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