5407
5408     Option("rgw_max_put_param_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
5409     .set_default(1_M)
5410     .set_description("The maximum size (in bytes) of data input of certain RESTful requests."),
5411
5412     Option("rgw_max_attr_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
5413     .set_default(0)
5414     .set_description("The maximum length of metadata value. 0 skips the check"),
5415
5416     Option("rgw_max_attr_name_len", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
5417     .set_default(0)
5418     .set_description("The maximum length of metadata name. 0 skips the check"),
5419
5420     Option("rgw_max_attrs_num_in_req", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
5421     .set_default(0)
5422     .set_description("The maximum number of metadata items that can be put via single request"),
5423
5424     Option("rgw_override_bucket_index_max_shards", Option::TYPE_UINT, Option::LEVEL_DEV)
5425     .set_default(0)
5426     .set_description("The default number of bucket index shards for newly-created "
5427         "buckets. This value overrides bucket_index_max_shards stored in the zone. "
5428         "Setting this value in the zone is preferred, because it applies globally "
5429         "to all radosgw daemons running in the zone."),
5430
5431     Option("rgw_bucket_index_max_aio", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
5432     .set_default(128)
5433     .set_description("Max number of concurrent RADOS requests when handling bucket shards."),
5434
5435     Option("rgw_enable_quota_threads", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5436     .set_default(true)
5437     .set_description("Enables the quota maintenance thread.")
5438     .set_long_description(
5439         "The quota maintenance thread is responsible for quota related maintenance work. "
5440         "The thread itself can be disabled, but in order for quota to work correctly, at "
5441         "least one RGW in each zone needs to have this thread running. Having the thread "
5442         "enabled on multiple RGW processes within the same zone can spread "
5443         "some of the maintenance work between them.")
5444     .add_see_also({"rgw_enable_gc_threads", "rgw_enable_lc_threads"}),
5445
5446     Option("rgw_enable_gc_threads", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5447     .set_default(true)
5448     .set_description("Enables the garbage collection maintenance thread.")
5449     .set_long_description(
5450         "The garbage collection maintenance thread is responsible for garbage collector "
5451         "maintenance work. The thread itself can be disabled, but in order for garbage "
5452         "collection to work correctly, at least one RGW in each zone needs to have this "
5453         "thread running.  Having the thread enabled on multiple RGW processes within the "
5454         "same zone can spread some of the maintenance work between them.")
5455     .add_see_also({"rgw_enable_quota_threads", "rgw_enable_lc_threads"}),
5456
5457     Option("rgw_enable_lc_threads", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5458     .set_default(true)
5459     .set_description("Enables the lifecycle maintenance thread. This is required on at least one rgw for each zone.")
5460     .set_long_description(
5461         "The lifecycle maintenance thread is responsible for lifecycle related maintenance "
5462         "work. The thread itself can be disabled, but in order for lifecycle to work "
5463         "correctly, at least one RGW in each zone needs to have this thread running. Having"
5464         "the thread enabled on multiple RGW processes within the same zone can spread "
5465         "some of the maintenance work between them.")
5466     .add_see_also({"rgw_enable_gc_threads", "rgw_enable_quota_threads"}),
5467
5468     Option("rgw_data", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5469     .set_default("/var/lib/ceph/radosgw/$cluster-$id")
5470     .set_flag(Option::FLAG_NO_MON_UPDATE)
5471     .set_description("Alternative location for RGW configuration.")
5472     .set_long_description(
5473         "If this is set, the different Ceph system configurables (such as the keyring file "
5474         "will be located in the path that is specified here. "),
5475
5476     Option("rgw_enable_apis", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5477     .set_default("s3, s3website, swift, swift_auth, admin, sts, iam, pubsub")
5478     .set_description("A list of set of RESTful APIs that rgw handles."),
5479
5480     Option("rgw_cache_enabled", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5481     .set_default(true)
5482     .set_description("Enable RGW metadata cache.")
5483     .set_long_description(
5484         "The metadata cache holds metadata entries that RGW requires for processing "
5485         "requests. Metadata entries can be user info, bucket info, and bucket instance "
5486         "info. If not found in the cache, entries will be fetched from the backing "
5487         "RADOS store.")
5488     .add_see_also("rgw_cache_lru_size"),
5489
5490     Option("rgw_cache_lru_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5491     .set_default(10000)
5492     .set_description("Max number of items in RGW metadata cache.")
5493     .set_long_description(
5494         "When full, the RGW metadata cache evicts least recently used entries.")
5495     .add_see_also("rgw_cache_enabled"),
5496
5497     Option("rgw_socket_path", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5498     .set_default("")
5499     .set_description("RGW FastCGI socket path (for FastCGI over Unix domain sockets).")
5500     .add_see_also("rgw_fcgi_socket_backlog"),
5501
5502     Option("rgw_host", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5503     .set_default("")
5504     .set_description("RGW FastCGI host name (for FastCGI over TCP)")
5505     .add_see_also({"rgw_port", "rgw_fcgi_socket_backlog"}),
5506
5507     Option("rgw_port", Option::TYPE_STR, Option::LEVEL_BASIC)
5508     .set_default("")
5509     .set_description("RGW FastCGI port number (for FastCGI over TCP)")
5510     .add_see_also({"rgw_host", "rgw_fcgi_socket_backlog"}),
5511
5512     Option("rgw_dns_name", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5513     .set_default("")
5514     .set_description("The host name that RGW uses.")
5515     .set_long_description(
5516         "This is Needed for virtual hosting of buckets to work properly, unless configured "
5517         "via zonegroup configuration."),
5518
5519     Option("rgw_dns_s3website_name", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5520     .set_default("")
5521     .set_description("The host name that RGW uses for static websites (S3)")
5522     .set_long_description(
5523         "This is needed for virtual hosting of buckets, unless configured via zonegroup "
5524         "configuration."),
5525
5526     Option("rgw_numa_node", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5527     .set_default(-1)
5528     .set_flag(Option::FLAG_STARTUP)
5529     .set_description("set rgw's cpu affinity to a numa node (-1 for none)"),
5530
5531     Option("rgw_service_provider_name", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5532     .set_default("")
5533     .set_description("Service provider name which is contained in http response headers")
5534     .set_long_description(
5535         "As S3 or other cloud storage providers do, http response headers should contain the name of the provider. "
5536         "This name will be placed in http header 'Server'."),
5537
5538     Option("rgw_content_length_compat", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5539     .set_default(false)
5540     .set_description("Multiple content length headers compatibility")
5541     .set_long_description(
5542         "Try to handle requests with abiguous multiple content length headers "
5543         "(Content-Length, Http-Content-Length)."),
5544
5545     Option("rgw_relaxed_region_enforcement", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5546     .set_default(false)
5547     .set_description("Disable region constraint enforcement")
5548     .set_long_description(
5549         "Enable requests such as bucket creation to succeed irrespective of region restrictions (Jewel compat)."),
5550
5551     Option("rgw_lifecycle_work_time", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5552     .set_default("00:00-06:00")
5553     .set_description("Lifecycle allowed work time")
5554     .set_long_description("Local time window in which the lifecycle maintenance thread can work."),
5555
5556     Option("rgw_lc_lock_max_time", Option::TYPE_INT, Option::LEVEL_DEV)
5557     .set_default(90)
5558     .set_description(""),
5559
5560     Option("rgw_lc_thread_delay", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5561     .set_default(0)
5562     .set_description("Delay after processing of bucket listing chunks (i.e., per 1000 entries) in milliseconds"),
5563
5564     Option("rgw_lc_max_worker", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5565     .set_default(3)
5566     .set_description("Number of LCWorker tasks that will be run in parallel")
5567     .set_long_description(
5568       "Number of LCWorker tasks that will run in parallel--used to permit >1 "
5569       "bucket/index shards to be processed simultaneously"),
5570
5571     Option("rgw_lc_max_wp_worker", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5572     .set_default(3)
5573     .set_description("Number of workpool threads per LCWorker")
5574     .set_long_description(
5575       "Number of threads in per-LCWorker workpools--used to accelerate "
5576       "per-bucket processing"),
5577
5578     Option("rgw_lc_max_objs", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5579     .set_default(32)
5580     .set_description("Number of lifecycle data shards")
5581     .set_long_description(
5582           "Number of RADOS objects to use for storing lifecycle index. This "
5583           "affects concurrency of lifecycle maintenance, as shards can be "
5584           "processed in parallel."),
5585
5586     Option("rgw_lc_max_rules", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
5587     .set_default(1000)
5588     .set_description("Max number of lifecycle rules set on one bucket")
5589     .set_long_description("Number of lifecycle rules set on one bucket should be limited."),
5590
5591     Option("rgw_lc_debug_interval", Option::TYPE_INT, Option::LEVEL_DEV)
5592     .set_default(-1)
5593     .set_description(""),
5594
5595     Option("rgw_mp_lock_max_time", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5596     .set_default(600)
5597     .set_description("Multipart upload max completion time")
5598     .set_long_description(
5599         "Time length to allow completion of a multipart upload operation. This is done "
5600         "to prevent concurrent completions on the same object with the same upload id."),
5601
5602     Option("rgw_script_uri", Option::TYPE_STR, Option::LEVEL_DEV)
5603     .set_default("")
5604     .set_description(""),
5605
5606     Option("rgw_request_uri", Option::TYPE_STR, Option::LEVEL_DEV)
5607     .set_default("")
5608     .set_description(""),
5609
5610     Option("rgw_ignore_get_invalid_range", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5611     .set_default(false)
5612     .set_description("Treat invalid (e.g., negative) range request as full")
5613     .set_long_description("Treat invalid (e.g., negative) range request "
5614                           "as request for the full object (AWS compatibility)"),
5615
5616     Option("rgw_swift_url", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5617     .set_default("")
5618     .set_description("Swift-auth storage URL")
5619     .set_long_description(
5620         "Used in conjunction with rgw internal swift authentication. This affects the "
5621         "X-Storage-Url response header value.")
5622     .add_see_also("rgw_swift_auth_entry"),
5623
5624     Option("rgw_swift_url_prefix", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5625     .set_default("swift")
5626     .set_description("Swift URL prefix")
5627     .set_long_description("The URL path prefix for swift requests."),
5628
5629     Option("rgw_swift_auth_url", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5630     .set_default("")
5631     .set_description("Swift auth URL")
5632     .set_long_description(
5633         "Default url to which RGW connects and verifies tokens for v1 auth (if not using "
5634         "internal swift auth)."),
5635
5636     Option("rgw_swift_auth_entry", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5637     .set_default("auth")
5638     .set_description("Swift auth URL prefix")
5639     .set_long_description("URL path prefix for internal swift auth requests.")
5640     .add_see_also("rgw_swift_url"),
5641
5642     Option("rgw_swift_tenant_name", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5643     .set_default("")
5644     .set_description("Swift tenant name")
5645     .set_long_description("Tenant name that is used when constructing the swift path.")
5646     .add_see_also("rgw_swift_account_in_url"),
5647
5648     Option("rgw_swift_account_in_url", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5649     .set_default(false)
5650     .set_description("Swift account encoded in URL")
5651     .set_long_description("Whether the swift account is encoded in the uri path (AUTH_<account>).")
5652     .add_see_also("rgw_swift_tenant_name"),
5653
5654     Option("rgw_swift_enforce_content_length", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5655     .set_default(false)
5656     .set_description("Send content length when listing containers (Swift)")
5657     .set_long_description(
5658         "Whether content length header is needed when listing containers. When this is "
5659         "set to false, RGW will send extra info for each entry in the response."),
5660
5661     Option("rgw_keystone_url", Option::TYPE_STR, Option::LEVEL_BASIC)
5662     .set_default("")
5663     .set_description("The URL to the Keystone server."),
5664
5665     Option("rgw_keystone_admin_token", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5666     .set_default("")
5667     .set_description("DEPRECATED: The admin token (shared secret) that is used for the Keystone requests."),
5668
5669     Option("rgw_keystone_admin_token_path", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5670     .set_default("")
5671     .set_description("Path to a file containing the admin token (shared secret) that is used for the Keystone requests."),
5672
5673     Option("rgw_keystone_admin_user", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5674     .set_default("")
5675     .set_description("Keystone admin user."),
5676
5677     Option("rgw_keystone_admin_password", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5678     .set_default("")
5679     .set_description("DEPRECATED: Keystone admin password."),
5680
5681     Option("rgw_keystone_admin_password_path", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5682     .set_default("")
5683     .set_description("Path to a file containing the Keystone admin password."),
5684
5685     Option("rgw_keystone_admin_tenant", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5686     .set_default("")
5687     .set_description("Keystone admin user tenant."),
5688
5689     Option("rgw_keystone_admin_project", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5690     .set_default("")
5691     .set_description("Keystone admin user project (for Keystone v3)."),
5692
5693     Option("rgw_keystone_admin_domain", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5694     .set_default("")
5695     .set_description("Keystone admin user domain (for Keystone v3)."),
5696
5697     Option("rgw_keystone_barbican_user", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5698     .set_default("")
5699     .set_description("Keystone user to access barbican secrets."),
5700
5701     Option("rgw_keystone_barbican_password", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5702     .set_default("")
5703     .set_description("Keystone password for barbican user."),
5704
5705     Option("rgw_keystone_barbican_tenant", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5706     .set_default("")
5707     .set_description("Keystone barbican user tenant (Keystone v2.0)."),
5708
5709     Option("rgw_keystone_barbican_project", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5710     .set_default("")
5711     .set_description("Keystone barbican user project (Keystone v3)."),
5712
5713     Option("rgw_keystone_barbican_domain", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5714     .set_default("")
5715     .set_description("Keystone barbican user domain."),
5716
5717     Option("rgw_keystone_api_version", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5718     .set_default(2)
5719     .set_description("Version of Keystone API to use (2 or 3)."),
5720
5721     Option("rgw_keystone_accepted_roles", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5722     .set_default("Member, admin")
5723     .set_description("Only users with one of these roles will be served when doing Keystone authentication."),
5724
5725     Option("rgw_keystone_accepted_admin_roles", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5726     .set_default("")
5727     .set_description("List of roles allowing user to gain admin privileges (Keystone)."),
5728
5729     Option("rgw_keystone_token_cache_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5730     .set_default(10000)
5731     .set_description("Keystone token cache size")
5732     .set_long_description(
5733         "Max number of Keystone tokens that will be cached. Token that is not cached "
5734         "requires RGW to access the Keystone server when authenticating."),
5735
5736     Option("rgw_keystone_verify_ssl", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5737     .set_default(true)
5738     .set_description("Should RGW verify the Keystone server SSL certificate."),
5739
5740     Option("rgw_keystone_implicit_tenants", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5741     .set_default("false")
5742     .set_enum_allowed( { "false", "true", "swift", "s3", "both", "0", "1", "none" } )
5743     .set_description("RGW Keystone implicit tenants creation")
5744     .set_long_description(
5745         "Implicitly create new users in their own tenant with the same name when "
5746         "authenticating via Keystone.  Can be limited to s3 or swift only."),
5747
5748     Option("rgw_cross_domain_policy", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5749     .set_default("<allow-access-from domain=\"*\" secure=\"false\" />")
5750     .set_description("RGW handle cross domain policy")
5751     .set_long_description("Returned cross domain policy when accessing the crossdomain.xml "
5752                           "resource (Swift compatiility)."),
5753
5754     Option("rgw_healthcheck_disabling_path", Option::TYPE_STR, Option::LEVEL_DEV)
5755     .set_default("")
5756     .set_description("Swift health check api can be disabled if a file can be accessed in this path."),
5757
5758     Option("rgw_s3_auth_use_rados", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5759     .set_default(true)
5760     .set_description("Should S3 authentication use credentials stored in RADOS backend."),
5761
5762     Option("rgw_s3_auth_use_keystone", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5763     .set_default(false)
5764     .set_description("Should S3 authentication use Keystone."),
5765
5766     Option("rgw_s3_auth_order", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5767      .set_default("sts, external, local")
5768      .set_description("Authentication strategy order to use for s3 authentication")
5769      .set_long_description(
5770           "Order of authentication strategies to try for s3 authentication, the allowed "
5771            "options are a comma separated list of engines external, local. The "
5772            "default order is to try all the externally configured engines before "
5773            "attempting local rados based authentication"),
5774
5775     Option("rgw_barbican_url", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5776     .set_default("")
5777     .set_description("URL to barbican server."),
5778
5779     Option("rgw_ldap_uri", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5780     .set_default("ldaps://<ldap.your.domain>")
5781     .set_description("Space-separated list of LDAP servers in URI format."),
5782
5783     Option("rgw_ldap_binddn", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5784     .set_default("uid=admin,cn=users,dc=example,dc=com")
5785     .set_description("LDAP entry RGW will bind with (user match)."),
5786
5787     Option("rgw_ldap_searchdn", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5788     .set_default("cn=users,cn=accounts,dc=example,dc=com")
5789     .set_description("LDAP search base (basedn)."),
5790
5791     Option("rgw_ldap_dnattr", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5792     .set_default("uid")
5793     .set_description("LDAP attribute containing RGW user names (to form binddns)."),
5794
5795     Option("rgw_ldap_secret", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5796     .set_default("/etc/openldap/secret")
5797     .set_description("Path to file containing credentials for rgw_ldap_binddn."),
5798
5799     Option("rgw_s3_auth_use_ldap", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5800     .set_default(false)
5801     .set_description("Should S3 authentication use LDAP."),
5802
5803     Option("rgw_ldap_searchfilter", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5804     .set_default("")
5805     .set_description("LDAP search filter."),
5806
5807     Option("rgw_opa_url", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5808     .set_default("")
5809     .set_description("URL to OPA server."),
5810
5811     Option("rgw_opa_token", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5812     .set_default("")
5813     .set_description("The Bearer token OPA uses to authenticate client requests."),
5814
5815     Option("rgw_opa_verify_ssl", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5816     .set_default(true)
5817     .set_description("Should RGW verify the OPA server SSL certificate."),
5818
5819     Option("rgw_use_opa_authz", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5820     .set_default(false)
5821     .set_description("Should OPA be used to authorize client requests."),
5822
5823     Option("rgw_admin_entry", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5824     .set_default("admin")
5825     .set_description("Path prefix to be used for accessing RGW RESTful admin API."),
5826
5827     Option("rgw_enforce_swift_acls", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5828     .set_default(true)
5829     .set_description("RGW enforce swift acls")
5830     .set_long_description(
5831         "Should RGW enforce special Swift-only ACLs. Swift has a special ACL that gives "
5832         "permission to access all objects in a container."),
5833
5834     Option("rgw_swift_token_expiration", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5835     .set_default(1_day)
5836     .set_description("Expiration time (in seconds) for token generated through RGW Swift auth."),
5837
5838     Option("rgw_print_continue", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5839     .set_default(true)
5840     .set_description("RGW support of 100-continue")
5841     .set_long_description(
5842         "Should RGW explicitly send 100 (continue) responses. This is mainly relevant when "
5843         "using FastCGI, as some FastCGI modules do not fully support this feature."),
5844
5845     Option("rgw_print_prohibited_content_length", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5846     .set_default(false)
5847     .set_description("RGW RFC-7230 compatibility")
5848     .set_long_description(
5849         "Specifies whether RGW violates RFC 7230 and sends Content-Length with 204 or 304 "
5850         "statuses."),
5851
5852     Option("rgw_remote_addr_param", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5853     .set_default("REMOTE_ADDR")
5854     .set_description("HTTP header that holds the remote address in incoming requests.")
5855     .set_long_description(
5856         "RGW will use this header to extract requests origin. When RGW runs behind "
5857         "a reverse proxy, the remote address header will point at the proxy's address "
5858         "and not at the originator's address. Therefore it is sometimes possible to "
5859         "have the proxy add the originator's address in a separate HTTP header, which "
5860         "will allow RGW to log it correctly."
5861         )
5862     .add_see_also("rgw_enable_ops_log"),
5863
5864     Option("rgw_op_thread_timeout", Option::TYPE_INT, Option::LEVEL_DEV)
5865     .set_default(10*60)
5866     .set_description("Timeout for async rados coroutine operations."),
5867
5868     Option("rgw_op_thread_suicide_timeout", Option::TYPE_INT, Option::LEVEL_DEV)
5869     .set_default(0)
5870     .set_description(""),
5871
5872     Option("rgw_thread_pool_size", Option::TYPE_INT, Option::LEVEL_BASIC)
5873     .set_default(512)
5874     .set_description("RGW requests handling thread pool size.")
5875     .set_long_description(
5876         "This parameter determines the number of concurrent requests RGW can process "
5877         "when using either the civetweb, or the fastcgi frontends. The higher this "
5878         "number is, RGW will be able to deal with more concurrent requests at the "
5879         "cost of more resource utilization."),
5880
5881     Option("rgw_num_control_oids", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5882     .set_default(8)
5883     .set_description("Number of control objects used for cross-RGW communication.")
5884     .set_long_description(
5885         "RGW uses certain control objects to send messages between different RGW "
5886         "processes running on the same zone. These messages include metadata cache "
5887         "invalidation info that is being sent when metadata is modified (such as "
5888         "user or bucket information). A higher number of control objects allows "
5889         "better concurrency of these messages, at the cost of more resource "
5890         "utilization."),
5891
5892     Option("rgw_num_rados_handles", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
5893     .set_default(1)
5894     .set_description("Number of librados handles that RGW uses.")
5895     .set_long_description(
5896         "This param affects the number of separate librados handles it uses to "
5897         "connect to the RADOS backend, which directly affects the number of connections "
5898         "RGW will have to each OSD. A higher number affects resource utilization."),
5899
5900     Option("rgw_verify_ssl", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5901     .set_default(true)
5902     .set_description("Should RGW verify SSL when connecing to a remote HTTP server")
5903     .set_long_description(
5904         "RGW can send requests to other RGW servers (e.g., in multi-site sync work). "
5905         "This configurable selects whether RGW should verify the certificate for "
5906         "the remote peer and host.")
5907     .add_see_also("rgw_keystone_verify_ssl"),
5908
5909     Option("rgw_nfs_lru_lanes", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5910     .set_default(5)
5911     .set_description(""),
5912
5913     Option("rgw_nfs_lru_lane_hiwat", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5914     .set_default(911)
5915     .set_description(""),
5916
5917     Option("rgw_nfs_fhcache_partitions", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5918     .set_default(3)
5919     .set_description(""),
5920
5921     Option("rgw_nfs_fhcache_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5922     .set_default(2017)
5923     .set_description(""),
5924
5925     Option("rgw_nfs_namespace_expire_secs", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5926     .set_default(300)
5927     .set_min(1)
5928     .set_description(""),
5929
5930     Option("rgw_nfs_max_gc", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5931     .set_default(300)
5932     .set_min(1)
5933     .set_description(""),
5934
5935     Option("rgw_nfs_write_completion_interval_s", Option::TYPE_INT, Option::LEVEL_ADVANCED)
5936     .set_default(10)
5937     .set_description(""),
5938
5939     Option("rgw_nfs_s3_fast_attrs", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5940     .set_default(false)
5941     .set_description("use fast S3 attrs from bucket index (immutable only)")
5942     .set_long_description("use fast S3 attrs from bucket index (assumes NFS "
5943                           "mounts are immutable)"),
5944
5945     Option("rgw_nfs_run_gc_threads", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5946     .set_default(false)
5947     .set_description("run GC threads in librgw (default off)"),
5948
5949     Option("rgw_nfs_run_lc_threads", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5950     .set_default(false)
5951     .set_description("run lifecycle threads in librgw (default off)"),
5952
5953     Option("rgw_nfs_run_quota_threads", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5954     .set_default(false)
5955     .set_description("run quota threads in librgw (default off)"),
5956
5957     Option("rgw_nfs_run_sync_thread", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
5958     .set_default(false)
5959     .set_description("run sync thread in librgw (default off)"),
5960
5961     Option("rgw_rados_pool_autoscale_bias", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
5962     .set_default(4.0)
5963     .set_min_max(0.01, 100000.0)
5964     .set_description("pg_autoscale_bias value for RGW metadata (omap-heavy) pools"),
5965
5966     Option("rgw_rados_pool_pg_num_min", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
5967     .set_default(8)
5968     .set_min_max(1, 1024)
5969     .set_description("pg_num_min value for RGW metadata (omap-heavy) pools"),
5970
5971     Option("rgw_rados_pool_recovery_priority", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
5972     .set_default(5)
5973     .set_min_max(-10, 10)
5974     .set_description("recovery_priority value for RGW metadata (omap-heavy) pools"),
5975
5976     Option("rgw_zone", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5977     .set_default("")
5978     .set_description("Zone name")
5979     .add_see_also({"rgw_zonegroup", "rgw_realm"}),
5980
5981     Option("rgw_zone_root_pool", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5982     .set_default(".rgw.root")
5983     .set_description("Zone root pool name")
5984     .set_long_description(
5985         "The zone root pool, is the pool where the RGW zone configuration located."
5986     )
5987     .add_see_also({"rgw_zonegroup_root_pool", "rgw_realm_root_pool", "rgw_period_root_pool"}),
5988
5989     Option("rgw_default_zone_info_oid", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5990     .set_default("default.zone")
5991     .set_description("Default zone info object id")
5992     .set_long_description(
5993         "Name of the RADOS object that holds the default zone information."
5994     ),
5995
5996     Option("rgw_region", Option::TYPE_STR, Option::LEVEL_ADVANCED)
5997     .set_default("")
5998     .set_description("Region name")
5999     .set_long_description(
6000         "Obsolete config option. The rgw_zonegroup option should be used instead.")
6001     .add_see_also("rgw_zonegroup"),
6002
6003     Option("rgw_region_root_pool", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6004     .set_default(".rgw.root")
6005     .set_description("Region root pool")
6006     .set_long_description(
6007         "Obsolete config option. The rgw_zonegroup_root_pool should be used instead.")
6008     .add_see_also("rgw_zonegroup_root_pool"),
6009
6010     Option("rgw_default_region_info_oid", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6011     .set_default("default.region")
6012     .set_description("Default region info object id")
6013     .set_long_description(
6014         "Obsolete config option. The rgw_default_zonegroup_info_oid should be used instead.")
6015     .add_see_also("rgw_default_zonegroup_info_oid"),
6016
6017     Option("rgw_zonegroup", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6018     .set_default("")
6019     .set_description("Zonegroup name")
6020     .add_see_also({"rgw_zone", "rgw_realm"}),
6021
6022     Option("rgw_zonegroup_root_pool", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6023     .set_default(".rgw.root")
6024     .set_description("Zonegroup root pool")
6025     .set_long_description(
6026         "The zonegroup root pool, is the pool where the RGW zonegroup configuration located."
6027     )
6028     .add_see_also({"rgw_zone_root_pool", "rgw_realm_root_pool", "rgw_period_root_pool"}),
6029
6030     Option("rgw_default_zonegroup_info_oid", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6031     .set_default("default.zonegroup")
6032     .set_description(""),
6033
6034     Option("rgw_realm", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6035     .set_default("")
6036     .set_description(""),
6037
6038     Option("rgw_realm_root_pool", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6039     .set_default(".rgw.root")
6040     .set_description("Realm root pool")
6041     .set_long_description(
6042         "The realm root pool, is the pool where the RGW realm configuration located."
6043     )
6044     .add_see_also({"rgw_zonegroup_root_pool", "rgw_zone_root_pool", "rgw_period_root_pool"}),
6045
6046     Option("rgw_default_realm_info_oid", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6047     .set_default("default.realm")
6048     .set_description(""),
6049
6050     Option("rgw_period_root_pool", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6051     .set_default(".rgw.root")
6052     .set_description("Period root pool")
6053     .set_long_description(
6054         "The period root pool, is the pool where the RGW period configuration located."
6055     )
6056     .add_see_also({"rgw_zonegroup_root_pool", "rgw_zone_root_pool", "rgw_realm_root_pool"}),
6057
6058     Option("rgw_period_latest_epoch_info_oid", Option::TYPE_STR, Option::LEVEL_DEV)
6059     .set_default(".latest_epoch")
6060     .set_description(""),
6061
6062     Option("rgw_log_nonexistent_bucket", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6063     .set_default(false)
6064     .set_description("Should RGW log operations on bucket that does not exist")
6065     .set_long_description(
6066         "This config option applies to the ops log. When this option is set, the ops log "
6067         "will log operations that are sent to non existing buckets. These operations "
6068         "inherently fail, and do not correspond to a specific user.")
6069     .add_see_also("rgw_enable_ops_log"),
6070
6071     Option("rgw_log_object_name", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6072     .set_default("%Y-%m-%d-%H-%i-%n")
6073     .set_description("Ops log object name format")
6074     .set_long_description(
6075         "Defines the format of the RADOS objects names that ops log uses to store ops "
6076         "log data")
6077     .add_see_also("rgw_enable_ops_log"),
6078
6079     Option("rgw_log_object_name_utc", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6080     .set_default(false)
6081     .set_description("Should ops log object name based on UTC")
6082     .set_long_description(
6083         "If set, the names of the RADOS objects that hold the ops log data will be based "
6084         "on UTC time zone. If not set, it will use the local time zone.")
6085     .add_see_also({"rgw_enable_ops_log", "rgw_log_object_name"}),
6086
6087     Option("rgw_usage_max_shards", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6088     .set_default(32)
6089     .set_description("Number of shards for usage log.")
6090     .set_long_description(
6091         "The number of RADOS objects that RGW will use in order to store the usage log "
6092         "data.")
6093     .add_see_also("rgw_enable_usage_log"),
6094
6095     Option("rgw_usage_max_user_shards", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6096     .set_default(1)
6097     .set_min(1)
6098     .set_description("Number of shards for single user in usage log")
6099     .set_long_description(
6100         "The number of shards that a single user will span over in the usage log.")
6101     .add_see_also("rgw_enable_usage_log"),
6102
6103     Option("rgw_enable_ops_log", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6104     .set_default(false)
6105     .set_description("Enable ops log")
6106     .add_see_also({"rgw_log_nonexistent_bucket", "rgw_log_object_name", "rgw_ops_log_rados",
6107                "rgw_ops_log_socket_path"}),
6108
6109     Option("rgw_enable_usage_log", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6110     .set_default(false)
6111     .set_description("Enable usage log")
6112     .add_see_also("rgw_usage_max_shards"),
6113
6114     Option("rgw_ops_log_rados", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6115     .set_default(true)
6116     .set_description("Use RADOS for ops log")
6117     .set_long_description(
6118        "If set, RGW will store ops log information in RADOS.")
6119     .add_see_also({"rgw_enable_ops_log"}),
6120
6121     Option("rgw_ops_log_socket_path", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6122     .set_default("")
6123     .set_description("Unix domain socket path for ops log.")
6124     .set_long_description(
6125         "Path to unix domain socket that RGW will listen for connection on. When connected, "
6126         "RGW will send ops log data through it.")
6127     .add_see_also({"rgw_enable_ops_log", "rgw_ops_log_data_backlog"}),
6128
6129     Option("rgw_ops_log_data_backlog", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
6130     .set_default(5 << 20)
6131     .set_description("Ops log socket backlog")
6132     .set_long_description(
6133         "Maximum amount of data backlog that RGW can keep when ops log is configured to "
6134         "send info through unix domain socket. When data backlog is higher than this, "
6135         "ops log entries will be lost. In order to avoid ops log information loss, the "
6136         "listener needs to clear data (by reading it) quickly enough.")
6137     .add_see_also({"rgw_enable_ops_log", "rgw_ops_log_socket_path"}),
6138
6139     Option("rgw_fcgi_socket_backlog", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6140     .set_default(1024)
6141     .set_description("FastCGI socket connection backlog")
6142     .set_long_description(
6143         "Size of FastCGI connection backlog. This reflects the maximum number of new "
6144         "connection requests that RGW can handle concurrently without dropping any. ")
6145     .add_see_also({"rgw_host", "rgw_socket_path"}),
6146
6147     Option("rgw_usage_log_flush_threshold", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6148     .set_default(1024)
6149     .set_description("Number of entries in usage log before flushing")
6150     .set_long_description(
6151         "This is the max number of entries that will be held in the usage log, before it "
6152         "will be flushed to the backend. Note that the usage log is periodically flushed, "
6153         "even if number of entries does not reach this threshold. A usage log entry "
6154         "corresponds to one or more operations on a single bucket.i")
6155     .add_see_also({"rgw_enable_usage_log", "rgw_usage_log_tick_interval"}),
6156
6157     Option("rgw_usage_log_tick_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6158     .set_default(30)
6159     .set_description("Number of seconds between usage log flush cycles")
6160     .set_long_description(
6161         "The number of seconds between consecutive usage log flushes. The usage log will "
6162         "also flush itself to the backend if the number of pending entries reaches a "
6163         "certain threshold.")
6164     .add_see_also({"rgw_enable_usage_log", "rgw_usage_log_flush_threshold"}),
6165
6166     Option("rgw_init_timeout", Option::TYPE_INT, Option::LEVEL_BASIC)
6167     .set_default(300)
6168     .set_description("Initialization timeout")
6169     .set_long_description(
6170         "The time length (in seconds) that RGW will allow for its initialization. RGW "
6171         "process will give up and quit if initialization is not complete after this amount "
6172         "of time."),
6173
6174     Option("rgw_mime_types_file", Option::TYPE_STR, Option::LEVEL_BASIC)
6175     .set_default("/etc/mime.types")
6176     .set_description("Path to local mime types file")
6177     .set_long_description(
6178         "The mime types file is needed in Swift when uploading an object. If object's "
6179         "content type is not specified, RGW will use data from this file to assign "
6180         "a content type to the object."),
6181
6182     Option("rgw_gc_max_objs", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6183     .set_default(32)
6184     .set_description("Number of shards for garbage collector data")
6185     .set_long_description(
6186         "The number of garbage collector data shards, is the number of RADOS objects that "
6187         "RGW will use to store the garbage collection information on.")
6188     .add_see_also({"rgw_gc_obj_min_wait", "rgw_gc_processor_max_time", "rgw_gc_processor_period", "rgw_gc_max_concurrent_io"}),
6189
6190     Option("rgw_gc_obj_min_wait", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6191     .set_default(2_hr)
6192     .set_description("Garbage collection object expiration time")
6193     .set_long_description(
6194        "The length of time (in seconds) that the RGW collector will wait before purging "
6195        "a deleted object's data. RGW will not remove object immediately, as object could "
6196        "still have readers. A mechanism exists to increase the object's expiration time "
6197        "when it's being read. The recommended value of its lower limit is 30 minutes")
6198     .add_see_also({"rgw_gc_max_objs", "rgw_gc_processor_max_time", "rgw_gc_processor_period", "rgw_gc_max_concurrent_io"}),
6199
6200     Option("rgw_gc_processor_max_time", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6201     .set_default(1_hr)
6202     .set_description("Length of time GC processor can lease shard")
6203     .set_long_description(
6204         "Garbage collection thread in RGW process holds a lease on its data shards. These "
6205         "objects contain the information about the objects that need to be removed. RGW "
6206         "takes a lease in order to prevent multiple RGW processes from handling the same "
6207         "objects concurrently. This time signifies that maximum amount of time (in seconds) that RGW "
6208         "is allowed to hold that lease. In the case where RGW goes down uncleanly, this "
6209         "is the amount of time where processing of that data shard will be blocked.")
6210     .add_see_also({"rgw_gc_max_objs", "rgw_gc_obj_min_wait", "rgw_gc_processor_period", "rgw_gc_max_concurrent_io"}),
6211
6212     Option("rgw_gc_processor_period", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6213     .set_default(1_hr)
6214     .set_description("Garbage collector cycle run time")
6215     .set_long_description(
6216         "The amount of time between the start of consecutive runs of the garbage collector "
6217         "threads. If garbage collector runs takes more than this period, it will not wait "
6218         "before running again.")
6219     .add_see_also({"rgw_gc_max_objs", "rgw_gc_obj_min_wait", "rgw_gc_processor_max_time", "rgw_gc_max_concurrent_io", "rgw_gc_max_trim_chunk"}),
6220
6221     Option("rgw_gc_max_concurrent_io", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6222     .set_default(10)
6223     .set_description("Max concurrent RADOS IO operations for garbage collection")
6224     .set_long_description(
6225         "The maximum number of concurrent IO operations that the RGW garbage collection "
6226         "thread will use when purging old data.")
6227     .add_see_also({"rgw_gc_max_objs", "rgw_gc_obj_min_wait", "rgw_gc_processor_max_time", "rgw_gc_max_trim_chunk"}),
6228
6229     Option("rgw_gc_max_trim_chunk", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6230     .set_default(16)
6231     .set_description("Max number of keys to remove from garbage collector log in a single operation")
6232     .add_see_also({"rgw_gc_max_objs", "rgw_gc_obj_min_wait", "rgw_gc_processor_max_time", "rgw_gc_max_concurrent_io"}),
6233
6234     Option("rgw_gc_max_deferred_entries_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
6235     .set_default(3072)
6236     .set_description("maximum allowed size of deferred entries in queue head for gc"),
6237
6238     Option("rgw_gc_max_queue_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
6239     .set_default(134213632)
6240     .set_description("Maximum allowed queue size for gc")
6241     .set_long_description(
6242         "The maximum allowed size of each gc queue, and its value should not "
6243         "be greater than (osd_max_object_size - rgw_gc_max_deferred_entries_size - 1K).")
6244     .add_see_also({"osd_max_object_size", "rgw_gc_max_deferred_entries_size"}),
6245
6246     Option("rgw_gc_max_deferred", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
6247     .set_default(50)
6248     .set_description("Number of maximum deferred data entries to be stored in queue for gc"),
6249
6250     Option("rgw_s3_success_create_obj_status", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6251     .set_default(0)
6252     .set_description("HTTP return code override for object creation")
6253     .set_long_description(
6254         "If not zero, this is the HTTP return code that will be returned on a successful S3 "
6255         "object creation."),
6256
6257     Option("rgw_resolve_cname", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6258     .set_default(false)
6259     .set_description("Support vanity domain names via CNAME")
6260     .set_long_description(
6261         "If true, RGW will query DNS when detecting that it's serving a request that was "
6262         "sent to a host in another domain. If a CNAME record is configured for that domain "
6263         "it will use it instead. This gives user to have the ability of creating a unique "
6264         "domain of their own to point at data in their bucket."),
6265
6266     Option("rgw_obj_stripe_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
6267     .set_default(4_M)
6268     .set_description("RGW object stripe size")
6269     .set_long_description(
6270         "The size of an object stripe for RGW objects. This is the maximum size a backing "
6271         "RADOS object will have. RGW objects that are larger than this will span over "
6272         "multiple objects."),
6273
6274     Option("rgw_extended_http_attrs", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6275     .set_default("")
6276     .set_description("RGW support extended HTTP attrs")
6277     .set_long_description(
6278         "Add new set of attributes that could be set on an object. These extra attributes "
6279         "can be set through HTTP header fields when putting the objects. If set, these "
6280         "attributes will return as HTTP fields when doing GET/HEAD on the object."),
6281
6282     Option("rgw_exit_timeout_secs", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6283     .set_default(120)
6284     .set_description("RGW shutdown timeout")
6285     .set_long_description("Number of seconds to wait for a process before exiting unconditionally."),
6286
6287     Option("rgw_get_obj_window_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
6288     .set_default(16_M)
6289     .set_description("RGW object read window size")
6290     .set_long_description("The window size in bytes for a single object read request"),
6291
6292     Option("rgw_get_obj_max_req_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
6293     .set_default(4_M)
6294     .set_description("RGW object read chunk size")
6295     .set_long_description(
6296         "The maximum request size of a single object read operation sent to RADOS"),
6297
6298     Option("rgw_relaxed_s3_bucket_names", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6299     .set_default(false)
6300     .set_description("RGW enable relaxed S3 bucket names")
6301     .set_long_description("RGW enable relaxed S3 bucket name rules for US region buckets."),
6302
6303     Option("rgw_defer_to_bucket_acls", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6304     .set_default("")
6305     .set_description("Bucket ACLs override object ACLs")
6306     .set_long_description(
6307         "If not empty, a string that selects that mode of operation. 'recurse' will use "
6308         "bucket's ACL for the authorizaton. 'full-control' will allow users that users "
6309         "that have full control permission on the bucket have access to the object."),
6310
6311     Option("rgw_list_buckets_max_chunk", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6312     .set_default(1000)
6313     .set_description("Max number of buckets to retrieve in a single listing operation")
6314     .set_long_description(
6315         "When RGW fetches lists of user's buckets from the backend, this is the max number "
6316         "of entries it will try to retrieve in a single operation. Note that the backend "
6317         "may choose to return a smaller number of entries."),
6318
6319     Option("rgw_md_log_max_shards", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6320     .set_default(64)
6321     .set_description("RGW number of metadata log shards")
6322     .set_long_description(
6323         "The number of shards the RGW metadata log entries will reside in. This affects "
6324         "the metadata sync parallelism as a shard can only be processed by a single "
6325         "RGW at a time"),
6326
6327     Option("rgw_curl_wait_timeout_ms", Option::TYPE_INT, Option::LEVEL_DEV)
6328     .set_default(1000)
6329     .set_description(""),
6330
6331     Option("rgw_curl_low_speed_limit", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6332     .set_default(1024)
6333     .set_long_description(
6334         "It contains the average transfer speed in bytes per second that the "
6335         "transfer should be below during rgw_curl_low_speed_time seconds for libcurl "
6336         "to consider it to be too slow and abort. Set it zero to disable this."),
6337
6338     Option("rgw_curl_low_speed_time", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6339     .set_default(300)
6340     .set_long_description(
6341         "It contains the time in number seconds that the transfer speed should be below "
6342         "the rgw_curl_low_speed_limit for the library to consider it too slow and abort. "
6343         "Set it zero to disable this."),
6344
6345     Option("rgw_copy_obj_progress", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6346     .set_default(true)
6347     .set_description("Send progress report through copy operation")
6348     .set_long_description(
6349         "If true, RGW will send progress information when copy operation is executed. "),
6350
6351     Option("rgw_copy_obj_progress_every_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
6352     .set_default(1_M)
6353     .set_description("Send copy-object progress info after these many bytes"),
6354
6355     Option("rgw_obj_tombstone_cache_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6356     .set_default(1000)
6357     .set_description("Max number of entries to keep in tombstone cache")
6358     .set_long_description(
6359         "The tombstone cache is used when doing a multi-zone data sync. RGW keeps "
6360         "there information about removed objects which is needed in order to prevent "
6361         "re-syncing of objects that were already removed."),
6362
6363     Option("rgw_data_log_window", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6364     .set_default(30)
6365     .set_description("Data log time window")
6366     .set_long_description(
6367         "The data log keeps information about buckets that have objectst that were "
6368         "modified within a specific timeframe. The sync process then knows which buckets "
6369         "are needed to be scanned for data sync."),
6370
6371     Option("rgw_data_log_changes_size", Option::TYPE_INT, Option::LEVEL_DEV)
6372     .set_default(1000)
6373     .set_description("Max size of pending changes in data log")
6374     .set_long_description(
6375         "RGW will trigger update to the data log if the number of pending entries reached "
6376         "this number."),
6377
6378     Option("rgw_data_log_num_shards", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6379     .set_default(128)
6380     .set_description("Number of data log shards")
6381     .set_long_description(
6382         "The number of shards the RGW data log entries will reside in. This affects the "
6383         "data sync parallelism as a shard can only be processed by a single RGW at a time."),
6384
6385     Option("rgw_data_log_obj_prefix", Option::TYPE_STR, Option::LEVEL_DEV)
6386     .set_default("data_log")
6387     .set_description(""),
6388
6389     Option("rgw_bucket_quota_ttl", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6390     .set_default(600)
6391     .set_description("Bucket quota stats cache TTL")
6392     .set_long_description(
6393         "Length of time for bucket stats to be cached within RGW instance."),
6394
6395     Option("rgw_bucket_quota_soft_threshold", Option::TYPE_FLOAT, Option::LEVEL_BASIC)
6396     .set_default(0.95)
6397     .set_description("RGW quota soft threshold")
6398     .set_long_description(
6399         "Threshold from which RGW doesn't rely on cached info for quota "
6400         "decisions. This is done for higher accuracy of the quota mechanism at "
6401         "cost of performance, when getting close to the quota limit. The value "
6402         "configured here is the ratio between the data usage to the max usage "
6403         "as specified by the quota."),
6404
6405     Option("rgw_bucket_quota_cache_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6406     .set_default(10000)
6407     .set_description("RGW quota stats cache size")
6408     .set_long_description(
6409         "Maximum number of entries in the quota stats cache."),
6410
6411     Option("rgw_bucket_default_quota_max_objects", Option::TYPE_INT, Option::LEVEL_BASIC)
6412     .set_default(-1)
6413     .set_description("Default quota for max objects in a bucket")
6414     .set_long_description(
6415         "The default quota configuration for max number of objects in a bucket. A "
6416         "negative number means 'unlimited'."),
6417
6418     Option("rgw_bucket_default_quota_max_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6419     .set_default(-1)
6420     .set_description("Default quota for total size in a bucket")
6421     .set_long_description(
6422         "The default quota configuration for total size of objects in a bucket. A "
6423         "negative number means 'unlimited'."),
6424
6425     Option("rgw_expose_bucket", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6426     .set_default(false)
6427     .set_description("Send Bucket HTTP header with the response")
6428     .set_long_description(
6429         "If true, RGW will send a Bucket HTTP header with the responses. The header will "
6430         "contain the name of the bucket the operation happened on."),
6431
6432     Option("rgw_frontends", Option::TYPE_STR, Option::LEVEL_BASIC)
6433     .set_default("beast port=7480")
6434     .set_description("RGW frontends configuration")
6435     .set_long_description(
6436         "A comma delimited list of frontends configuration. Each configuration contains "
6437         "the type of the frontend followed by an optional space delimited set of "
6438         "key=value config parameters."),
6439
6440     Option("rgw_frontend_defaults", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6441     .set_default("beast ssl_certificate=config://rgw/cert/$realm/$zone.crt ssl_private_key=config://rgw/cert/$realm/$zone.key")
6442     .set_description("RGW frontends default configuration")
6443     .set_long_description(
6444         "A comma delimited list of default frontends configuration."),
6445
6446     Option("rgw_user_quota_bucket_sync_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6447     .set_default(180)
6448     .set_description("User quota bucket sync interval")
6449     .set_long_description(
6450         "Time period for accumulating modified buckets before syncing these stats."),
6451
6452     Option("rgw_user_quota_sync_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6453     .set_default(1_day)
6454     .set_description("User quota sync interval")
6455     .set_long_description(
6456         "Time period for accumulating modified buckets before syncing entire user stats."),
6457
6458     Option("rgw_user_quota_sync_idle_users", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6459     .set_default(false)
6460     .set_description("Should sync idle users quota")
6461     .set_long_description(
6462         "Whether stats for idle users be fully synced."),
6463
6464     Option("rgw_user_quota_sync_wait_time", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6465     .set_default(1_day)
6466     .set_description("User quota full-sync wait time")
6467     .set_long_description(
6468         "Minimum time between two full stats sync for non-idle users."),
6469
6470     Option("rgw_user_default_quota_max_objects", Option::TYPE_INT, Option::LEVEL_BASIC)
6471     .set_default(-1)
6472     .set_description("User quota max objects")
6473     .set_long_description(
6474         "The default quota configuration for total number of objects for a single user. A "
6475         "negative number means 'unlimited'."),
6476
6477     Option("rgw_user_default_quota_max_size", Option::TYPE_INT, Option::LEVEL_BASIC)
6478     .set_default(-1)
6479     .set_description("User quota max size")
6480     .set_long_description(
6481         "The default quota configuration for total size of objects for a single user. A "
6482         "negative number means 'unlimited'."),
6483
6484     Option("rgw_multipart_min_part_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
6485     .set_default(5_M)
6486     .set_description("Minimum S3 multipart-upload part size")
6487     .set_long_description(
6488         "When doing a multipart upload, each part (other than the last part) should be "
6489         "at least this size."),
6490
6491     Option("rgw_multipart_part_upload_limit", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6492     .set_default(10000)
6493     .set_description("Max number of parts in multipart upload"),
6494
6495     Option("rgw_max_slo_entries", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6496     .set_default(1000)
6497     .set_description("Max number of entries in Swift Static Large Object manifest"),
6498
6499     Option("rgw_olh_pending_timeout_sec", Option::TYPE_INT, Option::LEVEL_DEV)
6500     .set_default(1_hr)
6501     .set_description("Max time for pending OLH change to complete")
6502     .set_long_description(
6503         "OLH is a versioned object's logical head. Operations on it are journaled and "
6504         "as pending before completion. If an operation doesn't complete with this amount "
6505         "of seconds, we remove the operation from the journal."),
6506
6507     Option("rgw_user_max_buckets", Option::TYPE_INT, Option::LEVEL_BASIC)
6508     .set_default(1000)
6509     .set_description("Max number of buckets per user")
6510     .set_long_description(
6511       "A user can create at most this number of buckets. Zero means "
6512       "no limit; a negative value means users cannot create any new "
6513       "buckets, although users will retain buckets already created."),
6514
6515     Option("rgw_objexp_gc_interval", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
6516     .set_default(10_min)
6517     .set_description("Swift objects expirer garbage collector interval"),
6518
6519     Option("rgw_objexp_hints_num_shards", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
6520     .set_default(127)
6521     .set_description("Number of object expirer data shards")
6522     .set_long_description(
6523         "The number of shards the (Swift) object expirer will store its data on."),
6524
6525     Option("rgw_objexp_chunk_size", Option::TYPE_UINT, Option::LEVEL_DEV)
6526     .set_default(100)
6527     .set_description(""),
6528
6529     Option("rgw_enable_static_website", Option::TYPE_BOOL, Option::LEVEL_BASIC)
6530     .set_default(false)
6531     .set_description("Enable static website APIs")
6532     .set_long_description(
6533         "This configurable controls whether RGW handles the website control APIs. RGW can "
6534         "server static websites if s3website hostnames are configured, and unrelated to "
6535         "this configurable."),
6536
6537      Option("rgw_user_unique_email", Option::TYPE_BOOL, Option::LEVEL_BASIC)
6538     .set_default(true)
6539     .set_description("Require local RGW users to have unique email addresses")
6540     .set_long_description(
6541         "Enforce builtin user accounts to have unique email addresses.  This "
6542         "setting is historical.  In future, non-enforcement of email address "
6543         "uniqueness is likely to become the default."),
6544
6545     Option("rgw_log_http_headers", Option::TYPE_STR, Option::LEVEL_BASIC)
6546     .set_default("")
6547     .set_description("List of HTTP headers to log")
6548     .set_long_description(
6549         "A comma delimited list of HTTP headers to log when seen, ignores case (e.g., "
6550         "http_x_forwarded_for)."),
6551
6552     Option("rgw_num_async_rados_threads", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6553     .set_default(32)
6554     .set_description("Number of concurrent RADOS operations in multisite sync")
6555     .set_long_description(
6556         "The number of concurrent RADOS IO operations that will be triggered for handling "
6557         "multisite sync operations. This includes control related work, and not the actual "
6558         "sync operations."),
6559
6560     Option("rgw_md_notify_interval_msec", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6561     .set_default(200)
6562     .set_description("Length of time to aggregate metadata changes")
6563     .set_long_description(
6564         "Length of time (in milliseconds) in which the master zone aggregates all the "
6565         "metadata changes that occurred, before sending notifications to all the other "
6566         "zones."),
6567
6568     Option("rgw_run_sync_thread", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6569     .set_default(true)
6570     .set_description("Should run sync thread"),
6571
6572     Option("rgw_sync_lease_period", Option::TYPE_INT, Option::LEVEL_DEV)
6573     .set_default(120)
6574     .set_description(""),
6575
6576     Option("rgw_sync_log_trim_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6577     .set_default(1200)
6578     .set_description("Sync log trim interval")
6579     .set_long_description(
6580         "Time in seconds between attempts to trim sync logs."),
6581
6582     Option("rgw_sync_log_trim_max_buckets", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6583     .set_default(16)
6584     .set_description("Maximum number of buckets to trim per interval")
6585     .set_long_description("The maximum number of buckets to consider for bucket index log trimming each trim interval, regardless of the number of bucket index shards. Priority is given to buckets with the most sync activity over the last trim interval.")
6586     .add_see_also("rgw_sync_log_trim_interval")
6587     .add_see_also("rgw_sync_log_trim_min_cold_buckets")
6588     .add_see_also("rgw_sync_log_trim_concurrent_buckets"),
6589
6590     Option("rgw_sync_log_trim_min_cold_buckets", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6591     .set_default(4)
6592     .set_description("Minimum number of cold buckets to trim per interval")
6593     .set_long_description("Of the `rgw_sync_log_trim_max_buckets` selected for bucket index log trimming each trim interval, at least this many of them must be 'cold' buckets. These buckets are selected in order from the list of all bucket instances, to guarantee that all buckets will be visited eventually.")
6594     .add_see_also("rgw_sync_log_trim_interval")
6595     .add_see_also("rgw_sync_log_trim_max_buckets")
6596     .add_see_also("rgw_sync_log_trim_concurrent_buckets"),
6597
6598     Option("rgw_sync_log_trim_concurrent_buckets", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6599     .set_default(4)
6600     .set_description("Maximum number of buckets to trim in parallel")
6601     .add_see_also("rgw_sync_log_trim_interval")
6602     .add_see_also("rgw_sync_log_trim_max_buckets")
6603     .add_see_also("rgw_sync_log_trim_min_cold_buckets"),
6604
6605     Option("rgw_sync_data_inject_err_probability", Option::TYPE_FLOAT, Option::LEVEL_DEV)
6606     .set_default(0)
6607     .set_description(""),
6608
6609     Option("rgw_sync_meta_inject_err_probability", Option::TYPE_FLOAT, Option::LEVEL_DEV)
6610     .set_default(0)
6611     .set_description(""),
6612
6613     Option("rgw_sync_trace_history_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
6614     .set_default(4096)
6615     .set_description("Sync trace history size")
6616     .set_long_description(
6617       "Maximum number of complete sync trace entries to keep."),
6618
6619     Option("rgw_sync_trace_per_node_log_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6620     .set_default(32)
6621     .set_description("Sync trace per-node log size")
6622     .set_long_description(
6623         "The number of log entries to keep per sync-trace node."),
6624
6625     Option("rgw_sync_trace_servicemap_update_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6626     .set_default(10)
6627     .set_description("Sync-trace service-map update interval")
6628     .set_long_description(
6629         "Number of seconds between service-map updates of sync-trace events."),
6630
6631     Option("rgw_period_push_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
6632     .set_default(2)
6633     .set_description("Period push interval")
6634     .set_long_description(
6635         "Number of seconds to wait before retrying 'period push' operation."),
6636
6637     Option("rgw_period_push_interval_max", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
6638     .set_default(30)
6639     .set_description("Period push maximum interval")
6640     .set_long_description(
6641         "The max number of seconds to wait before retrying 'period push' after exponential "
6642         "backoff."),
6643
6644     Option("rgw_safe_max_objects_per_shard", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6645     .set_default(100*1024)
6646     .set_description("Safe number of objects per shard")
6647     .set_long_description(
6648         "This is the max number of objects per bucket index shard that RGW considers "
6649         "safe. RGW will warn if it identifies a bucket where its per-shard count is "
6650         "higher than a percentage of this number.")
6651     .add_see_also("rgw_shard_warning_threshold"),
6652
6653     Option("rgw_shard_warning_threshold", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
6654     .set_default(90)
6655     .set_description("Warn about max objects per shard")
6656     .set_long_description(
6657         "Warn if number of objects per shard in a specific bucket passed this percentage "
6658         "of the safe number.")
6659     .add_see_also("rgw_safe_max_objects_per_shard"),
6660
6661     Option("rgw_swift_versioning_enabled", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6662     .set_default(false)
6663     .set_description("Enable Swift versioning"),
6664
6665     Option("rgw_swift_custom_header", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6666     .set_default("")
6667     .set_description("Enable swift custom header")
6668     .set_long_description(
6669         "If not empty, specifies a name of HTTP header that can include custom data. When "
6670         "uploading an object, if this header is passed RGW will store this header info "
6671         "and it will be available when listing the bucket."),
6672
6673     Option("rgw_swift_need_stats", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6674     .set_default(true)
6675     .set_description("Enable stats on bucket listing in Swift"),
6676
6677     Option("rgw_reshard_num_logs", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
6678     .set_default(16)
6679     .set_min(1)
6680     .set_description("")
6681     .add_service("rgw"),
6682
6683     Option("rgw_reshard_bucket_lock_duration", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
6684     .set_default(360)
6685     .set_min(30)
6686     .set_description("Number of seconds the timeout on the reshard locks (bucket reshard lock and reshard log lock) are set to. As a reshard proceeds these locks can be renewed/extended. If too short, reshards cannot complete and will fail, causing a future reshard attempt. If too long a hung or crashed reshard attempt will keep the bucket locked for an extended period, not allowing RGW to detect the failed reshard attempt and recover.")
6687     .add_tag("performance")
6688     .add_service("rgw"),
6689
6690     Option("rgw_reshard_batch_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
6691     .set_default(64)
6692     .set_min(8)
6693     .set_description("Number of reshard entries to batch together before sending the operations to the CLS back-end")
6694     .add_tag("performance")
6695     .add_service("rgw"),
6696
6697     Option("rgw_reshard_max_aio", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
6698     .set_default(128)
6699     .set_min(16)
6700     .set_description("Maximum number of outstanding asynchronous I/O operations to allow at a time during resharding")
6701     .add_tag("performance")
6702     .add_service("rgw"),
6703
6704     Option("rgw_trust_forwarded_https", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6705     .set_default(false)
6706     .set_description("Trust Forwarded and X-Forwarded-Proto headers")
6707     .set_long_description(
6708         "When a proxy in front of radosgw is used for ssl termination, radosgw "
6709         "does not know whether incoming http connections are secure. Enable "
6710         "this option to trust the Forwarded and X-Forwarded-Proto headers sent "
6711         "by the proxy when determining whether the connection is secure. This "
6712         "is required for some features, such as server side encryption.")
6713     .add_see_also("rgw_crypt_require_ssl"),
6714
6715     Option("rgw_crypt_require_ssl", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6716     .set_default(true)
6717     .set_description("Requests including encryption key headers must be sent over ssl"),
6718
6719     Option("rgw_crypt_default_encryption_key", Option::TYPE_STR, Option::LEVEL_DEV)
6720     .set_default("")
6721     .set_description(""),
6722
6723     Option("rgw_crypt_s3_kms_backend", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6724     .set_default("barbican")
6725     .set_enum_allowed({"barbican", "vault", "testing"})
6726     .set_description(
6727         "Where the SSE-KMS encryption keys are stored. Supported KMS "
6728         "systems are OpenStack Barbican ('barbican', the default) and HashiCorp "
6729         "Vault ('vault')."),
6730
6731     Option("rgw_crypt_s3_kms_encryption_keys", Option::TYPE_STR, Option::LEVEL_DEV)
6732     .set_default("")
6733     .set_description(""),
6734
6735     Option("rgw_crypt_vault_auth", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6736     .set_default("token")
6737     .set_enum_allowed({"token", "agent"})
6738     .set_description(
6739         "Type of authentication method to be used with Vault. ")
6740     .add_see_also({
6741         "rgw_crypt_s3_kms_backend",
6742         "rgw_crypt_vault_addr",
6743         "rgw_crypt_vault_token_file"}),
6744
6745     Option("rgw_crypt_vault_token_file", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6746     .set_default("")
6747     .set_description(
6748         "If authentication method is 'token', provide a path to the token file, "
6749         "which for security reasons should readable only by Rados Gateway.")
6750     .add_see_also({
6751       "rgw_crypt_s3_kms_backend",
6752       "rgw_crypt_vault_auth",
6753       "rgw_crypt_vault_addr"}),
6754
6755     Option("rgw_crypt_vault_addr", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6756     .set_default("")
6757     .set_description("Vault server base address.")
6758     .add_see_also({
6759       "rgw_crypt_s3_kms_backend",
6760       "rgw_crypt_vault_auth",
6761       "rgw_crypt_vault_prefix"}),
6762
6763     Option("rgw_crypt_vault_prefix", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6764     .set_default("")
6765     .set_description("Vault secret URL prefix, which can be used to restrict "
6766                      "access to a particular subset of the Vault secret space.")
6767     .add_see_also({
6768       "rgw_crypt_s3_kms_backend",
6769       "rgw_crypt_vault_addr",
6770       "rgw_crypt_vault_auth"}),
6771
6772
6773     Option("rgw_crypt_vault_secret_engine", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6774     .set_enum_allowed({"kv", "transit"})
6775     .set_default("transit")
6776     .set_description(
6777         "Vault Secret Engine to be used to retrieve encryption keys.")
6778     .add_see_also({
6779       "rgw_crypt_s3_kms_backend",
6780       "rgw_crypt_vault_auth",
6781       "rgw_crypt_vault_addr"}),
6782
6783     Option("rgw_crypt_vault_namespace", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6784     .set_default("")
6785     .set_description("Vault Namespace to be used to select your tenant")
6786     .add_see_also({
6787       "rgw_crypt_s3_kms_backend",
6788       "rgw_crypt_vault_auth",
6789       "rgw_crypt_vault_addr"}),
6790
6791     Option("rgw_crypt_suppress_logs", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6792     .set_default(true)
6793     .set_description("Suppress logs that might print client key"),
6794
6795     Option("rgw_list_bucket_min_readahead", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6796     .set_default(1000)
6797     .set_description("Minimum number of entries to request from rados for bucket listing"),
6798
6799     Option("rgw_rest_getusage_op_compat", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6800     .set_default(false)
6801     .set_description("REST GetUsage request backward compatibility"),
6802
6803     Option("rgw_torrent_flag", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6804     .set_default(false)
6805     .set_description("When true, uploaded objects will calculate and store "
6806                      "a SHA256 hash of object data so the object can be "
6807                      "retrieved as a torrent file"),
6808
6809     Option("rgw_torrent_tracker", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6810     .set_default("")
6811     .set_description("Torrent field announce and announce list"),
6812
6813     Option("rgw_torrent_createby", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6814     .set_default("")
6815     .set_description("torrent field created by"),
6816
6817     Option("rgw_torrent_comment", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6818     .set_default("")
6819     .set_description("Torrent field comment"),
6820
6821     Option("rgw_torrent_encoding", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6822     .set_default("")
6823     .set_description("torrent field encoding"),
6824
6825     Option("rgw_data_notify_interval_msec", Option::TYPE_INT, Option::LEVEL_ADVANCED)
6826     .set_default(200)
6827     .set_description("data changes notification interval to followers"),
6828
6829     Option("rgw_torrent_origin", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6830     .set_default("")
6831     .set_description("Torrent origin"),
6832
6833     Option("rgw_torrent_sha_unit", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
6834     .set_default(512*1024)
6835     .set_description(""),
6836
6837     Option("rgw_dynamic_resharding", Option::TYPE_BOOL, Option::LEVEL_BASIC)
6838     .set_default(true)
6839     .set_description("Enable dynamic resharding")
6840     .set_long_description(
6841         "If true, RGW will dynamicall increase the number of shards in buckets that have "
6842         "a high number of objects per shard.")
6843     .add_see_also("rgw_max_objs_per_shard")
6844     .add_see_also("rgw_max_dynamic_shards"),
6845
6846     Option("rgw_max_objs_per_shard", Option::TYPE_UINT, Option::LEVEL_BASIC)
6847     .set_default(100000)
6848     .set_description("Max objects per shard for dynamic resharding")
6849     .set_long_description(
6850         "This is the max number of objects per bucket index shard that RGW will "
6851         "allow with dynamic resharding. RGW will trigger an automatic reshard operation "
6852         "on the bucket if it exceeds this number.")
6853     .add_see_also("rgw_dynamic_resharding")
6854     .add_see_also("rgw_max_dynamic_shards"),
6855
6856     Option("rgw_max_dynamic_shards", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
6857     .set_default(1999)
6858     .set_min(1)
6859     .set_description("Max shards that dynamic resharding can create")
6860     .set_long_description(
6861         "This is the maximum number of bucket index shards that dynamic "
6862         "sharding is able to create on its own. This does not limit user "
6863         "requested resharding. Ideally this value is a prime number.")
6864     .add_see_also("rgw_dynamic_resharding")
6865     .add_see_also("rgw_max_objs_per_shard"),
6866
6867     Option("rgw_reshard_thread_interval", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
6868     .set_default(10_min)
6869     .set_min(10)
6870     .set_description("Number of seconds between processing of reshard log entries"),
6871
6872     Option("rgw_cache_expiry_interval", Option::TYPE_UINT,
6873            Option::LEVEL_ADVANCED)
6874     .set_default(15_min)
6875     .set_description("Number of seconds before entries in the cache are "
6876                      "assumed stale and re-fetched. Zero is never.")
6877     .add_tag("performance")
6878     .add_service("rgw")
6879     .set_long_description("The Rados Gateway stores metadata and objects in "
6880                           "an internal cache. This should be kept consistent "
6881                           "by the OSD's relaying notify events between "
6882                           "multiple watching RGW processes. In the event "
6883                           "that this notification protocol fails, bounding "
6884                           "the length of time that any data in the cache will "
6885                           "be assumed valid will ensure that any RGW instance "
6886                           "that falls out of sync will eventually recover. "
6887                           "This seems to be an issue mostly for large numbers "
6888                           "of RGW instances under heavy use. If you would like "
6889                           "to turn off cache expiry, set this value to zero."),
6890
6891     Option("rgw_inject_notify_timeout_probability", Option::TYPE_FLOAT,
6892            Option::LEVEL_DEV)
6893     .set_default(0)
6894     .add_tag("fault injection")
6895     .add_tag("testing")
6896     .add_service("rgw")
6897     .set_min_max(0.0, 1.0)
6898     .set_description("Likelihood of ignoring a notify")
6899     .set_long_description("This is the probability that the RGW cache will "
6900                           "ignore a cache notify message. It exists to help "
6901                           "with the development and testing of cache "
6902                           "consistency and recovery improvements. Please "
6903                           "do not set it in a production cluster, as it "
6904                           "actively causes failures. Set this to a floating "
6905                           "point value between 0 and 1."),
6906     Option("rgw_max_notify_retries", Option::TYPE_UINT,
6907            Option::LEVEL_ADVANCED)
6908     .set_default(3)
6909     .add_tag("error recovery")
6910     .add_service("rgw")
6911     .set_description("Number of attempts to notify peers before giving up.")
6912     .set_long_description("The number of times we will attempt to update "
6913                           "a peer's cache in the event of error before giving "
6914                           "up. This is unlikely to be an issue unless your "
6915                           "cluster is very heavily loaded. Beware that "
6916                           "increasing this value may cause some operations to "
6917                           "take longer in exceptional cases and thus may, "
6918                           "rarely, cause clients to time out."),
6919     Option("rgw_sts_entry", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6920     .set_default("sts")
6921     .set_description("STS URL prefix")
6922     .set_long_description("URL path prefix for internal STS requests."),
6923
6924     Option("rgw_sts_key", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6925     .set_default("sts")
6926     .set_description("STS Key")
6927     .set_long_description("Key used for encrypting/ decrypting session token."),
6928
6929     Option("rgw_s3_auth_use_sts", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
6930     .set_default(false)
6931     .set_description("Should S3 authentication use STS."),
6932
6933     Option("rgw_sts_max_session_duration", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
6934     .set_default(43200)
6935     .set_description("Session token max duration")
6936     .set_long_description("Max duration in seconds for which the session token is valid."),
6937
6938     Option("rgw_max_listing_results", Option::TYPE_UINT,
6939            Option::LEVEL_ADVANCED)
6940     .set_default(1000)
6941     .set_min_max(1, 100000)
6942     .add_service("rgw")
6943     .set_description("Upper bound on results in listing operations, ListBucket max-keys")
6944     .set_long_description("This caps the maximum permitted value for listing-like operations in RGW S3. "
6945                           "Affects ListBucket(max-keys), "
6946                           "ListBucketVersions(max-keys), "
6947                           "ListBucketMultipartUploads(max-uploads), "
6948                           "ListMultipartUploadParts(max-parts)"),
6949
6950     Option("rgw_sts_token_introspection_url", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6951     .set_default("")
6952     .set_description("STS Web Token introspection URL")
6953     .set_long_description("URL for introspecting an STS Web Token."),
6954
6955     Option("rgw_sts_client_id", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6956     .set_default("")
6957     .set_description("Client Id")
6958     .set_long_description("Client Id needed for introspecting a Web Token."),
6959
6960     Option("rgw_sts_client_secret", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6961     .set_default("")
6962     .set_description("Client Secret")
6963     .set_long_description("Client Secret needed for introspecting a Web Token."),
6964
6965     Option("rgw_max_concurrent_requests", Option::TYPE_INT, Option::LEVEL_BASIC)
6966     .set_default(1024)
6967     .set_description("Maximum number of concurrent HTTP requests.")
6968     .set_long_description(
6969         "Maximum number of concurrent HTTP requests that the beast frontend "
6970         "will process. Tuning this can help to limit memory usage under heavy "
6971         "load.")
6972     .add_tag("performance")
6973     .add_see_also("rgw_frontends"),
6974
6975     Option("rgw_scheduler_type", Option::TYPE_STR, Option::LEVEL_ADVANCED)
6976     .set_default("throttler")
6977     .set_description("Set the type of dmclock scheduler, defaults to throttler "
6978                      "Other valid values are dmclock which is experimental"),
6979
6980     Option("rgw_dmclock_admin_res", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
6981     .set_default(100.0)
6982     .set_description("mclock reservation for admin requests")
6983     .add_see_also("rgw_dmclock_admin_wgt")
6984     .add_see_also("rgw_dmclock_admin_lim"),
6985
6986     Option("rgw_dmclock_admin_wgt", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
6987     .set_default(100.0)
6988     .set_description("mclock weight for admin requests")
6989     .add_see_also("rgw_dmclock_admin_res")
6990     .add_see_also("rgw_dmclock_admin_lim"),
6991
6992     Option("rgw_dmclock_admin_lim", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
6993     .set_default(0.0)
6994     .set_description("mclock limit for admin requests")
6995     .add_see_also("rgw_dmclock_admin_res")
6996     .add_see_also("rgw_dmclock_admin_wgt"),
6997
6998     Option("rgw_dmclock_auth_res", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
6999     .set_default(200.0)
7000     .set_description("mclock reservation for object data requests")
7001     .add_see_also("rgw_dmclock_auth_wgt")
7002     .add_see_also("rgw_dmclock_auth_lim"),
7003
7004     Option("rgw_dmclock_auth_wgt", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7005     .set_default(100.0)
7006     .set_description("mclock weight for object data requests")
7007     .add_see_also("rgw_dmclock_auth_res")
7008     .add_see_also("rgw_dmclock_auth_lim"),
7009
7010     Option("rgw_dmclock_auth_lim", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7011     .set_default(0.0)
7012     .set_description("mclock limit for object data requests")
7013     .add_see_also("rgw_dmclock_auth_res")
7014     .add_see_also("rgw_dmclock_auth_wgt"),
7015
7016     Option("rgw_dmclock_data_res", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7017     .set_default(500.0)
7018     .set_description("mclock reservation for object data requests")
7019     .add_see_also("rgw_dmclock_data_wgt")
7020     .add_see_also("rgw_dmclock_data_lim"),
7021
7022     Option("rgw_dmclock_data_wgt", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7023     .set_default(500.0)
7024     .set_description("mclock weight for object data requests")
7025     .add_see_also("rgw_dmclock_data_res")
7026     .add_see_also("rgw_dmclock_data_lim"),
7027
7028     Option("rgw_dmclock_data_lim", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7029     .set_default(0.0)
7030     .set_description("mclock limit for object data requests")
7031     .add_see_also("rgw_dmclock_data_res")
7032     .add_see_also("rgw_dmclock_data_wgt"),
7033
7034     Option("rgw_dmclock_metadata_res", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7035     .set_default(500.0)
7036     .set_description("mclock reservation for metadata requests")
7037     .add_see_also("rgw_dmclock_metadata_wgt")
7038     .add_see_also("rgw_dmclock_metadata_lim"),
7039
7040     Option("rgw_dmclock_metadata_wgt", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7041     .set_default(500.0)
7042     .set_description("mclock weight for metadata requests")
7043     .add_see_also("rgw_dmclock_metadata_res")
7044     .add_see_also("rgw_dmclock_metadata_lim"),
7045
7046     Option("rgw_dmclock_metadata_lim", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7047     .set_default(0.0)
7048     .set_description("mclock limit for metadata requests")
7049     .add_see_also("rgw_dmclock_metadata_res")
7050     .add_see_also("rgw_dmclock_metadata_wgt"),
7051   });
7052 }
7053
7054 static std::vector<Option> get_rbd_options() {
7055   return std::vector<Option>({
7056     Option("rbd_default_pool", Option::TYPE_STR, Option::LEVEL_ADVANCED)
7057     .set_default("rbd")
7058     .set_description("default pool for storing new images")
7059     .set_validator([](std::string *value, std::string *error_message){
7060       std::regex pattern("^[^@/]+$");
7061       if (!std::regex_match (*value, pattern)) {
7062         *value = "rbd";
7063         *error_message = "invalid RBD default pool, resetting to 'rbd'";
7064       }
7065       return 0;
7066     }),
7067
7068     Option("rbd_default_data_pool", Option::TYPE_STR, Option::LEVEL_ADVANCED)
7069     .set_default("")
7070     .set_description("default pool for storing data blocks for new images")
7071     .set_validator([](std::string *value, std::string *error_message){
7072       std::regex pattern("^[^@/]*$");
7073       if (!std::regex_match (*value, pattern)) {
7074         *value = "";
7075         *error_message = "ignoring invalid RBD data pool";
7076       }
7077       return 0;
7078     }),
7079
7080     Option("rbd_default_features", Option::TYPE_STR, Option::LEVEL_ADVANCED)
7081     .set_default("layering,exclusive-lock,object-map,fast-diff,deep-flatten")
7082     .set_description("default v2 image features for new images")
7083     .set_long_description(
7084         "RBD features are only applicable for v2 images. This setting accepts "
7085         "either an integer bitmask value or comma-delimited string of RBD "
7086         "feature names. This setting is always internally stored as an integer "
7087         "bitmask value. The mapping between feature bitmask value and feature "
7088         "name is as follows: +1 -> layering, +2 -> striping, "
7089         "+4 -> exclusive-lock, +8 -> object-map, +16 -> fast-diff, "
7090         "+32 -> deep-flatten, +64 -> journaling, +128 -> data-pool")
7091     .set_flag(Option::FLAG_RUNTIME)
7092     .set_validator([](std::string *value, std::string *error_message) {
7093         ostringstream ss;
7094         uint64_t features = librbd::rbd_features_from_string(*value, &ss);
7095         // Leave this in integer form to avoid breaking Cinder.  Someday
7096         // we would like to present this in string form instead...
7097         *value = stringify(features);
7098         if (ss.str().size()) {
7099           return -EINVAL;
7100         }
7101         return 0;
7102       }),
7103
7104     Option("rbd_op_threads", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7105     .set_default(1)
7106     .set_description("number of threads to utilize for internal processing"),
7107
7108     Option("rbd_op_thread_timeout", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7109     .set_default(60)
7110     .set_description("time in seconds for detecting a hung thread"),
7111
7112     Option("rbd_disable_zero_copy_writes", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7113     .set_default(true)
7114     .set_description("Disable the use of zero-copy writes to ensure unstable "
7115                      "writes from clients cannot cause a CRC mismatch"),
7116
7117     Option("rbd_non_blocking_aio", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7118     .set_default(true)
7119     .set_description("process AIO ops from a dispatch thread to prevent blocking"),
7120
7121     Option("rbd_cache", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7122     .set_default(true)
7123     .set_description("whether to enable caching (writeback unless rbd_cache_max_dirty is 0)"),
7124
7125     Option("rbd_cache_policy", Option::TYPE_STR, Option::LEVEL_ADVANCED)
7126     .set_enum_allowed({"writethrough", "writeback", "writearound"})
7127     .set_default("writearound")
7128     .set_description("cache policy for handling writes."),
7129
7130     Option("rbd_cache_writethrough_until_flush", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7131     .set_default(true)
7132     .set_description("whether to make writeback caching writethrough until "
7133                      "flush is called, to be sure the user of librbd will send "
7134                      "flushes so that writeback is safe"),
7135
7136     Option("rbd_cache_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7137     .set_default(32_M)
7138     .set_description("cache size in bytes"),
7139
7140     Option("rbd_cache_max_dirty", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7141     .set_default(24_M)
7142     .set_description("dirty limit in bytes - set to 0 for write-through caching"),
7143
7144     Option("rbd_cache_target_dirty", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7145     .set_default(16_M)
7146     .set_description("target dirty limit in bytes"),
7147
7148     Option("rbd_cache_max_dirty_age", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7149     .set_default(1.0)
7150     .set_description("seconds in cache before writeback starts"),
7151
7152     Option("rbd_cache_max_dirty_object", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7153     .set_default(0)
7154     .set_description("dirty limit for objects - set to 0 for auto calculate from rbd_cache_size"),
7155
7156     Option("rbd_cache_block_writes_upfront", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7157     .set_default(false)
7158     .set_description("whether to block writes to the cache before the aio_write call completes"),
7159
7160     Option("rbd_parent_cache_enabled", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7161     .set_default(false)
7162     .set_description("whether to enable rbd shared ro cache"),
7163
7164     Option("rbd_concurrent_management_ops", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7165     .set_default(10)
7166     .set_min(1)
7167     .set_description("how many operations can be in flight for a management operation like deleting or resizing an image"),
7168
7169     Option("rbd_balance_snap_reads", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7170     .set_default(false)
7171     .set_description("distribute snap read requests to random OSD"),
7172
7173     Option("rbd_localize_snap_reads", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7174     .set_default(false)
7175     .set_description("localize snap read requests to closest OSD"),
7176
7177     Option("rbd_balance_parent_reads", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7178     .set_default(false)
7179     .set_description("distribute parent read requests to random OSD"),
7180
7181     Option("rbd_localize_parent_reads", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7182     .set_default(false)
7183     .set_description("localize parent requests to closest OSD"),
7184
7185     Option("rbd_sparse_read_threshold_bytes", Option::TYPE_SIZE,
7186            Option::LEVEL_ADVANCED)
7187     .set_default(64_K)
7188     .set_description("threshold for issuing a sparse-read")
7189     .set_long_description("minimum number of sequential bytes to read against "
7190                           "an object before issuing a sparse-read request to "
7191                           "the cluster. 0 implies it must be a full object read "
7192                           "to issue a sparse-read, 1 implies always use "
7193                           "sparse-read, and any value larger than the maximum "
7194                           "object size will disable sparse-read for all "
7195                           "requests"),
7196
7197     Option("rbd_readahead_trigger_requests", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7198     .set_default(10)
7199     .set_description("number of sequential requests necessary to trigger readahead"),
7200
7201     Option("rbd_readahead_max_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7202     .set_default(512_K)
7203     .set_description("set to 0 to disable readahead"),
7204
7205     Option("rbd_readahead_disable_after_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7206     .set_default(50_M)
7207     .set_description("how many bytes are read in total before readahead is disabled"),
7208
7209     Option("rbd_clone_copy_on_read", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7210     .set_default(false)
7211     .set_description("copy-up parent image blocks to clone upon read request"),
7212
7213     Option("rbd_blacklist_on_break_lock", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7214     .set_default(true)
7215     .set_description("whether to blacklist clients whose lock was broken"),
7216
7217     Option("rbd_blacklist_expire_seconds", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7218     .set_default(0)
7219     .set_description("number of seconds to blacklist - set to 0 for OSD default"),
7220
7221     Option("rbd_request_timed_out_seconds", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7222     .set_default(30)
7223     .set_description("number of seconds before maintenance request times out"),
7224
7225     Option("rbd_skip_partial_discard", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7226     .set_default(true)
7227     .set_description("skip discard (zero) of unaligned extents within an object"),
7228
7229     Option("rbd_discard_granularity_bytes", Option::TYPE_UINT,
7230            Option::LEVEL_ADVANCED)
7231     .set_default(64_K)
7232     .set_min_max(4_K, 32_M)
7233     .set_validator([](std::string *value, std::string *error_message){
7234         uint64_t f = strict_si_cast<uint64_t>(value->c_str(), error_message);
7235         if (!error_message->empty()) {
7236           return -EINVAL;
7237         } else if (!isp2(f)) {
7238           *error_message = "value must be a power of two";
7239           return -EINVAL;
7240         }
7241         return 0;
7242       })
7243     .set_description("minimum aligned size of discard operations"),
7244
7245     Option("rbd_enable_alloc_hint", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7246     .set_default(true)
7247     .set_description("when writing a object, it will issue a hint to osd backend to indicate the expected size object need"),
7248
7249     Option("rbd_compression_hint", Option::TYPE_STR, Option::LEVEL_BASIC)
7250     .set_enum_allowed({"none", "compressible", "incompressible"})
7251     .set_default("none")
7252     .set_description("Compression hint to send to the OSDs during writes")
7253     .set_flag(Option::FLAG_RUNTIME),
7254
7255     Option("rbd_read_from_replica_policy", Option::TYPE_STR, Option::LEVEL_BASIC)
7256     .set_enum_allowed({"default", "balance", "localize"})
7257     .set_default("default")
7258     .set_description("Read replica policy send to the OSDS during reads")
7259     .set_flag(Option::FLAG_RUNTIME),
7260
7261     Option("rbd_tracing", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7262     .set_default(false)
7263     .set_description("true if LTTng-UST tracepoints should be enabled"),
7264
7265     Option("rbd_blkin_trace_all", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7266     .set_default(false)
7267     .set_description("create a blkin trace for all RBD requests"),
7268
7269     Option("rbd_validate_pool", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7270     .set_default(true)
7271     .set_description("validate empty pools for RBD compatibility"),
7272
7273     Option("rbd_validate_names", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7274     .set_default(true)
7275     .set_description("validate new image names for RBD compatibility"),
7276
7277     Option("rbd_auto_exclusive_lock_until_manual_request", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7278     .set_default(true)
7279     .set_description("automatically acquire/release exclusive lock until it is explicitly requested"),
7280
7281     Option("rbd_move_to_trash_on_remove", Option::TYPE_BOOL, Option::LEVEL_BASIC)
7282     .set_default(false)
7283     .set_description("automatically move images to the trash when deleted"),
7284
7285     Option("rbd_move_to_trash_on_remove_expire_seconds", Option::TYPE_UINT, Option::LEVEL_BASIC)
7286     .set_default(0)
7287     .set_description("default number of seconds to protect deleted images in the trash"),
7288
7289     Option("rbd_move_parent_to_trash_on_remove", Option::TYPE_BOOL, Option::LEVEL_BASIC)
7290     .set_default(false)
7291     .set_description("move parent with clone format v2 children to the trash when deleted"),
7292
7293     Option("rbd_mirroring_resync_after_disconnect", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7294     .set_default(false)
7295     .set_description("automatically start image resync after mirroring is disconnected due to being laggy"),
7296
7297     Option("rbd_mirroring_delete_delay", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7298     .set_default(0)
7299     .set_description("time-delay in seconds for rbd-mirror delete propagation"),
7300
7301     Option("rbd_mirroring_replay_delay", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7302     .set_default(0)
7303     .set_description("time-delay in seconds for rbd-mirror asynchronous replication"),
7304
7305     Option("rbd_mirroring_max_mirroring_snapshots", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7306     .set_default(3)
7307     .set_min(3)
7308     .set_description("mirroring snapshots limit"),
7309
7310     Option("rbd_default_format", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7311     .set_default(2)
7312     .set_description("default image format for new images"),
7313
7314     Option("rbd_default_order", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7315     .set_default(22)
7316     .set_description("default order (data block object size) for new images"),
7317
7318     Option("rbd_default_stripe_count", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7319     .set_default(0)
7320     .set_description("default stripe count for new images"),
7321
7322     Option("rbd_default_stripe_unit", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7323     .set_default(0)
7324     .set_description("default stripe width for new images"),
7325
7326     Option("rbd_default_map_options", Option::TYPE_STR, Option::LEVEL_ADVANCED)
7327     .set_default("")
7328     .set_description("default krbd map options"),
7329
7330     Option("rbd_default_clone_format", Option::TYPE_STR, Option::LEVEL_ADVANCED)
7331     .set_enum_allowed({"1", "2", "auto"})
7332     .set_default("auto")
7333     .set_description("default internal format for handling clones")
7334     .set_long_description("This sets the internal format for tracking cloned "
7335                           "images. The setting of '1' requires attaching to "
7336                           "protected snapshots that cannot be removed until "
7337                           "the clone is removed/flattened. The setting of '2' "
7338                           "will allow clones to be attached to any snapshot "
7339                           "and permits removing in-use parent snapshots but "
7340                           "requires Mimic or later clients. The default "
7341                           "setting of 'auto' will use the v2 format if the "
7342                           "cluster is configured to require mimic or later "
7343                           "clients.")
7344     .set_flag(Option::FLAG_RUNTIME),
7345
7346     Option("rbd_journal_order", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7347     .set_min_max(12, 26)
7348     .set_default(24)
7349     .set_description("default order (object size) for journal data objects"),
7350
7351     Option("rbd_journal_splay_width", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7352     .set_default(4)
7353     .set_description("number of active journal objects"),
7354
7355     Option("rbd_journal_commit_age", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7356     .set_default(5)
7357     .set_description("commit time interval, seconds"),
7358
7359     Option("rbd_journal_object_writethrough_until_flush", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7360     .set_default(true)
7361     .set_description("when enabled, the rbd_journal_object_flush* configuration "
7362                      "options are ignored until the first flush so that batched "
7363                      "journal IO is known to be safe for consistency"),
7364
7365     Option("rbd_journal_object_flush_interval", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7366     .set_default(0)
7367     .set_description("maximum number of pending commits per journal object"),
7368
7369     Option("rbd_journal_object_flush_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7370     .set_default(1_M)
7371     .set_description("maximum number of pending bytes per journal object"),
7372
7373     Option("rbd_journal_object_flush_age", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7374     .set_default(0)
7375     .set_description("maximum age (in seconds) for pending commits"),
7376
7377     Option("rbd_journal_object_max_in_flight_appends", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7378     .set_default(0)
7379     .set_description("maximum number of in-flight appends per journal object"),
7380
7381     Option("rbd_journal_pool", Option::TYPE_STR, Option::LEVEL_ADVANCED)
7382     .set_default("")
7383     .set_description("pool for journal objects"),
7384
7385     Option("rbd_journal_max_payload_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7386     .set_default(16384)
7387     .set_description("maximum journal payload size before splitting"),
7388
7389     Option("rbd_journal_max_concurrent_object_sets", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7390     .set_default(0)
7391     .set_description("maximum number of object sets a journal client can be behind before it is automatically unregistered"),
7392
7393     Option("rbd_qos_iops_limit", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7394     .set_default(0)
7395     .set_description("the desired limit of IO operations per second"),
7396
7397     Option("rbd_qos_bps_limit", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7398     .set_default(0)
7399     .set_description("the desired limit of IO bytes per second"),
7400
7401     Option("rbd_qos_read_iops_limit", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7402     .set_default(0)
7403     .set_description("the desired limit of read operations per second"),
7404
7405     Option("rbd_qos_write_iops_limit", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7406     .set_default(0)
7407     .set_description("the desired limit of write operations per second"),
7408
7409     Option("rbd_qos_read_bps_limit", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7410     .set_default(0)
7411     .set_description("the desired limit of read bytes per second"),
7412
7413     Option("rbd_qos_write_bps_limit", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7414     .set_default(0)
7415     .set_description("the desired limit of write bytes per second"),
7416
7417     Option("rbd_qos_iops_burst", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7418     .set_default(0)
7419     .set_description("the desired burst limit of IO operations"),
7420
7421     Option("rbd_qos_bps_burst", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7422     .set_default(0)
7423     .set_description("the desired burst limit of IO bytes"),
7424
7425     Option("rbd_qos_read_iops_burst", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7426     .set_default(0)
7427     .set_description("the desired burst limit of read operations"),
7428
7429     Option("rbd_qos_write_iops_burst", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7430     .set_default(0)
7431     .set_description("the desired burst limit of write operations"),
7432
7433     Option("rbd_qos_read_bps_burst", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7434     .set_default(0)
7435     .set_description("the desired burst limit of read bytes"),
7436
7437     Option("rbd_qos_write_bps_burst", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7438     .set_default(0)
7439     .set_description("the desired burst limit of write bytes"),
7440
7441     Option("rbd_qos_schedule_tick_min", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7442     .set_default(50)
7443     .set_min(1)
7444     .set_description("minimum schedule tick (in milliseconds) for QoS"),
7445
7446     Option("rbd_discard_on_zeroed_write_same", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7447     .set_default(true)
7448     .set_description("discard data on zeroed write same instead of writing zero"),
7449
7450     Option("rbd_mtime_update_interval", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7451     .set_default(60)
7452     .set_min(0)
7453     .set_description("RBD Image modify timestamp refresh interval. Set to 0 to disable modify timestamp update."),
7454
7455     Option("rbd_atime_update_interval", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7456     .set_default(60)
7457     .set_min(0)
7458     .set_description("RBD Image access timestamp refresh interval. Set to 0 to disable access timestamp update."),
7459
7460     Option("rbd_io_scheduler", Option::TYPE_STR, Option::LEVEL_ADVANCED)
7461     .set_default("simple")
7462     .set_enum_allowed({"none", "simple"})
7463     .set_description("RBD IO scheduler"),
7464
7465     Option("rbd_io_scheduler_simple_max_delay", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7466     .set_default(0)
7467     .set_min(0)
7468     .set_description("maximum io delay (in milliseconds) for simple io scheduler (if set to 0 dalay is calculated based on latency stats)"),
7469
7470     Option("rbd_rwl_enabled", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7471     .set_default(false)
7472     .set_description("enable persistent write back cache for this volume"),
7473
7474     Option("rbd_rwl_log_periodic_stats", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7475     .set_default(false)
7476     .set_description("emit periodic perf stats to debug log"),
7477
7478     Option("rbd_rwl_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7479     .set_default(1073741824)
7480     .set_min(1073741824)
7481     .set_description("size of the persistent write back cache for this volume"),
7482
7483     Option("rbd_rwl_path", Option::TYPE_STR, Option::LEVEL_ADVANCED)
7484     .set_default("/tmp")
7485     .set_description("location of the persistent write back cache in a DAX-enabled filesystem on persistent memory"),
7486
7487     Option("rbd_config_pool_override_update_timestamp", Option::TYPE_UINT,
7488            Option::LEVEL_DEV)
7489     .set_default(0)
7490     .set_description("timestamp of last update to pool-level config overrides"),
7491
7492   });
7493 }
7494
7495 static std::vector<Option> get_rbd_mirror_options() {
7496   return std::vector<Option>({
7497     Option("rbd_mirror_journal_commit_age", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7498     .set_default(5)
7499     .set_description("commit time interval, seconds"),
7500
7501     Option("rbd_mirror_journal_poll_age", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7502     .set_default(5)
7503     .set_description("maximum age (in seconds) between successive journal polls"),
7504
7505     Option("rbd_mirror_sync_point_update_age", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7506     .set_default(30)
7507     .set_description("number of seconds between each update of the image sync point object number"),
7508
7509     Option("rbd_mirror_concurrent_image_syncs", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7510     .set_default(5)
7511     .set_description("maximum number of image syncs in parallel"),
7512
7513     Option("rbd_mirror_pool_replayers_refresh_interval", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7514     .set_default(30)
7515     .set_description("interval to refresh peers in rbd-mirror daemon"),
7516
7517     Option("rbd_mirror_concurrent_image_deletions", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7518     .set_default(1)
7519     .set_min(1)
7520     .set_description("maximum number of image deletions in parallel"),
7521
7522     Option("rbd_mirror_delete_retry_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7523     .set_default(30)
7524     .set_description("interval to check and retry the failed deletion requests"),
7525
7526     Option("rbd_mirror_image_state_check_interval", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7527     .set_default(30)
7528     .set_min(1)
7529     .set_description("interval to get images from pool watcher and set sources in replayer"),
7530
7531     Option("rbd_mirror_leader_heartbeat_interval", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7532     .set_default(5)
7533     .set_min(1)
7534     .set_description("interval (in seconds) between mirror leader heartbeats"),
7535
7536     Option("rbd_mirror_leader_max_missed_heartbeats", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7537     .set_default(2)
7538     .set_description("number of missed heartbeats for non-lock owner to attempt to acquire lock"),
7539
7540     Option("rbd_mirror_leader_max_acquire_attempts_before_break", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7541     .set_default(3)
7542     .set_description("number of failed attempts to acquire lock after missing heartbeats before breaking lock"),
7543
7544     Option("rbd_mirror_image_policy_type", Option::TYPE_STR, Option::LEVEL_ADVANCED)
7545     .set_default("simple")
7546     .set_enum_allowed({"none", "simple"})
7547     .set_description("active/active policy type for mapping images to instances"),
7548
7549     Option("rbd_mirror_image_policy_migration_throttle", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7550     .set_default(300)
7551     .set_description("number of seconds after which an image can be reshuffled (migrated) again"),
7552
7553     Option("rbd_mirror_image_policy_update_throttle_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7554     .set_default(1)
7555     .set_min(1)
7556     .set_description("interval (in seconds) to throttle images for mirror daemon peer updates"),
7557
7558     Option("rbd_mirror_image_policy_rebalance_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7559     .set_default(0)
7560     .set_description("number of seconds policy should be idle before trigerring reshuffle (rebalance) of images"),
7561
7562     Option("rbd_mirror_perf_stats_prio", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7563     .set_default((int64_t)PerfCountersBuilder::PRIO_USEFUL)
7564     .set_description("Priority level for mirror daemon replication perf counters")
7565     .set_long_description("The daemon will send perf counter data to the "
7566                           "manager daemon if the priority is not lower than "
7567                           "mgr_stats_threshold.")
7568     .set_min_max((int64_t)PerfCountersBuilder::PRIO_DEBUGONLY,
7569                  (int64_t)PerfCountersBuilder::PRIO_CRITICAL + 1),
7570
7571     Option("rbd_mirror_image_perf_stats_prio", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7572     .set_default((int64_t)PerfCountersBuilder::PRIO_USEFUL)
7573     .set_description("Priority level for mirror daemon per-image replication perf counters")
7574     .set_long_description("The daemon will send per-image perf counter data to the "
7575                           "manager daemon if the priority is not lower than "
7576                           "mgr_stats_threshold.")
7577     .set_min_max((int64_t)PerfCountersBuilder::PRIO_DEBUGONLY,
7578                  (int64_t)PerfCountersBuilder::PRIO_CRITICAL + 1),
7579
7580     Option("rbd_mirror_memory_autotune", Option::TYPE_BOOL, Option::LEVEL_DEV)
7581     .set_default(true)
7582     .add_see_also("rbd_mirror_memory_target")
7583     .set_description("Automatically tune the ratio of caches while respecting min values."),
7584
7585     Option("rbd_mirror_memory_target", Option::TYPE_SIZE, Option::LEVEL_BASIC)
7586     .set_default(4_G)
7587     .add_see_also("rbd_mirror_memory_autotune")
7588     .set_description("When tcmalloc and cache autotuning is enabled, try to keep this many bytes mapped in memory."),
7589
7590     Option("rbd_mirror_memory_base", Option::TYPE_SIZE, Option::LEVEL_DEV)
7591     .set_default(768_M)
7592     .add_see_also("rbd_mirror_memory_autotune")
7593     .set_description("When tcmalloc and cache autotuning is enabled, estimate the minimum amount of memory in bytes the rbd-mirror daemon will need."),
7594
7595     Option("rbd_mirror_memory_expected_fragmentation", Option::TYPE_FLOAT, Option::LEVEL_DEV)
7596     .set_default(0.15)
7597     .set_min_max(0.0, 1.0)
7598     .add_see_also("rbd_mirror_memory_autotune")
7599     .set_description("When tcmalloc and cache autotuning is enabled, estimate the percent of memory fragmentation."),
7600
7601     Option("rbd_mirror_memory_cache_min", Option::TYPE_SIZE, Option::LEVEL_DEV)
7602     .set_default(128_M)
7603     .add_see_also("rbd_mirror_memory_autotune")
7604     .set_description("When tcmalloc and cache autotuning is enabled, set the minimum amount of memory used for cache."),
7605
7606     Option("rbd_mirror_memory_cache_resize_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
7607     .set_default(5)
7608     .add_see_also("rbd_mirror_memory_autotune")
7609     .set_description("When tcmalloc and cache autotuning is enabled, wait this many seconds between resizing caches."),
7610
7611     Option("rbd_mirror_memory_cache_autotune_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
7612     .set_default(30)
7613     .add_see_also("rbd_mirror_memory_autotune")
7614     .set_description("The number of seconds to wait between rebalances when cache autotune is enabled."),
7615   });
7616 }
7617
7618 static std::vector<Option> get_immutable_object_cache_options() {
7619   return std::vector<Option>({
7620     Option("immutable_object_cache_path", Option::TYPE_STR, Option::LEVEL_ADVANCED)
7621     .set_default("/tmp/ceph_immutable_object_cache")
7622     .set_description("immutable object cache data dir"),
7623
7624     Option("immutable_object_cache_sock", Option::TYPE_STR, Option::LEVEL_ADVANCED)
7625     .set_default("/var/run/ceph/immutable_object_cache_sock")
7626     .set_description("immutable object cache domain socket"),
7627
7628     Option("immutable_object_cache_max_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7629     .set_default(1_G)
7630     .set_description("max immutable object cache data size"),
7631
7632     Option("immutable_object_cache_max_inflight_ops", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7633     .set_default(128)
7634     .set_description("max inflight promoting requests for immutable object cache daemon"),
7635
7636     Option("immutable_object_cache_client_dedicated_thread_num", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7637     .set_default(2)
7638     .set_description("immutable object cache client dedicated thread number"),
7639
7640     Option("immutable_object_cache_watermark", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7641     .set_default(0.1)
7642     .set_description("immutable object cache water mark"),
7643   });
7644 }
7645
7646 std::vector<Option> get_mds_options() {
7647   return std::vector<Option>({
7648     Option("mds_numa_node", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7649     .set_default(-1)
7650     .set_flag(Option::FLAG_STARTUP)
7651     .set_description("set mds's cpu affinity to a numa node (-1 for none)"),
7652
7653     Option("mds_data", Option::TYPE_STR, Option::LEVEL_ADVANCED)
7654     .set_default("/var/lib/ceph/mds/$cluster-$id")
7655     .set_flag(Option::FLAG_NO_MON_UPDATE)
7656     .set_description("path to MDS data and keyring"),
7657
7658     Option("mds_join_fs", Option::TYPE_STR, Option::LEVEL_BASIC)
7659     .set_default("")
7660     .set_description("file system MDS prefers to join")
7661     .set_long_description("This setting indicates which file system name the MDS should prefer to join (affinity). The monitors will try to have the MDS cluster safely reach a state where all MDS have strong affinity, even via failovers to a standby.")
7662     .set_flag(Option::FLAG_RUNTIME),
7663
7664     Option("mds_max_xattr_pairs_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7665     .set_default(64_K)
7666     .set_description("maximum aggregate size of extended attributes on a file"),
7667
7668     Option("mds_cache_trim_interval", Option::TYPE_SECS, Option::LEVEL_ADVANCED)
7669     .set_default(1)
7670     .set_description("interval in seconds between cache trimming")
7671     .set_flag(Option::FLAG_RUNTIME),
7672
7673     Option("mds_cache_release_free_interval", Option::TYPE_SECS, Option::LEVEL_DEV)
7674     .set_default(10)
7675     .set_description("interval in seconds between heap releases")
7676     .set_flag(Option::FLAG_RUNTIME),
7677
7678     Option("mds_cache_memory_limit", Option::TYPE_SIZE, Option::LEVEL_BASIC)
7679     .set_default(4_G)
7680     .set_description("target maximum memory usage of MDS cache")
7681     .set_flag(Option::FLAG_RUNTIME)
7682     .set_long_description("This sets a target maximum memory usage of the MDS cache and is the primary tunable to limit the MDS memory usage. The MDS will try to stay under a reservation of this limit (by default 95%; 1 - mds_cache_reservation) by trimming unused metadata in its cache and recalling cached items in the client caches. It is possible for the MDS to exceed this limit due to slow recall from clients. The mds_health_cache_threshold (150%) sets a cache full threshold for when the MDS signals a cluster health warning."),
7683
7684     Option("mds_cache_reservation", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7685     .set_default(.05)
7686     .set_description("amount of memory to reserve for future cached objects")
7687     .set_flag(Option::FLAG_RUNTIME),
7688
7689     Option("mds_health_cache_threshold", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7690     .set_default(1.5)
7691     .set_description("threshold for cache size to generate health warning"),
7692
7693     Option("mds_cache_mid", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7694     .set_default(.7)
7695     .set_description("midpoint for MDS cache LRU"),
7696
7697     Option("mds_cache_trim_decay_rate", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7698     .set_default(1)
7699     .set_description("decay rate for trimming MDS cache throttle")
7700     .set_flag(Option::FLAG_RUNTIME),
7701
7702     Option("mds_cache_trim_threshold", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7703     .set_default(64_K)
7704     .set_description("threshold for number of dentries that can be trimmed")
7705     .set_flag(Option::FLAG_RUNTIME),
7706
7707     Option("mds_max_file_recover", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7708     .set_default(32)
7709     .set_description("maximum number of files to recover file sizes in parallel"),
7710
7711     Option("mds_dir_max_commit_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7712     .set_default(10)
7713     .set_description("maximum size in megabytes for a RADOS write to a directory"),
7714
7715     Option("mds_dir_keys_per_op", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7716     .set_default(16384)
7717     .set_description("number of directory entries to read in one RADOS operation"),
7718
7719     Option("mds_decay_halflife", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7720     .set_default(5)
7721     .set_description("rate of decay for temperature counters on each directory for balancing"),
7722
7723     Option("mds_beacon_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7724     .set_default(4)
7725     .set_description("interval in seconds between MDS beacons to monitors"),
7726
7727     Option("mds_beacon_grace", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7728     .set_default(15)
7729     .set_description("tolerance in seconds for missed MDS beacons to monitors"),
7730
7731     Option("mds_heartbeat_grace", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7732     .set_default(15)
7733     .set_description("tolerance in seconds for MDS internal heartbeat"),
7734
7735     Option("mds_enforce_unique_name", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7736     .set_default(true)
7737     .set_description("require MDS name is unique in the cluster"),
7738
7739     Option("mds_session_blacklist_on_timeout", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7740     .set_default(true)
7741     .set_description("blacklist clients whose sessions have become stale"),
7742
7743     Option("mds_session_blacklist_on_evict", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7744     .set_default(true)
7745     .set_description("blacklist clients that have been evicted"),
7746
7747     Option("mds_sessionmap_keys_per_op", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7748     .set_default(1024)
7749     .set_description("number of omap keys to read from the SessionMap in one operation"),
7750
7751     Option("mds_recall_max_caps", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7752     .set_default(5000)
7753     .set_description("maximum number of caps to recall from client session in single recall")
7754     .set_flag(Option::FLAG_RUNTIME),
7755
7756     Option("mds_recall_max_decay_rate", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7757     .set_default(2.5)
7758     .set_description("decay rate for throttle on recalled caps on a session")
7759     .set_flag(Option::FLAG_RUNTIME),
7760
7761     Option("mds_recall_max_decay_threshold", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7762     .set_default(16_K)
7763     .set_description("decay threshold for throttle on recalled caps on a session")
7764     .set_flag(Option::FLAG_RUNTIME),
7765
7766     Option("mds_recall_global_max_decay_threshold", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7767     .set_default(64_K)
7768     .set_description("decay threshold for throttle on recalled caps globally")
7769     .set_flag(Option::FLAG_RUNTIME),
7770
7771     Option("mds_recall_warning_threshold", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7772     .set_default(32_K)
7773     .set_description("decay threshold for warning on slow session cap recall")
7774     .set_flag(Option::FLAG_RUNTIME),
7775
7776     Option("mds_recall_warning_decay_rate", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7777     .set_default(60.0)
7778     .set_description("decay rate for warning on slow session cap recall")
7779     .set_flag(Option::FLAG_RUNTIME),
7780
7781     Option("mds_session_cache_liveness_decay_rate", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7782     .add_see_also("mds_session_cache_liveness_magnitude")
7783     .set_default(5_min)
7784     .set_description("decay rate for session liveness leading to preemptive cap recall")
7785     .set_flag(Option::FLAG_RUNTIME)
7786     .set_long_description("This determines how long a session needs to be quiescent before the MDS begins preemptively recalling capabilities. The default of 5 minutes will cause 10 halvings of the decay counter after 1 hour, or 1/1024. The default magnitude of 10 (1^10 or 1024) is chosen so that the MDS considers a previously chatty session (approximately) to be quiescent after 1 hour."),
7787
7788     Option("mds_session_cache_liveness_magnitude", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7789     .add_see_also("mds_session_cache_liveness_decay_rate")
7790     .set_default(10)
7791     .set_description("decay magnitude for preemptively recalling caps on quiet client")
7792     .set_flag(Option::FLAG_RUNTIME)
7793     .set_long_description("This is the order of magnitude difference (in base 2) of the internal liveness decay counter and the number of capabilities the session holds. When this difference occurs, the MDS treats the session as quiescent and begins recalling capabilities."),
7794
7795     Option("mds_freeze_tree_timeout", Option::TYPE_FLOAT, Option::LEVEL_DEV)
7796     .set_default(30)
7797     .set_description(""),
7798
7799     Option("mds_health_summarize_threshold", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7800     .set_default(10)
7801     .set_description("threshold of number of clients to summarize late client recall"),
7802
7803     Option("mds_reconnect_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7804     .set_default(45)
7805     .set_description("timeout in seconds to wait for clients to reconnect during MDS reconnect recovery state"),
7806
7807     Option("mds_tick_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7808     .set_default(5)
7809     .set_description("time in seconds between upkeep tasks"),
7810
7811     Option("mds_dirstat_min_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
7812     .set_default(1)
7813     .set_description(""),
7814
7815     Option("mds_scatter_nudge_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7816     .set_default(5)
7817     .set_description("minimum interval between scatter lock updates"),
7818
7819     Option("mds_client_prealloc_inos", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7820     .set_default(1000)
7821     .set_description("number of unused inodes to pre-allocate to clients for file creation"),
7822
7823     Option("mds_client_delegate_inos_pct", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7824     .set_default(50)
7825     .set_flag(Option::FLAG_RUNTIME)
7826     .set_description("percentage of preallocated inos to delegate to client"),
7827
7828     Option("mds_early_reply", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7829     .set_default(true)
7830     .set_description("additional reply to clients that metadata requests are complete but not yet durable"),
7831
7832     Option("mds_replay_unsafe_with_closed_session", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7833     .set_default(false)
7834     .set_flag(Option::FLAG_RUNTIME)
7835     .set_description("complete all the replay request when mds is restarted, no matter the session is closed or not"),
7836
7837     Option("mds_default_dir_hash", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7838     .set_default(CEPH_STR_HASH_RJENKINS)
7839     .set_description("hash function to select directory fragment for dentry name"),
7840
7841     Option("mds_log_pause", Option::TYPE_BOOL, Option::LEVEL_DEV)
7842     .set_default(false)
7843     .set_description(""),
7844
7845     Option("mds_log_skip_corrupt_events", Option::TYPE_BOOL, Option::LEVEL_DEV)
7846     .set_default(false)
7847     .set_description(""),
7848
7849     Option("mds_log_max_events", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7850     .set_default(-1)
7851     .set_description("maximum number of events in the MDS journal (-1 is unlimited)"),
7852
7853     Option("mds_log_events_per_segment", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7854     .set_default(1024)
7855     .set_description("maximum number of events in an MDS journal segment"),
7856
7857     Option("mds_log_segment_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
7858     .set_default(0)
7859     .set_description("size in bytes of each MDS log segment"),
7860
7861     Option("mds_log_max_segments", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
7862     .set_default(128)
7863     .set_description("maximum number of segments which may be untrimmed"),
7864
7865     Option("mds_bal_export_pin", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7866     .set_default(true)
7867     .set_description("allow setting directory export pins to particular ranks"),
7868
7869     Option("mds_export_ephemeral_random", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7870     .set_default(false)
7871     .set_flag(Option::FLAG_RUNTIME)
7872     .set_description("allow ephemeral random pinning of the loaded subtrees")
7873     .set_long_description("probabilistically pin the loaded directory inode and the subtree beneath it to an MDS based on the consistent hash of the inode number. The higher this value the more likely the loaded subtrees get pinned"),
7874
7875     Option("mds_export_ephemeral_random_max", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7876     .set_default(0.01)
7877     .set_flag(Option::FLAG_RUNTIME)
7878     .set_description("the maximum percent permitted for random ephemeral pin policy")
7879     .set_min_max(0.0, 1.0)
7880     .add_see_also("mds_export_ephemeral_random"),
7881
7882     Option("mds_export_ephemeral_distributed", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7883     .set_default(false)
7884     .set_flag(Option::FLAG_RUNTIME)
7885     .set_description("allow ephemeral distributed pinning of the loaded subtrees")
7886     .set_long_description("pin the immediate child directories of the loaded directory inode based on the consistent hash of the child's inode number. "),
7887
7888     Option("mds_bal_sample_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7889     .set_default(3.0)
7890     .set_description("interval in seconds between balancer ticks"),
7891
7892     Option("mds_bal_replicate_threshold", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7893     .set_default(8000)
7894     .set_description("hot popularity threshold to replicate a subtree"),
7895
7896     Option("mds_bal_unreplicate_threshold", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7897     .set_default(0)
7898     .set_description("cold popularity threshold to merge subtrees"),
7899
7900     Option("mds_bal_split_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7901     .set_default(10000)
7902     .set_description("minimum size of directory fragment before splitting"),
7903
7904     Option("mds_bal_split_rd", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7905     .set_default(25000)
7906     .set_description("hot read popularity threshold for splitting a directory fragment"),
7907
7908     Option("mds_bal_split_wr", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7909     .set_default(10000)
7910     .set_description("hot write popularity threshold for splitting a directory fragment"),
7911
7912     Option("mds_bal_split_bits", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7913     .set_default(3)
7914     .set_min_max(1, 24)
7915     .set_description("power of two child fragments for a fragment on split"),
7916
7917     Option("mds_bal_merge_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7918     .set_default(50)
7919     .set_description("size of fragments where merging should occur"),
7920
7921     Option("mds_bal_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7922     .set_default(10)
7923     .set_description("interval between MDS balancer cycles"),
7924
7925     Option("mds_bal_fragment_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7926     .set_default(5)
7927     .set_description("delay in seconds before interrupting client IO to perform splits"),
7928
7929     Option("mds_bal_fragment_size_max", Option::TYPE_INT, Option::LEVEL_ADVANCED)
7930     .set_default(10000*10)
7931     .set_description("maximum size of a directory fragment before new creat/links fail"),
7932
7933     Option("mds_bal_fragment_fast_factor", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7934     .set_default(1.5)
7935     .set_description("ratio of mds_bal_split_size at which fast fragment splitting occurs"),
7936
7937     Option("mds_bal_fragment_dirs", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
7938     .set_default(true)
7939     .set_description("enable directory fragmentation")
7940     .set_long_description("Directory fragmentation is a standard feature of CephFS that allows sharding directories across multiple objects for performance and stability. Additionally, this allows fragments to be distributed across multiple active MDSs to increase throughput. Disabling (new) fragmentation should only be done in exceptional circumstances and may lead to performance issues."),
7941
7942     Option("mds_bal_idle_threshold", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7943     .set_default(0)
7944     .set_description("idle metadata popularity threshold before rebalancing"),
7945
7946     Option("mds_bal_max", Option::TYPE_INT, Option::LEVEL_DEV)
7947     .set_default(-1)
7948     .set_description(""),
7949
7950     Option("mds_bal_max_until", Option::TYPE_INT, Option::LEVEL_DEV)
7951     .set_default(-1)
7952     .set_description(""),
7953
7954     Option("mds_bal_mode", Option::TYPE_INT, Option::LEVEL_DEV)
7955     .set_default(0)
7956     .set_description(""),
7957
7958     Option("mds_bal_min_rebalance", Option::TYPE_FLOAT, Option::LEVEL_DEV)
7959     .set_default(.1)
7960     .set_description("amount overloaded over internal target before balancer begins offloading"),
7961
7962     Option("mds_bal_min_start", Option::TYPE_FLOAT, Option::LEVEL_DEV)
7963     .set_default(.2)
7964     .set_description(""),
7965
7966     Option("mds_bal_need_min", Option::TYPE_FLOAT, Option::LEVEL_DEV)
7967     .set_default(.8)
7968     .set_description(""),
7969
7970     Option("mds_bal_need_max", Option::TYPE_FLOAT, Option::LEVEL_DEV)
7971     .set_default(1.2)
7972     .set_description(""),
7973
7974     Option("mds_bal_midchunk", Option::TYPE_FLOAT, Option::LEVEL_DEV)
7975     .set_default(.3)
7976     .set_description(""),
7977
7978     Option("mds_bal_minchunk", Option::TYPE_FLOAT, Option::LEVEL_DEV)
7979     .set_default(.001)
7980     .set_description(""),
7981
7982     Option("mds_bal_target_decay", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7983     .set_default(10.0)
7984     .set_description("rate of decay for export targets communicated to clients"),
7985
7986     Option("mds_replay_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
7987     .set_default(1.0)
7988     .set_description("time in seconds between replay of updates to journal by standby replay MDS"),
7989
7990     Option("mds_shutdown_check", Option::TYPE_INT, Option::LEVEL_DEV)
7991     .set_default(0)
7992     .set_description(""),
7993
7994     Option("mds_thrash_exports", Option::TYPE_INT, Option::LEVEL_DEV)
7995     .set_default(0)
7996     .set_description(""),
7997
7998     Option("mds_thrash_fragments", Option::TYPE_INT, Option::LEVEL_DEV)
7999     .set_default(0)
8000     .set_description(""),
8001
8002     Option("mds_dump_cache_on_map", Option::TYPE_BOOL, Option::LEVEL_DEV)
8003     .set_default(false)
8004     .set_description(""),
8005
8006     Option("mds_dump_cache_after_rejoin", Option::TYPE_BOOL, Option::LEVEL_DEV)
8007     .set_default(false)
8008     .set_description(""),
8009
8010     Option("mds_verify_scatter", Option::TYPE_BOOL, Option::LEVEL_DEV)
8011     .set_default(false)
8012     .set_description(""),
8013
8014     Option("mds_debug_scatterstat", Option::TYPE_BOOL, Option::LEVEL_DEV)
8015     .set_default(false)
8016     .set_description(""),
8017
8018     Option("mds_debug_frag", Option::TYPE_BOOL, Option::LEVEL_DEV)
8019     .set_default(false)
8020     .set_description(""),
8021
8022     Option("mds_debug_auth_pins", Option::TYPE_BOOL, Option::LEVEL_DEV)
8023     .set_default(false)
8024     .set_description(""),
8025
8026     Option("mds_debug_subtrees", Option::TYPE_BOOL, Option::LEVEL_DEV)
8027     .set_default(false)
8028     .set_description(""),
8029
8030     Option("mds_kill_mdstable_at", Option::TYPE_INT, Option::LEVEL_DEV)
8031     .set_default(0)
8032     .set_description(""),
8033
8034     Option("mds_max_export_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
8035     .set_default(20_M)
8036     .set_description(""),
8037
8038     Option("mds_kill_export_at", Option::TYPE_INT, Option::LEVEL_DEV)
8039     .set_default(0)
8040     .set_description(""),
8041
8042     Option("mds_kill_import_at", Option::TYPE_INT, Option::LEVEL_DEV)
8043     .set_default(0)
8044     .set_description(""),
8045
8046     Option("mds_kill_link_at", Option::TYPE_INT, Option::LEVEL_DEV)
8047     .set_default(0)
8048     .set_description(""),
8049
8050     Option("mds_kill_rename_at", Option::TYPE_INT, Option::LEVEL_DEV)
8051     .set_default(0)
8052     .set_description(""),
8053
8054     Option("mds_kill_openc_at", Option::TYPE_INT, Option::LEVEL_DEV)
8055     .set_default(0)
8056     .set_description(""),
8057
8058     Option("mds_kill_journal_at", Option::TYPE_INT, Option::LEVEL_DEV)
8059     .set_default(0)
8060     .set_description(""),
8061
8062     Option("mds_kill_journal_expire_at", Option::TYPE_INT, Option::LEVEL_DEV)
8063     .set_default(0)
8064     .set_description(""),
8065
8066     Option("mds_kill_journal_replay_at", Option::TYPE_INT, Option::LEVEL_DEV)
8067     .set_default(0)
8068     .set_description(""),
8069
8070     Option("mds_journal_format", Option::TYPE_UINT, Option::LEVEL_DEV)
8071     .set_default(1)
8072     .set_description(""),
8073
8074     Option("mds_kill_create_at", Option::TYPE_INT, Option::LEVEL_DEV)
8075     .set_default(0)
8076     .set_description(""),
8077
8078     Option("mds_inject_traceless_reply_probability", Option::TYPE_FLOAT, Option::LEVEL_DEV)
8079     .set_default(0)
8080     .set_description(""),
8081
8082     Option("mds_wipe_sessions", Option::TYPE_BOOL, Option::LEVEL_DEV)
8083     .set_default(0)
8084     .set_description(""),
8085
8086     Option("mds_wipe_ino_prealloc", Option::TYPE_BOOL, Option::LEVEL_DEV)
8087     .set_default(0)
8088     .set_description(""),
8089
8090     Option("mds_skip_ino", Option::TYPE_INT, Option::LEVEL_DEV)
8091     .set_default(0)
8092     .set_description(""),
8093
8094     Option("mds_enable_op_tracker", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8095     .set_default(true)
8096     .set_description("track remote operation progression and statistics"),
8097
8098     Option("mds_op_history_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
8099     .set_default(20)
8100     .set_description("maximum size for list of historical operations"),
8101
8102     Option("mds_op_history_duration", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
8103     .set_default(600)
8104     .set_description("expiration time in seconds of historical operations"),
8105
8106     Option("mds_op_complaint_time", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
8107     .set_default(30)
8108     .set_description("time in seconds to consider an operation blocked after no updates"),
8109
8110     Option("mds_op_log_threshold", Option::TYPE_INT, Option::LEVEL_DEV)
8111     .set_default(5)
8112     .set_description(""),
8113
8114     Option("mds_snap_min_uid", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
8115     .set_default(0)
8116     .set_description("minimum uid of client to perform snapshots"),
8117
8118     Option("mds_snap_max_uid", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
8119     .set_default(4294967294)
8120     .set_description("maximum uid of client to perform snapshots"),
8121
8122     Option("mds_snap_rstat", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8123     .set_default(false)
8124     .set_description("enabled nested rstat for snapshots"),
8125
8126     Option("mds_verify_backtrace", Option::TYPE_UINT, Option::LEVEL_DEV)
8127     .set_default(1)
8128     .set_description(""),
8129
8130     Option("mds_max_completed_flushes", Option::TYPE_UINT, Option::LEVEL_DEV)
8131     .set_default(100000)
8132     .set_description(""),
8133
8134     Option("mds_max_completed_requests", Option::TYPE_UINT, Option::LEVEL_DEV)
8135     .set_default(100000)
8136     .set_description(""),
8137
8138     Option("mds_action_on_write_error", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
8139     .set_default(1)
8140     .set_description("action to take when MDS cannot write to RADOS (0:ignore, 1:read-only, 2:suicide)"),
8141
8142     Option("mds_mon_shutdown_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
8143     .set_default(5)
8144     .set_description("time to wait for mon to receive damaged MDS rank notification"),
8145
8146     Option("mds_max_purge_files", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
8147     .set_default(64)
8148     .set_description("maximum number of deleted files to purge in parallel"),
8149
8150     Option("mds_max_purge_ops", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
8151     .set_default(8192)
8152     .set_description("maximum number of purge operations performed in parallel"),
8153
8154     Option("mds_max_purge_ops_per_pg", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
8155     .set_default(0.5)
8156     .set_description("number of parallel purge operations performed per PG"),
8157
8158     Option("mds_purge_queue_busy_flush_period", Option::TYPE_FLOAT, Option::LEVEL_DEV)
8159     .set_default(1.0)
8160     .set_description(""),
8161
8162     Option("mds_root_ino_uid", Option::TYPE_INT, Option::LEVEL_ADVANCED)
8163     .set_default(0)
8164     .set_description("default uid for new root directory"),
8165
8166     Option("mds_root_ino_gid", Option::TYPE_INT, Option::LEVEL_ADVANCED)
8167     .set_default(0)
8168     .set_description("default gid for new root directory"),
8169
8170     Option("mds_max_scrub_ops_in_progress", Option::TYPE_INT, Option::LEVEL_ADVANCED)
8171     .set_default(5)
8172     .set_description("maximum number of scrub operations performed in parallel"),
8173
8174     Option("mds_forward_all_requests_to_auth", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8175     .set_default(false)
8176     .set_flag(Option::FLAG_RUNTIME)
8177     .set_description("always process op on auth mds"),
8178
8179     Option("mds_damage_table_max_entries", Option::TYPE_INT, Option::LEVEL_ADVANCED)
8180     .set_default(10000)
8181     .set_description("maximum number of damage table entries"),
8182
8183     Option("mds_client_writeable_range_max_inc_objs", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
8184     .set_default(1024)
8185     .set_description("maximum number of objects in writeable range of a file for a client"),
8186
8187     Option("mds_min_caps_per_client", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
8188     .set_default(100)
8189     .set_description("minimum number of capabilities a client may hold"),
8190
8191     Option("mds_max_caps_per_client", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
8192     .set_default(1_M)
8193     .set_description("maximum number of capabilities a client may hold"),
8194
8195     Option("mds_hack_allow_loading_invalid_metadata", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8196      .set_default(0)
8197      .set_description("INTENTIONALLY CAUSE DATA LOSS by bypasing checks for invalid metadata on disk. Allows testing repair tools."),
8198
8199     Option("mds_defer_session_stale", Option::TYPE_BOOL, Option::LEVEL_DEV)
8200      .set_default(true),
8201
8202     Option("mds_inject_migrator_session_race", Option::TYPE_BOOL, Option::LEVEL_DEV)
8203      .set_default(false),
8204
8205     Option("mds_request_load_average_decay_rate", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
8206     .set_default(60)
8207     .set_description("rate of decay in seconds for calculating request load average"),
8208
8209     Option("mds_cap_revoke_eviction_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
8210      .set_default(0)
8211      .set_description("number of seconds after which clients which have not responded to cap revoke messages by the MDS are evicted."),
8212
8213     Option("mds_max_retries_on_remount_failure", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
8214      .set_default(5)
8215      .set_description("number of consecutive failed remount attempts for invalidating kernel dcache after which client would abort."),
8216
8217     Option("mds_dump_cache_threshold_formatter", Option::TYPE_SIZE, Option::LEVEL_DEV)
8218      .set_default(1_G)
8219      .set_description("threshold for cache usage to disallow \"dump cache\" operation to formatter")
8220      .set_long_description("Disallow MDS from dumping caches to formatter via \"dump cache\" command if cache usage exceeds this threshold."),
8221
8222     Option("mds_dump_cache_threshold_file", Option::TYPE_SIZE, Option::LEVEL_DEV)
8223      .set_default(0)
8224      .set_description("threshold for cache usage to disallow \"dump cache\" operation to file")
8225      .set_long_description("Disallow MDS from dumping caches to file via \"dump cache\" command if cache usage exceeds this threshold."),
8226
8227     Option("mds_task_status_update_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
8228      .set_default(2.0)
8229      .set_description("task status update interval to manager")
8230      .set_long_description("interval (in seconds) for sending mds task status to ceph manager"),
8231
8232     Option("mds_max_snaps_per_dir", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
8233      .set_default(100)
8234      .set_min_max(0, 4096)
8235      .set_flag(Option::FLAG_RUNTIME)
8236      .set_description("max snapshots per directory")
8237      .set_long_description("maximum number of snapshots that can be created per directory"),
8238   });
8239 }
8240
8241 std::vector<Option> get_mds_client_options() {
8242   return std::vector<Option>({
8243     Option("client_cache_size", Option::TYPE_SIZE, Option::LEVEL_BASIC)
8244     .set_default(16384)
8245     .set_description("soft maximum number of directory entries in client cache"),
8246
8247     Option("client_cache_mid", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
8248     .set_default(.75)
8249     .set_description("mid-point of client cache LRU"),
8250
8251     Option("client_use_random_mds", Option::TYPE_BOOL, Option::LEVEL_DEV)
8252     .set_default(false)
8253     .set_description("issue new requests to a random active MDS"),
8254
8255     Option("client_mount_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
8256     .set_default(300.0)
8257     .set_description("timeout for mounting CephFS (seconds)"),
8258
8259     Option("client_tick_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
8260     .set_default(1.0)
8261     .set_description("seconds between client upkeep ticks"),
8262
8263     Option("client_trace", Option::TYPE_STR, Option::LEVEL_DEV)
8264     .set_default("")
8265     .set_description("file containing trace of client operations"),
8266
8267     Option("client_readahead_min", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
8268     .set_default(128*1024)
8269     .set_description("minimum bytes to readahead in a file"),
8270
8271     Option("client_readahead_max_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
8272     .set_default(0)
8273     .set_description("maximum bytes to readahead in a file (zero is unlimited)"),
8274
8275     Option("client_readahead_max_periods", Option::TYPE_INT, Option::LEVEL_ADVANCED)
8276     .set_default(4)
8277     .set_description("maximum stripe periods to readahead in a file"),
8278
8279     Option("client_reconnect_stale", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8280     .set_default(false)
8281     .set_description("reconnect when the session becomes stale"),
8282
8283     Option("client_snapdir", Option::TYPE_STR, Option::LEVEL_ADVANCED)
8284     .set_default(".snap")
8285     .set_description("pseudo directory for snapshot access to a directory"),
8286
8287     Option("client_mountpoint", Option::TYPE_STR, Option::LEVEL_ADVANCED)
8288     .set_default("/")
8289     .set_description("default mount-point"),
8290
8291     Option("client_mount_uid", Option::TYPE_INT, Option::LEVEL_ADVANCED)
8292     .set_default(-1)
8293     .set_description("uid to mount as"),
8294
8295     Option("client_mount_gid", Option::TYPE_INT, Option::LEVEL_ADVANCED)
8296     .set_default(-1)
8297     .set_description("gid to mount as"),
8298
8299     /* RADOS client option */
8300     Option("client_notify_timeout", Option::TYPE_INT, Option::LEVEL_DEV)
8301     .set_default(10)
8302     .set_description(""),
8303
8304     /* RADOS client option */
8305     Option("osd_client_watch_timeout", Option::TYPE_INT, Option::LEVEL_DEV)
8306     .set_default(30)
8307     .set_description(""),
8308
8309     Option("client_caps_release_delay", Option::TYPE_INT, Option::LEVEL_DEV)
8310     .set_default(5)
8311     .set_description(""),
8312
8313     Option("client_quota_df", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8314     .set_default(true)
8315     .set_description("show quota usage for statfs (df)"),
8316
8317     Option("client_oc", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8318     .set_default(true)
8319     .set_description("enable object caching"),
8320
8321     Option("client_oc_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
8322     .set_default(200_M)
8323     .set_description("maximum size of object cache"),
8324
8325     Option("client_oc_max_dirty", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
8326     .set_default(100_M)
8327     .set_description("maximum size of dirty pages in object cache"),
8328
8329     Option("client_oc_target_dirty", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
8330     .set_default(8_M)
8331     .set_description("target size of dirty pages object cache"),
8332
8333     Option("client_oc_max_dirty_age", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
8334     .set_default(5.0)
8335     .set_description("maximum age of dirty pages in object cache (seconds)"),
8336
8337     Option("client_oc_max_objects", Option::TYPE_INT, Option::LEVEL_ADVANCED)
8338     .set_default(1000)
8339     .set_description("maximum number of objects in cache"),
8340
8341     Option("client_debug_getattr_caps", Option::TYPE_BOOL, Option::LEVEL_DEV)
8342     .set_default(false)
8343     .set_description(""),
8344
8345     Option("client_debug_force_sync_read", Option::TYPE_BOOL, Option::LEVEL_DEV)
8346     .set_default(false)
8347     .set_description(""),
8348
8349     Option("client_debug_inject_tick_delay", Option::TYPE_INT, Option::LEVEL_DEV)
8350     .set_default(0)
8351     .set_description(""),
8352
8353     Option("client_max_inline_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
8354     .set_default(4_K)
8355     .set_description(""),
8356
8357     Option("client_inject_release_failure", Option::TYPE_BOOL, Option::LEVEL_DEV)
8358     .set_default(false)
8359     .set_description(""),
8360
8361     Option("client_inject_fixed_oldest_tid", Option::TYPE_BOOL, Option::LEVEL_DEV)
8362     .set_default(false)
8363     .set_description(""),
8364
8365     Option("client_metadata", Option::TYPE_STR, Option::LEVEL_ADVANCED)
8366     .set_default("")
8367     .set_description("metadata key=value comma-delimited pairs appended to session metadata"),
8368
8369     Option("client_acl_type", Option::TYPE_STR, Option::LEVEL_ADVANCED)
8370     .set_default("")
8371     .set_description("ACL type to enforce (none or \"posix_acl\")"),
8372
8373     Option("client_permissions", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8374     .set_default(true)
8375     .set_description("client-enforced permission checking"),
8376
8377     Option("client_dirsize_rbytes", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8378     .set_default(true)
8379     .set_description("set the directory size as the number of file bytes recursively used")
8380     .set_long_description("This option enables a CephFS feature that stores the recursive directory size (the bytes used by files in the directory and its descendents) in the st_size field of the stat structure."),
8381
8382     Option("client_force_lazyio", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8383     .set_default(false)
8384     .set_description(""),
8385
8386     // note: the max amount of "in flight" dirty data is roughly (max - target)
8387     Option("fuse_use_invalidate_cb", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8388     .set_default(true)
8389     .set_description("use fuse 2.8+ invalidate callback to keep page cache consistent"),
8390
8391     Option("fuse_disable_pagecache", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8392     .set_default(false)
8393     .set_description("disable page caching in the kernel for this FUSE mount"),
8394
8395     Option("fuse_allow_other", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8396     .set_default(true)
8397     .set_description("pass allow_other to FUSE on mount"),
8398
8399     Option("fuse_default_permissions", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8400     .set_default(false)
8401     .set_description("pass default_permisions to FUSE on mount")
8402     .set_flag(Option::FLAG_STARTUP),
8403
8404     Option("fuse_big_writes", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8405     .set_default(true)
8406     .set_description("big_writes is deprecated in libfuse 3.0.0"),
8407
8408     Option("fuse_max_write", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
8409     .set_default(0)
8410     .set_description("set the maximum number of bytes in a single write operation")
8411     .set_long_description("Set the maximum number of bytes in a single write operation that may pass atomically through FUSE. The FUSE default is 128kB and may be indicated by setting this option to 0."),
8412
8413     Option("fuse_atomic_o_trunc", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8414     .set_default(true)
8415     .set_description("pass atomic_o_trunc flag to FUSE on mount"),
8416
8417     Option("fuse_debug", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8418     .set_default(false)
8419     .set_flag(Option::FLAG_STARTUP)
8420     .set_flag(Option::FLAG_NO_MON_UPDATE)
8421     .set_description("enable debugging for the libfuse"),
8422
8423     Option("fuse_multithreaded", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8424     .set_default(true)
8425     .set_description("allow parallel processing through FUSE library"),
8426
8427     Option("fuse_require_active_mds", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8428     .set_default(true)
8429     .set_description("require active MDSs in the file system when mounting"),
8430
8431     Option("fuse_syncfs_on_mksnap", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8432     .set_default(true)
8433     .set_description("synchronize all local metadata/file changes after snapshot"),
8434
8435     Option("fuse_set_user_groups", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8436     .set_default(true)
8437     .set_description("check for ceph-fuse to consider supplementary groups for permissions"),
8438
8439     Option("client_try_dentry_invalidate", Option::TYPE_BOOL, Option::LEVEL_DEV)
8440     .set_default(false)
8441     .set_description(""),
8442
8443     Option("client_die_on_failed_remount", Option::TYPE_BOOL, Option::LEVEL_DEV)
8444     .set_default(false)
8445     .set_description(""),
8446
8447     Option("client_die_on_failed_dentry_invalidate", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8448     .set_default(true)
8449     .set_description("kill the client when no dentry invalidation options are available")
8450     .set_long_description("The CephFS client requires a mechanism to invalidate dentries in the caller (e.g. the kernel for ceph-fuse) when capabilities must be recalled. If the client cannot do this then the MDS cache cannot shrink which can cause the MDS to fail."),
8451
8452     Option("client_check_pool_perm", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
8453     .set_default(true)
8454     .set_description("confirm access to inode's data pool/namespace described in file layout"),
8455
8456     Option("client_use_faked_inos", Option::TYPE_BOOL, Option::LEVEL_DEV)
8457     .set_default(false)
8458     .set_description(""),
8459
8460     Option("client_fs", Option::TYPE_STR, Option::LEVEL_ADVANCED)
8461     .set_flag(Option::FLAG_STARTUP)
8462     .set_default("")
8463     .set_description("CephFS file system name to mount")
8464     .set_long_description("Use this with ceph-fuse, or with any process "
8465         "that uses libcephfs.  Programs using libcephfs may also pass "
8466         "the filesystem name into mount(), which will override this setting. "
8467         "If no filesystem name is given in mount() or this setting, the default "
8468         "filesystem will be mounted (usually the first created)."),
8469
8470     /* Alias for client_fs. Deprecated */
8471     Option("client_mds_namespace", Option::TYPE_STR, Option::LEVEL_DEV)
8472     .set_flag(Option::FLAG_STARTUP)
8473     .set_default(""),
8474
8475     Option("fake_statfs_for_testing", Option::TYPE_INT, Option::LEVEL_DEV)
8476     .set_default(0)
8477     .set_description("Set a value for kb and compute kb_used from total of num_bytes"),
8478
8479     Option("debug_allow_any_pool_priority", Option::TYPE_BOOL, Option::LEVEL_DEV)
8480     .set_default(false)
8481     .set_description("Allow any pool priority to be set to test conversion to new range"),
8482
8483     Option("client_shutdown_timeout", Option::TYPE_SECS, Option::LEVEL_ADVANCED)
8484     .set_flag(Option::FLAG_RUNTIME)
8485     .set_default(30)
8486     .set_min(0)
8487     .set_description("timeout for shutting down CephFS")
8488     .set_long_description("Timeout for shutting down CephFS via unmount or shutdown.")
8489     .add_tag("client")
8490   });
8491 }
8492
8493
8494 std::vector<Option> get_cephfs_shell_options() {
8495   return std::vector<Option>({
8496     Option("allow_ansi", Option::TYPE_STR, Option::LEVEL_BASIC)
8497     .set_default("Terminal")
8498     .set_description("Allow ANSI escape sequences in output. Values: "
8499                      "Terminal, Always, Never"),
8500
8501     Option("colors", Option::TYPE_STR, Option::LEVEL_BASIC)
8502     .set_default("Terminal")
8503     .set_description("Colouring CephFS shell input and output. Values: "
8504                      "Terminal, Always, Never"),
8505
8506     Option("continuation_prompt", Option::TYPE_STR, Option::LEVEL_BASIC)
8507     .set_default(">")
8508     .set_description("Prompt string when a command continue to second line"),
8509
8510     Option("debug_shell", Option::TYPE_BOOL, Option::LEVEL_BASIC)
8511     .set_default(false)
8512     .set_description("Allow tracebacks on error for CephFS Shell"),
8513
8514     Option("echo", Option::TYPE_BOOL, Option::LEVEL_BASIC)
8515     .set_default(false)
8516     .set_description("Print command issued on prompt before execution"),
8517
8518     Option("editor", Option::TYPE_STR, Option::LEVEL_BASIC)
8519     .set_default("vim")
8520     .set_description("Default text editor for shell"),
8521
8522     Option("feedback_to_output", Option::TYPE_BOOL, Option::LEVEL_BASIC)
8523     .set_default(false)
8524     .set_description("include '|' and '>' in result"),
8525
8526     Option("max_completion_items", Option::TYPE_INT, Option::LEVEL_BASIC)
8527     .set_default(50)
8528     .set_description("Maximum number of items to be displayed by tab "
8529                      "completion"),
8530
8531     Option("prompt", Option::TYPE_STR, Option::LEVEL_BASIC)
8532     .set_default("\x1b[01;33mCephFS:~\x1b[96m/\x1b[0m\x1b[01;33m>>>\x1b[00m ")
8533     .set_description("Whether non-essential feedback should be printed."),
8534
8535     Option("quiet", Option::TYPE_BOOL, Option::LEVEL_BASIC)
8536     .set_default(false)
8537     .set_description("Whether non-essential feedback should be printed."),
8538
8539     Option("timing", Option::TYPE_BOOL, Option::LEVEL_BASIC)
8540     .set_default(false)
8541     .set_description("Whether execution time should be reported"),
8542   });
8543 }
8544
8545 static std::vector<Option> build_options()
8546 {
8547   std::vector<Option> result = get_global_options();
8548
8549   auto ingest = [&result](std::vector<Option>&& options, const char* svc) {
8550     for (auto &o : options) {
8551       o.add_service(svc);
8552       result.push_back(std::move(o));
8553     }
8554   };
8555
8556   ingest(get_rgw_options(), "rgw");
8557   ingest(get_rbd_options(), "rbd");
8558   ingest(get_rbd_mirror_options(), "rbd-mirror");
8559   ingest(get_immutable_object_cache_options(), "immutable-objet-cache");
8560   ingest(get_mds_options(), "mds");
8561   ingest(get_mds_client_options(), "mds_client");
8562   ingest(get_cephfs_shell_options(), "cephfs-shell");
8563
8564   return result;
8565 }
8566
8567 const std::vector<Option> ceph_options = build_options();