// Helpers for validators
#include "include/stringify.h"
+#include "include/common_fwd.h"
#include <boost/algorithm/string.hpp>
#include <boost/lexical_cast.hpp>
#include <regex>
return -EINVAL;
}
*out = addr;
- } else if (type == Option::TYPE_ADDR) {
+ } else if (type == Option::TYPE_ADDRVEC) {
entity_addrvec_t addr;
if (!addr.parse(val.c_str())){
return -EINVAL;
constexpr unsigned long long operator"" _G (unsigned long long n) {
return n << 30;
}
+constexpr unsigned long long operator"" _T (unsigned long long n) {
+ return n << 40;
+}
std::vector<Option> get_global_options() {
return std::vector<Option>({
.set_flag(Option::FLAG_STARTUP)
.add_service({"mon", "mds", "osd", "mgr"}),
+ Option("public_addrv", Option::TYPE_ADDRVEC, Option::LEVEL_BASIC)
+ .set_description("public-facing address to bind to")
+ .set_flag(Option::FLAG_STARTUP)
+ .add_service({"mon", "mds", "osd", "mgr"}),
+
Option("public_bind_addr", Option::TYPE_ADDR, Option::LEVEL_ADVANCED)
.set_default(entity_addr_t())
.set_flag(Option::FLAG_STARTUP)
Option("mon_dns_srv_name", Option::TYPE_STR, Option::LEVEL_ADVANCED)
.set_default("ceph-mon")
.set_description("name of DNS SRV record to check for monitor addresses")
- .set_flag(Option::FLAG_NO_MON_UPDATE)
.set_flag(Option::FLAG_STARTUP)
.add_service("common")
.add_tag("network")
.add_see_also("mon_host"),
+ Option("container_image", Option::TYPE_STR, Option::LEVEL_BASIC)
+ .set_description("container image (used by cephadm orchestrator)")
+ .set_flag(Option::FLAG_STARTUP)
+ .set_default("docker.io/ceph/ceph:v15"),
+
+ Option("no_config_file", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(false)
+ .set_flag(Option::FLAG_NO_MON_UPDATE)
+ .set_flag(Option::FLAG_STARTUP)
+ .add_service("common")
+ .add_tag("config")
+ .set_description("signal that we don't require a config file to be present")
+ .set_long_description("When specified, we won't be looking for a "
+ "configuration file, and will instead expect that "
+ "whatever options or values are required for us to "
+ "work will be passed as arguments."),
+
// lockdep
Option("lockdep", Option::TYPE_BOOL, Option::LEVEL_DEV)
.set_description("enable lockdep lock dependency analyzer")
.add_service({"mon", "osd"})
.set_description("Base directory for dynamically loaded plugins"),
- // XIO
- Option("xio_trace_mempool", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
- .set_default(false)
- .set_description(""),
-
- Option("xio_trace_msgcnt", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
- .set_default(false)
- .set_description(""),
-
- Option("xio_trace_xcon", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
- .set_default(false)
- .set_description(""),
-
- Option("xio_queue_depth", Option::TYPE_INT, Option::LEVEL_ADVANCED)
- .set_default(128)
- .set_description(""),
-
- Option("xio_mp_min", Option::TYPE_INT, Option::LEVEL_ADVANCED)
- .set_default(128)
- .set_description(""),
-
- Option("xio_mp_max_64", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
- .set_default(65536)
- .set_description(""),
-
- Option("xio_mp_max_256", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
- .set_default(8192)
- .set_description(""),
-
- Option("xio_mp_max_1k", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
- .set_default(8192)
- .set_description(""),
-
- Option("xio_mp_max_page", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
- .set_default(4096)
- .set_description(""),
-
- Option("xio_mp_max_hint", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
- .set_default(4096)
- .set_description(""),
-
- Option("xio_portal_threads", Option::TYPE_INT, Option::LEVEL_ADVANCED)
- .set_default(2)
- .set_description(""),
-
- Option("xio_max_conns_per_portal", Option::TYPE_INT, Option::LEVEL_ADVANCED)
- .set_default(32)
- .set_description(""),
-
- Option("xio_transport_type", Option::TYPE_STR, Option::LEVEL_ADVANCED)
- .set_default("rdma")
- .set_description(""),
-
- Option("xio_max_send_inline", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
- .set_default(512)
- .set_description(""),
-
// Compressor
Option("compressor_zlib_isal", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(false)
.set_flag(Option::FLAG_NO_MON_UPDATE)
.set_description(""),
+ Option("thp", Option::TYPE_BOOL, Option::LEVEL_DEV)
+ .set_default(false)
+ .set_flag(Option::FLAG_STARTUP)
+ .set_description("enable transparent huge page (THP) support")
+ .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."),
+
Option("key", Option::TYPE_STR, Option::LEVEL_ADVANCED)
.set_default("")
.set_description("Authentication key")
.set_default(100_M)
.set_description("Limit messages that are read off the network but still being processed"),
- Option("ms_msgr2_sign_messages", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
- .set_default(false)
- .set_description("Sign msgr2 frames' payload")
- .add_see_also("ms_msgr2_encrypt_messages"),
-
- Option("ms_msgr2_encrypt_messages", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
- .set_default(false)
- .set_description("Encrypt msgr2 frames' payload")
- .add_see_also("ms_msgr2_sign_messages"),
-
Option("ms_bind_ipv4", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(true)
.set_description("Bind servers to IPv4 address(es)")
.set_default(512)
.set_description("Size of queue of incoming connections for accept(2)"),
- Option("ms_rwthread_stack_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
- .set_default(1_M)
- .set_description("Size of stack for SimpleMessenger read/write threads"),
Option("ms_connection_ready_timeout", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
.set_default(10)
.set_default(0)
.set_description("Inject various internal delays to induce races (seconds)"),
+ Option("ms_blackhole_osd", Option::TYPE_BOOL, Option::LEVEL_DEV)
+ .set_default(false)
+ .set_description(""),
+
+ Option("ms_blackhole_mon", Option::TYPE_BOOL, Option::LEVEL_DEV)
+ .set_default(false)
+ .set_description(""),
+
+ Option("ms_blackhole_mds", Option::TYPE_BOOL, Option::LEVEL_DEV)
+ .set_default(false)
+ .set_description(""),
+
+ Option("ms_blackhole_mgr", Option::TYPE_BOOL, Option::LEVEL_DEV)
+ .set_default(false)
+ .set_description(""),
+
+ Option("ms_blackhole_client", Option::TYPE_BOOL, Option::LEVEL_DEV)
+ .set_default(false)
+ .set_description(""),
+
Option("ms_dump_on_send", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(false)
.set_description("Hexdump message to debug log on message send"),
.set_default(1000)
.set_description(""),
+ Option("ms_async_rdma_gid_idx", Option::TYPE_INT, Option::LEVEL_ADVANCED)
+ .set_default(0)
+ .set_description("use gid_idx to select GID for choosing RoCEv1 or RoCEv2"),
+
Option("ms_async_rdma_local_gid", Option::TYPE_STR, Option::LEVEL_ADVANCED)
.set_default("")
.set_description(""),
.add_service("mon")
.set_description("maximum number of OSDMaps to cache in memory"),
+ Option("mon_osd_cache_size_min", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
+ .set_default(128_M)
+ .add_service("mon")
+ .set_description("The minimum amount of bytes to be kept mapped in memory for osd monitor caches."),
+
+ Option("mon_memory_target", Option::TYPE_SIZE, Option::LEVEL_BASIC)
+ .set_default(2_G)
+ .set_flag(Option::FLAG_RUNTIME)
+ .add_service("mon")
+ .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"),
+
+ Option("mon_memory_autotune", Option::TYPE_BOOL, Option::LEVEL_BASIC)
+ .set_default(true)
+ .set_flag(Option::FLAG_RUNTIME)
+ .add_service("mon")
+ .set_description("Autotune the cache memory being used for osd monitors and kv database"),
+
Option("mon_cpu_threads", Option::TYPE_INT, Option::LEVEL_ADVANCED)
.set_default(4)
.add_service("mon")
.add_service("mon")
.set_description("granularity of PG placement calculation background work"),
- Option("mon_clean_pg_upmaps_per_chunk", Option::TYPE_INT, Option::LEVEL_DEV)
+ Option("mon_clean_pg_upmaps_per_chunk", Option::TYPE_UINT, Option::LEVEL_DEV)
.set_default(256)
.add_service("mon")
.set_description("granularity of PG upmap validation background work"),
.add_service("mgr"),
Option("mon_pg_warn_min_per_osd", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
- .set_default(30)
+ .set_default(0)
.add_service("mgr")
.set_description("minimal number PGs per (in) osd before we warn the admin"),
.add_service("mgr")
.set_description("issue POOL_APP_NOT_ENABLED health warning if pool has not application enabled"),
+ Option("mon_warn_on_pool_pg_num_not_power_of_two", Option::TYPE_BOOL, Option::LEVEL_DEV)
+ .set_default(true)
+ .add_service("mon")
+ .set_description("issue POOL_PG_NUM_NOT_POWER_OF_TWO warning if pool has a non-power-of-two pg_num value"),
+
+ Option("mon_warn_on_pool_no_redundancy", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(true)
+ .add_service("mon")
+ .set_description("Issue a health warning if any pool is configured with no replicas")
+ .add_see_also("osd_pool_default_size")
+ .add_see_also("osd_pool_default_min_size"),
+
Option("mon_warn_on_misplaced", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(false)
.add_service("mgr")
.set_description("Issue a health warning if there are misplaced objects"),
+ Option("mon_warn_on_too_few_osds", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(true)
+ .add_service("mgr")
+ .set_description("Issue a health warning if there are fewer OSDs than osd_pool_default_size"),
+
+ Option("mon_warn_on_slow_ping_time", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
+ .set_default(0)
+ .add_service("mgr")
+ .set_description("Override mon_warn_on_slow_ping_ratio with specified threshold in milliseconds")
+ .add_see_also("mon_warn_on_slow_ping_ratio"),
+
+ Option("mon_warn_on_slow_ping_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
+ .set_default(.05)
+ .add_service("mgr")
+ .set_description("Issue a health warning if heartbeat ping longer than percentage of osd_heartbeat_grace")
+ .add_see_also("osd_heartbeat_grace")
+ .add_see_also("mon_warn_on_slow_ping_time"),
+
Option("mon_max_snap_prune_per_epoch", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
.set_default(100)
.add_service("mon")
.add_service("mon")
.set_description("target max message payload for mon sync"),
+ Option("mon_sync_max_payload_keys", Option::TYPE_INT, Option::LEVEL_ADVANCED)
+ .set_default(2000)
+ .add_service("mon")
+ .set_description("target max keys in message payload for mon sync"),
+
Option("mon_sync_debug", Option::TYPE_BOOL, Option::LEVEL_DEV)
.set_default(false)
.add_service("mon")
.set_description("file to dump paxos transactions to")
.add_see_also("mon_debug_dump_transactions"),
- Option("mon_debug_no_require_mimic", Option::TYPE_BOOL, Option::LEVEL_DEV)
+ Option("mon_debug_no_require_nautilus", Option::TYPE_BOOL, Option::LEVEL_DEV)
.set_default(false)
.add_service("mon")
.set_flag(Option::FLAG_CLUSTER_CREATE)
- .set_description("do not set mimic feature for new mon clusters"),
+ .set_description("do not set nautilus feature for new mon clusters"),
- Option("mon_debug_no_require_nautilus", Option::TYPE_BOOL, Option::LEVEL_DEV)
+ Option("mon_debug_no_require_octopus", Option::TYPE_BOOL, Option::LEVEL_DEV)
.set_default(false)
.add_service("mon")
.set_flag(Option::FLAG_CLUSTER_CREATE)
- .set_description("do not set nautilus feature for new mon clusters"),
+ .set_description("do not set octopus feature for new mon clusters"),
Option("mon_debug_no_require_bluestore_for_ec_overwrites", Option::TYPE_BOOL, Option::LEVEL_DEV)
.set_default(false)
.set_min(1_hr)
.add_service("mon")
.set_description("Duration in seconds that blacklist entries for MDS "
- "daemons remain in the OSD map"),
+ "daemons remain in the OSD map")
+ .set_flag(Option::FLAG_RUNTIME),
+
+ Option("mon_mgr_blacklist_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
+ .set_default(1_day)
+ .set_min(1_hr)
+ .add_service("mon")
+ .set_description("Duration in seconds that blacklist entries for mgr "
+ "daemons remain in the OSD map")
+ .set_flag(Option::FLAG_RUNTIME),
Option("mon_osd_crush_smoke_test", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(true)
.set_default(3.0)
.set_description(""),
+ Option("mon_client_log_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
+ .set_default(1.0)
+ .set_description("How frequently we send queued cluster log messages to mon"),
+
Option("mon_client_ping_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
.set_default(10.0)
.set_description(""),
.set_default(1000)
.set_description(""),
+ Option("mon_client_directed_command_retry", Option::TYPE_INT, Option::LEVEL_DEV)
+ .set_default(2)
+ .set_description("Number of times to try sending a comamnd directed at a specific monitor"),
+
Option("mon_max_pool_pg_num", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
.set_default(65536)
.set_description(""),
"by doing a fairly exhaustive search of existing PGs "
"that can be unmapped or upmapped"),
- Option("osd_calc_pg_upmaps_max_stddev", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(1.0)
- .set_flag(Option::FLAG_RUNTIME)
- .set_description("standard deviation below which there is no attempt made "
- "while trying to calculate PG upmaps"),
-
Option("osd_calc_pg_upmaps_local_fallback_retries", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
.set_default(100)
.set_flag(Option::FLAG_RUNTIME)
Option("osd_crush_chooseleaf_type", Option::TYPE_INT, Option::LEVEL_DEV)
.set_default(1)
- .set_flag(Option::FLAG_CREATE)
+ .set_flag(Option::FLAG_CLUSTER_CREATE)
.set_description("default chooseleaf type for osdmaptool --create"),
Option("osd_pool_use_gmt_hitset", Option::TYPE_BOOL, Option::LEVEL_DEV)
.add_service("mon"),
Option("osd_pool_default_pg_num", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
- .set_default(8)
+ .set_default(32)
.set_description("number of PGs for new pools")
.set_flag(Option::FLAG_RUNTIME)
.add_service("mon"),
.add_service("mon"),
Option("osd_pool_default_erasure_code_profile", Option::TYPE_STR, Option::LEVEL_ADVANCED)
- .set_default("plugin=jerasure technique=reed_sol_van k=2 m=1")
+ .set_default("plugin=jerasure technique=reed_sol_van k=2 m=2")
.set_flag(Option::FLAG_RUNTIME)
.set_description("default erasure code profile for new erasure-coded pools")
.add_service("mon"),
.set_description(""),
Option("osd_pool_default_pg_autoscale_mode", Option::TYPE_STR, Option::LEVEL_ADVANCED)
- .set_default("warn")
+ .set_default("on")
.set_flag(Option::FLAG_RUNTIME)
.set_enum_allowed({"off", "warn", "on"})
.set_description("Default PG autoscaling behavior for new pools"),
+ Option("osd_pool_default_read_lease_ratio", Option::TYPE_FLOAT, Option::LEVEL_DEV)
+ .set_default(.8)
+ .set_flag(Option::FLAG_RUNTIME)
+ .set_description("Default read_lease_ratio for a pool, as a multiple of osd_heartbeat_grace")
+ .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.")
+ .add_see_also("osd_heartbeat_grace"),
+
Option("osd_hit_set_min_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
.set_default(1000)
.set_description(""),
.set_default(true)
.set_description(""),
+ Option("osd_num_cache_shards", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
+ .set_default(32)
+ .set_flag(Option::FLAG_STARTUP)
+ .set_description("The number of cache shards to use in the object store."),
+
Option("osd_op_num_threads_per_shard", Option::TYPE_INT, Option::LEVEL_ADVANCED)
.set_default(0)
.set_flag(Option::FLAG_STARTUP)
Option("osd_op_queue", Option::TYPE_STR, Option::LEVEL_ADVANCED)
.set_default("wpq")
- .set_enum_allowed( { "wpq", "prioritized", "mclock_opclass", "mclock_client", "debug_random" } )
- .set_description("which operation queue algorithm to use")
- .set_long_description("which operation queue algorithm to use; mclock_opclass and mclock_client are currently experimental")
- .set_flag(Option::FLAG_STARTUP)
+ .set_enum_allowed( { "wpq", "prioritized",
+ "mclock_opclass", "mclock_client", "mclock_scheduler",
+ "debug_random" } )
+ .set_description("which operation priority queue algorithm to use")
+ .set_long_description("which operation priority queue algorithm to use; "
+ "mclock_opclass mclock_client, and "
+ "mclock_client_profile are currently experimental")
.add_see_also("osd_op_queue_cut_off"),
Option("osd_op_queue_cut_off", Option::TYPE_STR, Option::LEVEL_ADVANCED)
- .set_default("low")
+ .set_default("high")
.set_enum_allowed( { "low", "high", "debug_random" } )
.set_description("the threshold between high priority ops and low priority ops")
.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")
.add_see_also("osd_op_queue"),
- Option("osd_op_queue_mclock_client_op_res", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(1000.0)
- .set_description("mclock reservation of client operator requests")
- .set_long_description("mclock reservation of client operator requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the reservation")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim")
- .add_see_also("osd_op_queue_mclock_anticipation_timeout"),
-
- Option("osd_op_queue_mclock_client_op_wgt", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(500.0)
- .set_description("mclock weight of client operator requests")
- .set_long_description("mclock weight of client operator requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the weight")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim")
- .add_see_also("osd_op_queue_mclock_anticipation_timeout"),
-
- Option("osd_op_queue_mclock_client_op_lim", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(0.0)
- .set_description("mclock limit of client operator requests")
- .set_long_description("mclock limit of client operator requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the limit")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim")
- .add_see_also("osd_op_queue_mclock_anticipation_timeout"),
-
- Option("osd_op_queue_mclock_osd_rep_op_res", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(1000.0)
- .set_description("mclock reservation of osd replication operation requests and replies")
- .set_long_description("mclock reservation of replication operation requests and replies when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the reservation")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim")
- .add_see_also("osd_op_queue_mclock_anticipation_timeout"),
-
- Option("osd_op_queue_mclock_osd_rep_op_wgt", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(500.0)
- .set_description("mclock weight of osd replication operation requests and replies")
- .set_long_description("mclock weight of osd replication operation requests and replies when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the weight")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim")
- .add_see_also("osd_op_queue_mclock_anticipation_timeout"),
-
- Option("osd_op_queue_mclock_osd_rep_op_lim", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(0.0)
- .set_description("mclock limit of osd replication operation requests and replies")
- .set_long_description("mclock limit of osd sub-operation requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the limit")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim")
- .add_see_also("osd_op_queue_mclock_anticipation_timeout"),
-
- Option("osd_op_queue_mclock_snap_res", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(0.0)
- .set_description("mclock reservation of snaptrim requests")
- .set_long_description("mclock reservation of snaptrim requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the reservation")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim")
- .add_see_also("osd_op_queue_mclock_anticipation_timeout"),
-
- Option("osd_op_queue_mclock_snap_wgt", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(1.0)
- .set_description("mclock weight of snaptrim requests")
- .set_long_description("mclock weight of snaptrim requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the weight")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim")
- .add_see_also("osd_op_queue_mclock_anticipation_timeout"),
-
- Option("osd_op_queue_mclock_snap_lim", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(0.001)
- .set_description("")
- .set_description("mclock limit of snaptrim requests")
- .set_long_description("mclock limit of snaptrim requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the limit")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim")
- .add_see_also("osd_op_queue_mclock_anticipation_timeout"),
-
- Option("osd_op_queue_mclock_recov_res", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(0.0)
- .set_description("mclock reservation of recovery requests")
- .set_long_description("mclock reservation of recovery requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the reservation")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim")
- .add_see_also("osd_op_queue_mclock_anticipation_timeout"),
-
- Option("osd_op_queue_mclock_recov_wgt", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(1.0)
- .set_description("mclock weight of recovery requests")
- .set_long_description("mclock weight of recovery requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the weight")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim")
- .add_see_also("osd_op_queue_mclock_anticipation_timeout"),
-
- Option("osd_op_queue_mclock_recov_lim", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(0.001)
- .set_description("mclock limit of recovery requests")
- .set_long_description("mclock limit of recovery requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the limit")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim")
- .add_see_also("osd_op_queue_mclock_anticipation_timeout"),
-
- Option("osd_op_queue_mclock_scrub_res", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(0.0)
- .set_description("mclock reservation of scrub requests")
- .set_long_description("mclock reservation of scrub requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the reservation")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim")
- .add_see_also("osd_op_queue_mclock_anticipation_timeout"),
-
- Option("osd_op_queue_mclock_scrub_wgt", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(1.0)
- .set_description("mclock weight of scrub requests")
- .set_long_description("mclock weight of scrub requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the weight")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_lim")
- .add_see_also("osd_op_queue_mclock_anticipation_timeout"),
-
- Option("osd_op_queue_mclock_scrub_lim", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(0.001)
- .set_description("mclock weight of limit requests")
- .set_long_description("mclock weight of limit requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the limit")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_anticipation_timeout"),
-
- Option("osd_op_queue_mclock_anticipation_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
+ Option("osd_mclock_scheduler_client_res", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(1)
+ .set_description("IO proportion reserved for each client (default)")
+ .set_long_description("Only considered for osd_op_queue = mClockScheduler")
+ .add_see_also("osd_op_queue"),
+
+ Option("osd_mclock_scheduler_client_wgt", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(1)
+ .set_description("IO share for each client (default) over reservation")
+ .set_long_description("Only considered for osd_op_queue = mClockScheduler")
+ .add_see_also("osd_op_queue"),
+
+ Option("osd_mclock_scheduler_client_lim", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(999999)
+ .set_description("IO limit for each client (default) over reservation")
+ .set_long_description("Only considered for osd_op_queue = mClockScheduler")
+ .add_see_also("osd_op_queue"),
+
+ Option("osd_mclock_scheduler_background_recovery_res", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(1)
+ .set_description("IO proportion reserved for background recovery (default)")
+ .set_long_description("Only considered for osd_op_queue = mClockScheduler")
+ .add_see_also("osd_op_queue"),
+
+ Option("osd_mclock_scheduler_background_recovery_wgt", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(1)
+ .set_description("IO share for each background recovery over reservation")
+ .set_long_description("Only considered for osd_op_queue = mClockScheduler")
+ .add_see_also("osd_op_queue"),
+
+ Option("osd_mclock_scheduler_background_recovery_lim", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(999999)
+ .set_description("IO limit for background recovery over reservation")
+ .set_long_description("Only considered for osd_op_queue = mClockScheduler")
+ .add_see_also("osd_op_queue"),
+
+ Option("osd_mclock_scheduler_background_best_effort_res", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(1)
+ .set_description("IO proportion reserved for background best_effort (default)")
+ .set_long_description("Only considered for osd_op_queue = mClockScheduler")
+ .add_see_also("osd_op_queue"),
+
+ Option("osd_mclock_scheduler_background_best_effort_wgt", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(1)
+ .set_description("IO share for each background best_effort over reservation")
+ .set_long_description("Only considered for osd_op_queue = mClockScheduler")
+ .add_see_also("osd_op_queue"),
+
+ Option("osd_mclock_scheduler_background_best_effort_lim", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(999999)
+ .set_description("IO limit for background best_effort over reservation")
+ .set_long_description("Only considered for osd_op_queue = mClockScheduler")
+ .add_see_also("osd_op_queue"),
+
+ Option("osd_mclock_scheduler_anticipation_timeout", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
.set_default(0.0)
.set_description("mclock anticipation timeout in seconds")
- .set_long_description("the amount of time that mclock waits until the unused resource is forfeited")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim"),
-
- Option("osd_op_queue_mclock_pg_delete_res", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(0.0)
- .set_description("mclock reservation of pg delete work")
- .set_long_description("mclock reservation of pg delete work when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the reservation")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim"),
-
- Option("osd_op_queue_mclock_pg_delete_wgt", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(1.0)
- .set_description("mclock weight of pg delete work")
- .set_long_description("mclock weight of pg delete work when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the weight")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_lim"),
-
- Option("osd_op_queue_mclock_pg_delete_lim", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(0.001)
- .set_description("mclock weight of pg delete work limit requests")
- .set_long_description("mclock weight of limit pg delete work when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the limit")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt"),
-
- Option("osd_op_queue_mclock_peering_event_res", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(0.0)
- .set_description("mclock reservation of peering events")
- .set_long_description("mclock reservation of scrub requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the reservation")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_wgt")
- .add_see_also("osd_op_queue_mclock_scrub_lim"),
-
- Option("osd_op_queue_mclock_peering_event_wgt", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(1.0)
- .set_description("mclock weight of peering events")
- .set_long_description("mclock weight of scrub requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the weight")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_lim"),
-
- Option("osd_op_queue_mclock_peering_event_lim", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
- .set_default(0.001)
- .set_description("mclock weight of limit peering events")
- .set_long_description("mclock weight of limit requests when osd_op_queue is either 'mclock_opclass' or 'mclock_client'; higher values increase the limit")
- .add_see_also("osd_op_queue")
- .add_see_also("osd_op_queue_mclock_client_op_res")
- .add_see_also("osd_op_queue_mclock_client_op_wgt")
- .add_see_also("osd_op_queue_mclock_client_op_lim")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_res")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_wgt")
- .add_see_also("osd_op_queue_mclock_osd_rep_op_lim")
- .add_see_also("osd_op_queue_mclock_snap_res")
- .add_see_also("osd_op_queue_mclock_snap_wgt")
- .add_see_also("osd_op_queue_mclock_snap_lim")
- .add_see_also("osd_op_queue_mclock_recov_res")
- .add_see_also("osd_op_queue_mclock_recov_wgt")
- .add_see_also("osd_op_queue_mclock_recov_lim")
- .add_see_also("osd_op_queue_mclock_scrub_res")
- .add_see_also("osd_op_queue_mclock_scrub_wgt"),
+ .set_long_description("the amount of time that mclock waits until the unused resource is forfeited"),
Option("osd_ignore_stale_divergent_priors", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(false)
.set_default(false)
.set_description(""),
- Option("osd_recover_clone_overlap_limit", Option::TYPE_INT, Option::LEVEL_ADVANCED)
+ // Only use clone_overlap for recovery if there are fewer than
+ // osd_recover_clone_overlap_limit entries in the overlap set
+ Option("osd_recover_clone_overlap_limit", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
.set_default(10)
- .set_description(""),
+ .set_description("")
+ .set_flag(Option::FLAG_RUNTIME),
Option("osd_debug_feed_pullee", Option::TYPE_INT, Option::LEVEL_DEV)
.set_default(-1)
.set_default(15_min)
.set_description(""),
- Option("osd_heartbeat_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
+ Option("osd_heartbeat_interval", Option::TYPE_INT, Option::LEVEL_DEV)
.set_default(6)
- .set_min_max(1, 86400)
+ .set_min_max(1, 60)
.set_description("Interval (in seconds) between peer pings"),
Option("osd_heartbeat_grace", Option::TYPE_INT, Option::LEVEL_ADVANCED)
.set_default(30)
.set_description(""),
+ Option("osd_mon_heartbeat_stat_stale", Option::TYPE_INT, Option::LEVEL_ADVANCED)
+ .set_default(1_hr)
+ .set_description("Stop reporting on heartbeat ping times not updated for this many seconds.")
+ .set_long_description("Stop reporting on old heartbeat information unless this is set to zero"),
+
Option("osd_mon_report_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
.set_default(5)
.set_description("Frequency of OSD reports to mon for peer failures, fullness status changes"),
.set_description(""),
Option("osd_recovery_max_active", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(0)
+ .set_description("Number of simultaneous active recovery operations per OSD (overrides _ssd and _hdd if non-zero)")
+ .add_see_also("osd_recovery_max_active_hdd")
+ .add_see_also("osd_recovery_max_active_ssd"),
+
+ Option("osd_recovery_max_active_hdd", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
.set_default(3)
- .set_description(""),
+ .set_description("Number of simultaneous active recovery oeprations per OSD (for rotational devices)")
+ .add_see_also("osd_recovery_max_active")
+ .add_see_also("osd_recovery_max_active_ssd"),
+
+ Option("osd_recovery_max_active_ssd", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(10)
+ .set_description("Number of simultaneous active recovery oeprations per OSD (for non-rotational solid state devices)")
+ .add_see_also("osd_recovery_max_active")
+ .add_see_also("osd_recovery_max_active_hdd"),
Option("osd_recovery_max_single_start", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
.set_default(1)
.set_default(false)
.set_description("Allow scrubbing when PGs on the OSD are undergoing recovery"),
+ Option("osd_repair_during_recovery", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(false)
+ .set_description("Allow requested repairing when PGs on the OSD are undergoing recovery"),
+
Option("osd_scrub_begin_hour", Option::TYPE_INT, Option::LEVEL_ADVANCED)
.set_default(0)
.set_description("Restrict scrubbing to this hour of the day or later")
.set_default(0)
.set_description("Duration to inject a delay during scrubbing"),
+ Option("osd_scrub_extended_sleep", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
+ .set_default(0)
+ .set_description("Duration to inject a delay during scrubbing out of scrubbing hours")
+ .add_see_also("osd_scrub_begin_hour")
+ .add_see_also("osd_scrub_end_hour")
+ .add_see_also("osd_scrub_begin_week_day")
+ .add_see_also("osd_scrub_end_week_day"),
+
Option("osd_scrub_auto_repair", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(false)
.set_description("Automatically repair damaged objects detected during scrub"),
.set_description(""),
Option("osd_class_load_list", Option::TYPE_STR, Option::LEVEL_ADVANCED)
- .set_default("cephfs hello journal lock log numops " "otp rbd refcount rgw timeindex user version cas")
+ .set_default("cephfs hello journal lock log numops " "otp rbd refcount rgw rgw_gc timeindex user version cas")
.set_description(""),
Option("osd_class_default_list", Option::TYPE_STR, Option::LEVEL_ADVANCED)
- .set_default("cephfs hello journal lock log numops " "otp rbd refcount rgw timeindex user version cas")
+ .set_default("cephfs hello journal lock log numops " "otp rbd refcount rgw rgw_gc timeindex user version cas")
.set_description(""),
Option("osd_check_for_log_corruption", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(40)
.set_description(""),
+ Option("osd_target_pg_log_entries_per_osd", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(3000 * 100)
+ .set_description("target number of PG entries total on an OSD")
+ .add_see_also("osd_max_pg_log_entries")
+ .add_see_also("osd_min_pg_log_entries"),
+
Option("osd_min_pg_log_entries", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
- .set_default(3000)
+ .set_default(250)
.set_description("minimum number of entries to maintain in the PG log")
.add_service("osd")
.add_see_also("osd_max_pg_log_entries")
.add_see_also("osd_pg_log_dups_tracked"),
Option("osd_max_pg_log_entries", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
- .set_default(3000)
+ .set_default(10000)
.set_description("maximum number of entries to maintain in the PG log when degraded before we trim")
.add_service("osd")
.add_see_also("osd_min_pg_log_entries")
.add_see_also("osd_min_pg_log_entries")
.add_see_also("osd_max_pg_log_entries"),
+ Option("osd_object_clean_region_max_num_intervals", Option::TYPE_INT, Option::LEVEL_ADVANCED)
+ .set_default(10)
+ .set_description("number of intervals in clean_offsets")
+ .set_long_description("partial recovery uses multiple intervals to record the clean part of the object"
+ "when the number of intervals is greater than osd_object_clean_region_max_num_intervals, minimum interval will be trimmed"
+ "(0 will recovery the entire object data interval)")
+ .add_service("osd"),
+
Option("osd_force_recovery_pg_log_entries_factor", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
.set_default(1.3)
.set_description(""),
.set_default(5)
.set_description(""),
- Option("osd_verify_sparse_read_holes", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
- .set_default(false)
- .set_description(""),
-
Option("osd_backoff_on_unfound", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(true)
.set_description(""),
Option("osd_debug_no_purge_strays", Option::TYPE_BOOL, Option::LEVEL_DEV)
.set_default(false),
+ Option("osd_debug_pretend_recovery_active", Option::TYPE_BOOL, Option::LEVEL_DEV)
+ .set_default(false)
+ .set_description(""),
+
Option("osd_enable_op_tracker", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(true)
.set_description(""),
.set_default(.97)
.set_description(""),
+ Option("osd_fast_shutdown", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(true)
+ .set_description("Fast, immediate shutdown")
+ .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."),
+
Option("osd_fast_fail_on_connection_refused", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(true)
.set_description(""),
.set_default(false)
.set_description(""),
- Option("kinetic_host", Option::TYPE_STR, Option::LEVEL_ADVANCED)
- .set_default("")
- .set_description(""),
-
- Option("kinetic_port", Option::TYPE_INT, Option::LEVEL_ADVANCED)
- .set_default(8123)
- .set_description(""),
-
- Option("kinetic_user_id", Option::TYPE_INT, Option::LEVEL_ADVANCED)
- .set_default(1)
- .set_description(""),
-
- Option("kinetic_hmac_key", Option::TYPE_STR, Option::LEVEL_ADVANCED)
- .set_default("asdfasdf")
- .set_description(""),
-
- Option("kinetic_use_ssl", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
- .set_default(false)
- .set_description(""),
-
Option("rocksdb_log_to_ceph_log", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(true)
.set_description(""),
Option("rocksdb_cache_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
.set_default(512_M)
+ .set_flag(Option::FLAG_RUNTIME)
.set_description(""),
Option("rocksdb_cache_row_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
.set_default(false)
.set_description(""),
- Option("rocksdb_enable_rmrange", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
- .set_default(true)
- .set_description("Refer to github.com/facebook/rocksdb/wiki/DeleteRange-Implementation"),
-
- Option("rocksdb_max_items_rmrange", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
- .set_default(1024)
- .set_description("Delete Range will be called if number of keys exceeded, must enable rocksdb_enable_rmrange first")
- .add_see_also("rocksdb_enable_rmrange"),
+ Option("rocksdb_delete_range_threshold", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(1048576)
+ .set_description("The number of keys required to invoke DeleteRange when deleting muliple keys."),
Option("rocksdb_bloom_bits_per_key", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
.set_default(20)
.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."),
Option("rocksdb_pin_l0_filter_and_index_blocks_in_cache", Option::TYPE_BOOL, Option::LEVEL_DEV)
- .set_default(true)
+ .set_default(false)
.set_description("Whether to pin Level 0 indices and bloom filters in the block cache")
.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."),
.set_description(""),
Option("osd_max_omap_entries_per_request", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
- .set_default(131072)
+ .set_default(1024)
.set_description(""),
Option("osd_max_omap_bytes_per_request", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
.set_default(1_G)
.set_description(""),
+ Option("osd_max_write_op_reply_len", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
+ .set_default(32)
+ .set_description("Max size of the per-op payload for requests with the RETURNVEC flag set")
+ .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."),
+
Option("osd_objectstore", Option::TYPE_STR, Option::LEVEL_ADVANCED)
.set_default("bluestore")
.set_enum_allowed({"bluestore", "filestore", "memstore", "kstore"})
Option("osd_memory_target", Option::TYPE_SIZE, Option::LEVEL_BASIC)
.set_default(4_G)
+ .set_min(896_M)
+ .set_flag(Option::FLAG_RUNTIME)
.add_see_also("bluestore_cache_autotune")
- .set_description("When tcmalloc and cache autotuning is enabled, try to keep this many bytes mapped in memory."),
+ .add_see_also("osd_memory_cache_min")
+ .add_see_also("osd_memory_base")
+ .set_description("When tcmalloc and cache autotuning is enabled, try to keep this many bytes mapped in memory.")
+ .set_long_description("The minimum value must be at least equal to osd_memory_base + osd_memory_cache_min."),
Option("osd_memory_target_cgroup_limit_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
.set_default(0.8)
Option("osd_memory_base", Option::TYPE_SIZE, Option::LEVEL_DEV)
.set_default(768_M)
+ .set_flag(Option::FLAG_RUNTIME)
.add_see_also("bluestore_cache_autotune")
.set_description("When tcmalloc and cache autotuning is enabled, estimate the minimum amount of memory in bytes the OSD will need."),
Option("osd_memory_expected_fragmentation", Option::TYPE_FLOAT, Option::LEVEL_DEV)
.set_default(0.15)
.set_min_max(0.0, 1.0)
+ .set_flag(Option::FLAG_RUNTIME)
.add_see_also("bluestore_cache_autotune")
.set_description("When tcmalloc and cache autotuning is enabled, estimate the percent of memory fragmentation."),
Option("osd_memory_cache_min", Option::TYPE_SIZE, Option::LEVEL_DEV)
.set_default(128_M)
+ .set_min(128_M)
+ .set_flag(Option::FLAG_RUNTIME)
.add_see_also("bluestore_cache_autotune")
.set_description("When tcmalloc and cache autotuning is enabled, set the minimum amount of memory used for caches."),
.set_default(64_K)
.set_description(""),
+ Option("memstore_debug_omit_block_device_write", Option::TYPE_BOOL, Option::LEVEL_DEV)
+ .set_default(false)
+ .add_see_also("bluestore_debug_omit_block_device_write")
+ .set_description("write metadata only"),
+
Option("objectstore_blackhole", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(false)
.set_description(""),
Option("bluefs_alloc_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
.set_default(1_M)
- .set_description(""),
+ .set_description("Allocation unit size for DB and WAL devices"),
+
+ Option("bluefs_shared_alloc_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
+ .set_default(64_K)
+ .set_description("Allocation unit size for primary/shared device"),
Option("bluefs_max_prefetch", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
.set_default(1_M)
.set_description(""),
Option("bluefs_buffered_io", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
- .set_default(true)
- .set_description(""),
+ .set_default(false)
+ .set_description("Enabled buffered IO for bluefs reads.")
+ .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."),
Option("bluefs_sync_write", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(false)
Option("bluefs_preextend_wal_files", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(false)
- .set_description(""),
+ .set_description("Preextent rocksdb wal files on mkfs to avoid performance penalty"),
+
+ Option("bluefs_log_replay_check_allocations", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(true)
+ .set_description("Enables checks for allocations consistency during log replay"),
Option("bluestore_bluefs", Option::TYPE_BOOL, Option::LEVEL_DEV)
.set_default(true)
.set_default(1_G)
.set_description("minimum free space allocated to BlueFS"),
+ Option("bluestore_bluefs_max_free", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
+ .set_default(10_G)
+ .set_description("Maximum free space allocated to BlueFS"),
+
Option("bluestore_bluefs_min_ratio", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
.set_default(.02)
.set_description("Minimum fraction of free space devoted to BlueFS"),
.set_description("Path to block device/file"),
Option("bluestore_block_size", Option::TYPE_SIZE, Option::LEVEL_DEV)
- .set_default(10_G)
+ .set_default(100_G)
.set_flag(Option::FLAG_CREATE)
.set_description("Size of file to create for backing bluestore"),
.add_see_also("bluestore_min_alloc_size"),
Option("bluestore_min_alloc_size_ssd", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
- .set_default(16_K)
+ .set_default(4_K)
.set_flag(Option::FLAG_CREATE)
.set_description("Default min_alloc_size value for non-rotational (solid state) media")
.add_see_also("bluestore_min_alloc_size"),
.add_see_also("bluestore_cache_autotune")
.set_description("The number of seconds to wait between rebalances when cache autotune is enabled."),
+ Option("bluestore_alloc_stats_dump_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
+ .set_default(3600 * 24)
+ .set_description("The period (in second) for logging allocation statistics."),
+
Option("bluestore_kvbackend", Option::TYPE_STR, Option::LEVEL_DEV)
.set_default("rocksdb")
.set_flag(Option::FLAG_CREATE)
Option("bluestore_allocator", Option::TYPE_STR, Option::LEVEL_ADVANCED)
.set_default("bitmap")
- .set_enum_allowed({"bitmap", "stupid"})
+ .set_enum_allowed({"bitmap", "stupid", "avl"})
.set_description("Allocator policy")
.set_long_description("Allocator to use for bluestore. Stupid should only be used for testing."),
.set_default(32)
.set_description("Max transactions with deferred writes that can accumulate before we force flush deferred writes"),
+ Option("bluestore_max_defer_interval", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
+ .set_default(3)
+ .set_description("max duration to force deferred submit"),
+
Option("bluestore_rocksdb_options", Option::TYPE_STR, Option::LEVEL_ADVANCED)
.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")
.set_description("Rocksdb options"),
.set_default(false)
.set_description("Run deep fsck at mount when bluestore_fsck_on_mount is set to true"),
+ Option("bluestore_fsck_quick_fix_on_mount", Option::TYPE_BOOL, Option::LEVEL_DEV)
+ .set_default(true)
+ .set_description("Do quick-fix for the store at mount"),
+
Option("bluestore_fsck_on_umount", Option::TYPE_BOOL, Option::LEVEL_DEV)
.set_default(false)
.set_description("Run fsck at umount"),
.set_flag(Option::FLAG_RUNTIME)
.set_description("Maximum bytes read at once by deep fsck"),
+ Option("bluestore_fsck_quick_fix_threads", Option::TYPE_INT, Option::LEVEL_ADVANCED)
+ .set_default(2)
+ .set_description("Number of additional threads to perform quick-fix (shallow fsck) command"),
+
Option("bluestore_throttle_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
.set_default(64_M)
.set_flag(Option::FLAG_RUNTIME)
.set_default(0)
.set_description(""),
+ Option("bluestore_debug_too_many_blobs_threshold", Option::TYPE_INT, Option::LEVEL_DEV)
+ .set_default(24*1024)
+ .set_description(""),
+
Option("bluestore_debug_freelist", Option::TYPE_BOOL, Option::LEVEL_DEV)
.set_default(false)
.set_description(""),
.set_default(0.0)
.set_description("inject crc verification errors into bluestore device reads"),
- Option("bluestore_no_per_pool_stats_tolerance", Option::TYPE_STR, Option::LEVEL_ADVANCED)
- .set_default("until_repair")
- .set_flag(Option::FLAG_RUNTIME)
- .set_enum_allowed({"enforce", "until_repair", "until_fsck"})
- .set_description("Specified how to treat lack of per-pool stats, e.g. caused by an upgrade")
- .set_long_description(
- "'until_fsck' will tolerate the case for regular ops and fail on fsck or repair, the latter will fix the issue, "
- "'until_repair' will tolerate for regular ops and fsck. Repair indicates and fixes the issue, "
- "'enforce' will unconditionally use global stats mode."),
+ Option("bluestore_fsck_error_on_no_per_pool_stats", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(false)
+ .set_description("Make fsck error (instead of warn) when bluestore lacks per-pool stats, e.g., after an upgrade"),
Option("bluestore_warn_on_bluefs_spillover", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(true)
.set_default(true)
.set_description("Enable health indication on lack of per-pool statfs reporting from bluestore"),
+ Option("bluestore_fsck_error_on_no_per_pool_omap", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(false)
+ .set_description("Make fsck error (instead of warn) when objects without per-pool omap are found"),
+
+ Option("bluestore_warn_on_no_per_pool_omap", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(true)
+ .set_description("Enable health indication on lack of per-pool omap"),
+
Option("bluestore_log_op_age", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
.set_default(5)
.set_description("log operation if it's slower than this age (seconds)"),
.set_default(60)
.set_description("log collection list operation if it's slower than this age (seconds)"),
+ Option("bluestore_debug_enforce_settings", Option::TYPE_STR, Option::LEVEL_DEV)
+ .set_default("default")
+ .set_enum_allowed({"default", "hdd", "ssd"})
+ .set_description("Enforces specific hw profile settings")
+ .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."),
+
+ Option("bluestore_avl_alloc_bf_threshold", Option::TYPE_UINT, Option::LEVEL_DEV)
+ .set_default(131072)
+ .set_description(""),
+
+ Option("bluestore_avl_alloc_bf_free_pct", Option::TYPE_UINT, Option::LEVEL_DEV)
+ .set_default(4)
+ .set_description(""),
+
+ Option("bluestore_volume_selection_policy", Option::TYPE_STR, Option::LEVEL_DEV)
+ .set_default("rocksdb_original")
+ .set_enum_allowed({ "rocksdb_original", "use_some_extra" })
+ .set_description("Determines bluefs volume selection policy")
+ .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"),
+
+ Option("bluestore_volume_selection_reserved_factor", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
+ .set_flag(Option::FLAG_STARTUP)
+ .set_default(2.0)
+ .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"),
+
+ Option("bluestore_volume_selection_reserved", Option::TYPE_INT, Option::LEVEL_ADVANCED)
+ .set_flag(Option::FLAG_STARTUP)
+ .set_default(0)
+ .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."),
+
+ Option("bluestore_ioring", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(false)
+ .set_description("Enables Linux io_uring API instead of libaio"),
+
// -----------------------------------------
// kstore
.set_default(0)
.set_description(""),
- Option("mgr_stats_threshold", Option::TYPE_INT, Option::LEVEL_ADVANCED)
- .set_default((int64_t)PerfCountersBuilder::PRIO_USEFUL)
- .set_description("Lowest perfcounter priority collected by mgr")
- .set_long_description("Daemons only set perf counter data to the manager "
- "daemon if the counter has a priority higher than this.")
- .set_min_max((int64_t)PerfCountersBuilder::PRIO_DEBUGONLY,
- (int64_t)PerfCountersBuilder::PRIO_CRITICAL + 1),
+ Option("mgr_stats_threshold", Option::TYPE_INT, Option::LEVEL_ADVANCED)
+ .set_default((int64_t)PerfCountersBuilder::PRIO_USEFUL)
+ .set_description("Lowest perfcounter priority collected by mgr")
+ .set_long_description("Daemons only set perf counter data to the manager "
+ "daemon if the counter has a priority higher than this.")
+ .set_min_max((int64_t)PerfCountersBuilder::PRIO_DEBUGONLY,
+ (int64_t)PerfCountersBuilder::PRIO_CRITICAL + 1),
Option("journal_zero_on_create", Option::TYPE_BOOL, Option::LEVEL_DEV)
.set_default(false)
.set_default(false)
.set_description(""),
- Option("nss_db_path", Option::TYPE_STR, Option::LEVEL_ADVANCED)
- .set_default("")
- .set_description(""),
+ Option("cephadm_path", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_default("/usr/sbin/cephadm")
+ .add_service("mgr")
+ .set_description("Path to cephadm utility"),
Option("mgr_module_path", Option::TYPE_STR, Option::LEVEL_ADVANCED)
.set_default(CEPH_DATADIR "/mgr")
.set_default(false)
.set_description(""),
+ Option("bluestore_tracing", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(false)
+ .set_description("Enable bluestore event tracing."),
+
+ Option("bluestore_throttle_trace_rate", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
+ .set_default(0)
+ .set_description("Rate at which to sample bluestore transactions (per second)"),
+
Option("debug_deliberately_leak_memory", Option::TYPE_BOOL, Option::LEVEL_DEV)
.set_default(false)
.set_description(""),
.set_description("")
.add_service({"mon", "osd"})
.set_long_description("This sets the gss target service name."),
+
+ Option("debug_disable_randomized_ping", Option::TYPE_BOOL, Option::LEVEL_DEV)
+ .set_default(false)
+ .set_description("Disable heartbeat ping randomization for testing purposes"),
+
+ Option("debug_heartbeat_testing_span", Option::TYPE_INT, Option::LEVEL_DEV)
+ .set_default(0)
+ .set_description("Override 60 second periods for testing only"),
+
+ // ----------------------------
+ // Crimson specific options
+
+ Option("crimson_osd_obc_lru_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(10)
+ .set_description("Number of obcs to cache")
+
});
}
Option("rgw_override_bucket_index_max_shards", Option::TYPE_UINT, Option::LEVEL_DEV)
.set_default(0)
- .set_description(""),
+ .set_description("The default number of bucket index shards for newly-created "
+ "buckets. This value overrides bucket_index_max_shards stored in the zone. "
+ "Setting this value in the zone is preferred, because it applies globally "
+ "to all radosgw daemons running in the zone."),
Option("rgw_bucket_index_max_aio", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
- .set_default(8)
+ .set_default(128)
.set_description("Max number of concurrent RADOS requests when handling bucket shards."),
Option("rgw_enable_quota_threads", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
"will be located in the path that is specified here. "),
Option("rgw_enable_apis", Option::TYPE_STR, Option::LEVEL_ADVANCED)
- .set_default("s3, s3website, swift, swift_auth, admin, sts")
+ .set_default("s3, s3website, swift, swift_auth, admin, sts, iam, pubsub")
.set_description("A list of set of RESTful APIs that rgw handles."),
Option("rgw_cache_enabled", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_long_description(
"This is needed for virtual hosting of buckets, unless configured via zonegroup "
"configuration."),
+
+ Option("rgw_numa_node", Option::TYPE_INT, Option::LEVEL_ADVANCED)
+ .set_default(-1)
+ .set_flag(Option::FLAG_STARTUP)
+ .set_description("set rgw's cpu affinity to a numa node (-1 for none)"),
Option("rgw_service_provider_name", Option::TYPE_STR, Option::LEVEL_ADVANCED)
.set_default("")
"Max number of Keystone tokens that will be cached. Token that is not cached "
"requires RGW to access the Keystone server when authenticating."),
- Option("rgw_keystone_revocation_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
- .set_default(15_min)
- .set_description("Keystone cache revocation interval")
- .set_long_description(
- "Time (in seconds) that RGW waits between requests to Keystone for getting a list "
- "of revoked tokens. A revoked token might still be considered valid by RGW for "
- "this amount of time."),
-
Option("rgw_keystone_verify_ssl", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(true)
.set_description("Should RGW verify the Keystone server SSL certificate."),
- Option("rgw_keystone_implicit_tenants", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
- .set_default(false)
+ Option("rgw_keystone_implicit_tenants", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_default("false")
+ .set_enum_allowed( { "false", "true", "swift", "s3", "both", "0", "1", "none" } )
.set_description("RGW Keystone implicit tenants creation")
.set_long_description(
"Implicitly create new users in their own tenant with the same name when "
- "authenticating via Keystone."),
+ "authenticating via Keystone. Can be limited to s3 or swift only."),
Option("rgw_cross_domain_policy", Option::TYPE_STR, Option::LEVEL_ADVANCED)
.set_default("<allow-access-from domain=\"*\" secure=\"false\" />")
.set_default(10)
.set_description(""),
+ Option("rgw_nfs_s3_fast_attrs", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(false)
+ .set_description("use fast S3 attrs from bucket index (immutable only)")
+ .set_long_description("use fast S3 attrs from bucket index (assumes NFS "
+ "mounts are immutable)"),
+
Option("rgw_rados_pool_autoscale_bias", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
.set_default(4.0)
.set_min_max(0.01, 100000.0)
.set_min_max(1, 1024)
.set_description("pg_num_min value for RGW metadata (omap-heavy) pools"),
+ Option("rgw_rados_pool_recovery_priority", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(5)
+ .set_min_max(-10, 10)
+ .set_description("recovery_priority value for RGW metadata (omap-heavy) pools"),
+
Option("rgw_zone", Option::TYPE_STR, Option::LEVEL_ADVANCED)
.set_default("")
.set_description("Zone name")
"The length of time (in seconds) that the RGW collector will wait before purging "
"a deleted object's data. RGW will not remove object immediately, as object could "
"still have readers. A mechanism exists to increase the object's expiration time "
- "when it's being read.")
+ "when it's being read. The recommended value of its lower limit is 30 minutes")
.add_see_also({"rgw_gc_max_objs", "rgw_gc_processor_max_time", "rgw_gc_processor_period", "rgw_gc_max_concurrent_io"}),
Option("rgw_gc_processor_max_time", Option::TYPE_INT, Option::LEVEL_ADVANCED)
.set_description("Max number of keys to remove from garbage collector log in a single operation")
.add_see_also({"rgw_gc_max_objs", "rgw_gc_obj_min_wait", "rgw_gc_processor_max_time", "rgw_gc_max_concurrent_io"}),
+ Option("rgw_gc_max_deferred_entries_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(3072)
+ .set_description("maximum allowed size of deferred entries in queue head for gc"),
+
+ Option("rgw_gc_max_queue_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(134213632)
+ .set_description("Maximum allowed queue size for gc")
+ .set_long_description(
+ "The maximum allowed size of each gc queue, and its value should not "
+ "be greater than (osd_max_object_size - rgw_gc_max_deferred_entries_size - 1K).")
+ .add_see_also({"osd_max_object_size", "rgw_gc_max_deferred_entries_size"}),
+
+ Option("rgw_gc_max_deferred", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(50)
+ .set_description("Number of maximum deferred data entries to be stored in queue for gc"),
+
Option("rgw_s3_success_create_obj_status", Option::TYPE_INT, Option::LEVEL_ADVANCED)
.set_default(0)
.set_description("HTTP return code override for object creation")
"the type of the frontend followed by an optional space delimited set of "
"key=value config parameters."),
+ Option("rgw_frontend_defaults", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_default("beast ssl_certificate=config://rgw/cert/$realm/$zone.crt ssl_private_key=config://rgw/cert/$realm/$zone.key")
+ .set_description("RGW frontends default configuration")
+ .set_long_description(
+ "A comma delimited list of default frontends configuration."),
+
Option("rgw_user_quota_bucket_sync_interval", Option::TYPE_INT, Option::LEVEL_ADVANCED)
.set_default(180)
.set_description("User quota bucket sync interval")
.set_default(1000)
.set_description("Max number of buckets per user")
.set_long_description(
- "A user can create this many buckets. Zero means unlimmited, negative number means "
+ "A user can create this many buckets. Zero means unlimited, negative number means "
"user cannot create any buckets (although user will retain buckets already created."),
Option("rgw_objexp_gc_interval", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
.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.")
.add_tag("performance")
.add_service("rgw"),
-
+
Option("rgw_reshard_batch_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
.set_default(64)
.set_min(8)
.set_default("")
.set_description(""),
+ Option("rgw_crypt_s3_kms_backend", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_default("barbican")
+ .set_enum_allowed({"barbican", "vault", "testing"})
+ .set_description(
+ "Where the SSE-KMS encryption keys are stored. Supported KMS "
+ "systems are OpenStack Barbican ('barbican', the default) and HashiCorp "
+ "Vault ('vault')."),
+
Option("rgw_crypt_s3_kms_encryption_keys", Option::TYPE_STR, Option::LEVEL_DEV)
.set_default("")
.set_description(""),
+ Option("rgw_crypt_vault_auth", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_default("token")
+ .set_enum_allowed({"token", "agent"})
+ .set_description(
+ "Type of authentication method to be used with Vault. ")
+ .add_see_also({
+ "rgw_crypt_s3_kms_backend",
+ "rgw_crypt_vault_addr",
+ "rgw_crypt_vault_token_file"}),
+
+ Option("rgw_crypt_vault_token_file", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_default("")
+ .set_description(
+ "If authentication method is 'token', provide a path to the token file, "
+ "which for security reasons should readable only by Rados Gateway.")
+ .add_see_also({
+ "rgw_crypt_s3_kms_backend",
+ "rgw_crypt_vault_auth",
+ "rgw_crypt_vault_addr"}),
+
+ Option("rgw_crypt_vault_addr", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_default("")
+ .set_description("Vault server base address.")
+ .add_see_also({
+ "rgw_crypt_s3_kms_backend",
+ "rgw_crypt_vault_auth",
+ "rgw_crypt_vault_prefix"}),
+
+ Option("rgw_crypt_vault_prefix", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_default("")
+ .set_description("Vault secret URL prefix, which can be used to restrict "
+ "access to a particular subset of the Vault secret space.")
+ .add_see_also({
+ "rgw_crypt_s3_kms_backend",
+ "rgw_crypt_vault_addr",
+ "rgw_crypt_vault_auth"}),
+
+
+ Option("rgw_crypt_vault_secret_engine", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_enum_allowed({"kv", "transit"})
+ .set_default("transit")
+ .set_description(
+ "Vault Secret Engine to be used to retrieve encryption keys.")
+ .add_see_also({
+ "rgw_crypt_s3_kms_backend",
+ "rgw_crypt_vault_auth",
+ "rgw_crypt_vault_addr"}),
+
+ Option("rgw_crypt_vault_namespace", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_default("")
+ .set_description("Vault Namespace to be used to select your tenant")
+ .add_see_also({
+ "rgw_crypt_s3_kms_backend",
+ "rgw_crypt_vault_auth",
+ "rgw_crypt_vault_addr"}),
+
Option("rgw_crypt_suppress_logs", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(true)
.set_description("Suppress logs that might print client key"),
.set_long_description(
"If true, RGW will dynamicall increase the number of shards in buckets that have "
"a high number of objects per shard.")
- .add_see_also("rgw_max_objs_per_shard"),
+ .add_see_also("rgw_max_objs_per_shard")
+ .add_see_also("rgw_max_dynamic_shards"),
Option("rgw_max_objs_per_shard", Option::TYPE_UINT, Option::LEVEL_BASIC)
.set_default(100000)
"This is the max number of objects per bucket index shard that RGW will "
"allow with dynamic resharding. RGW will trigger an automatic reshard operation "
"on the bucket if it exceeds this number.")
- .add_see_also("rgw_dynamic_resharding"),
+ .add_see_also("rgw_dynamic_resharding")
+ .add_see_also("rgw_max_dynamic_shards"),
+
+ Option("rgw_max_dynamic_shards", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(1999)
+ .set_min(1)
+ .set_description("Max shards that dynamic resharding can create")
+ .set_long_description(
+ "This is the maximum number of bucket index shards that dynamic "
+ "sharding is able to create on its own. This does not limit user "
+ "requested resharding. Ideally this value is a prime number.")
+ .add_see_also("rgw_dynamic_resharding")
+ .add_see_also("rgw_max_objs_per_shard"),
Option("rgw_reshard_thread_interval", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
.set_default(10_min)
- .set_min(10_min)
+ .set_min(10)
.set_description("Number of seconds between processing of reshard log entries"),
Option("rgw_cache_expiry_interval", Option::TYPE_UINT,
.set_long_description("This caps the maximum permitted value for listing-like operations in RGW S3. "
"Affects ListBucket(max-keys), "
"ListBucketVersions(max-keys), "
- "ListBucketMultiPartUploads(max-uploads), "
+ "ListBucketMultipartUploads(max-uploads), "
"ListMultipartUploadParts(max-parts)"),
Option("rgw_sts_token_introspection_url", Option::TYPE_STR, Option::LEVEL_ADVANCED)
.set_description("mclock limit for metadata requests")
.add_see_also("rgw_dmclock_metadata_res")
.add_see_also("rgw_dmclock_metadata_wgt"),
-
});
}
.set_default(60)
.set_description("time in seconds for detecting a hung thread"),
+ Option("rbd_disable_zero_copy_writes", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(true)
+ .set_description("Disable the use of zero-copy writes to ensure unstable "
+ "writes from clients cannot cause a CRC mismatch"),
+
Option("rbd_non_blocking_aio", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(true)
.set_description("process AIO ops from a dispatch thread to prevent blocking"),
.set_default(true)
.set_description("whether to enable caching (writeback unless rbd_cache_max_dirty is 0)"),
+ Option("rbd_cache_policy", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_enum_allowed({"writethrough", "writeback", "writearound"})
+ .set_default("writearound")
+ .set_description("cache policy for handling writes."),
+
Option("rbd_cache_writethrough_until_flush", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(true)
.set_description("whether to make writeback caching writethrough until "
.set_default(false)
.set_description("whether to block writes to the cache before the aio_write call completes"),
+ Option("rbd_parent_cache_enabled", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(false)
+ .set_description("whether to enable rbd shared ro cache"),
+
Option("rbd_concurrent_management_ops", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
.set_default(10)
.set_min(1)
.set_default(true)
.set_description("when writing a object, it will issue a hint to osd backend to indicate the expected size object need"),
+ Option("rbd_compression_hint", Option::TYPE_STR, Option::LEVEL_BASIC)
+ .set_enum_allowed({"none", "compressible", "incompressible"})
+ .set_default("none")
+ .set_description("Compression hint to send to the OSDs during writes")
+ .set_flag(Option::FLAG_RUNTIME),
+
+ Option("rbd_read_from_replica_policy", Option::TYPE_STR, Option::LEVEL_BASIC)
+ .set_enum_allowed({"default", "balance", "localize"})
+ .set_default("default")
+ .set_description("Read replica policy send to the OSDS during reads")
+ .set_flag(Option::FLAG_RUNTIME),
+
Option("rbd_tracing", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(false)
.set_description("true if LTTng-UST tracepoints should be enabled"),
.set_default(0)
.set_description("default number of seconds to protect deleted images in the trash"),
+ Option("rbd_move_parent_to_trash_on_remove", Option::TYPE_BOOL, Option::LEVEL_BASIC)
+ .set_default(false)
+ .set_description("move parent with clone format v2 children to the trash when deleted"),
+
Option("rbd_mirroring_resync_after_disconnect", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(false)
.set_description("automatically start image resync after mirroring is disconnected due to being laggy"),
.set_default(0)
.set_description("time-delay in seconds for rbd-mirror asynchronous replication"),
+ Option("rbd_mirroring_max_mirroring_snapshots", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(3)
+ .set_min(3)
+ .set_description("mirroring snapshots limit"),
+
Option("rbd_default_format", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
.set_default(2)
.set_description("default image format for new images"),
.set_default(60)
.set_min(0)
.set_description("RBD Image access timestamp refresh interval. Set to 0 to disable access timestamp update."),
+
+ Option("rbd_io_scheduler", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_default("simple")
+ .set_enum_allowed({"none", "simple"})
+ .set_description("RBD IO scheduler"),
+
+ Option("rbd_io_scheduler_simple_max_delay", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(0)
+ .set_min(0)
+ .set_description("maximum io delay (in milliseconds) for simple io scheduler (if set to 0 dalay is calculated based on latency stats)"),
+
+ Option("rbd_rwl_enabled", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(false)
+ .set_description("enable persistent write back cache for this volume"),
+
+ Option("rbd_rwl_log_periodic_stats", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(false)
+ .set_description("emit periodic perf stats to debug log"),
+
+ Option("rbd_rwl_size", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(1073741824)
+ .set_min(1073741824)
+ .set_description("size of the persistent write back cache for this volume"),
+
+ Option("rbd_rwl_path", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_default("/tmp")
+ .set_description("location of the persistent write back cache in a DAX-enabled filesystem on persistent memory"),
});
}
.set_default(5)
.set_description("maximum age (in seconds) between successive journal polls"),
- Option("rbd_mirror_journal_max_fetch_bytes", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
- .set_default(32768)
- .set_description("maximum bytes to read from each journal data object per fetch"),
-
Option("rbd_mirror_sync_point_update_age", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
.set_default(30)
.set_description("number of seconds between each update of the image sync point object number"),
"mgr_stats_threshold.")
.set_min_max((int64_t)PerfCountersBuilder::PRIO_DEBUGONLY,
(int64_t)PerfCountersBuilder::PRIO_CRITICAL + 1),
+
+ Option("rbd_mirror_image_perf_stats_prio", Option::TYPE_INT, Option::LEVEL_ADVANCED)
+ .set_default((int64_t)PerfCountersBuilder::PRIO_USEFUL)
+ .set_description("Priority level for mirror daemon per-image replication perf counters")
+ .set_long_description("The daemon will send per-image perf counter data to the "
+ "manager daemon if the priority is not lower than "
+ "mgr_stats_threshold.")
+ .set_min_max((int64_t)PerfCountersBuilder::PRIO_DEBUGONLY,
+ (int64_t)PerfCountersBuilder::PRIO_CRITICAL + 1),
+
+ Option("rbd_mirror_memory_autotune", Option::TYPE_BOOL, Option::LEVEL_DEV)
+ .set_default(true)
+ .add_see_also("rbd_mirror_memory_target")
+ .set_description("Automatically tune the ratio of caches while respecting min values."),
+
+ Option("rbd_mirror_memory_target", Option::TYPE_SIZE, Option::LEVEL_BASIC)
+ .set_default(4_G)
+ .add_see_also("rbd_mirror_memory_autotune")
+ .set_description("When tcmalloc and cache autotuning is enabled, try to keep this many bytes mapped in memory."),
+
+ Option("rbd_mirror_memory_base", Option::TYPE_SIZE, Option::LEVEL_DEV)
+ .set_default(768_M)
+ .add_see_also("rbd_mirror_memory_autotune")
+ .set_description("When tcmalloc and cache autotuning is enabled, estimate the minimum amount of memory in bytes the rbd-mirror daemon will need."),
+
+ Option("rbd_mirror_memory_expected_fragmentation", Option::TYPE_FLOAT, Option::LEVEL_DEV)
+ .set_default(0.15)
+ .set_min_max(0.0, 1.0)
+ .add_see_also("rbd_mirror_memory_autotune")
+ .set_description("When tcmalloc and cache autotuning is enabled, estimate the percent of memory fragmentation."),
+
+ Option("rbd_mirror_memory_cache_min", Option::TYPE_SIZE, Option::LEVEL_DEV)
+ .set_default(128_M)
+ .add_see_also("rbd_mirror_memory_autotune")
+ .set_description("When tcmalloc and cache autotuning is enabled, set the minimum amount of memory used for cache."),
+
+ Option("rbd_mirror_memory_cache_resize_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
+ .set_default(5)
+ .add_see_also("rbd_mirror_memory_autotune")
+ .set_description("When tcmalloc and cache autotuning is enabled, wait this many seconds between resizing caches."),
+
+ Option("rbd_mirror_memory_cache_autotune_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
+ .set_default(30)
+ .add_see_also("rbd_mirror_memory_autotune")
+ .set_description("The number of seconds to wait between rebalances when cache autotune is enabled."),
+ });
+}
+
+static std::vector<Option> get_immutable_object_cache_options() {
+ return std::vector<Option>({
+ Option("immutable_object_cache_path", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_default("/tmp")
+ .set_description("immutable object cache data dir"),
+
+ Option("immutable_object_cache_sock", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_default("/var/run/ceph/immutable_object_cache_sock")
+ .set_description("immutable object cache domain socket"),
+
+ Option("immutable_object_cache_max_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
+ .set_default(1_G)
+ .set_description("max immutable object cache data size"),
+
+ Option("immutable_object_cache_max_inflight_ops", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(128)
+ .set_description("max inflight promoting requests for immutable object cache daemon"),
+
+ Option("immutable_object_cache_client_dedicated_thread_num", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(2)
+ .set_description("immutable object cache client dedicated thread number"),
+
+ Option("immutable_object_cache_watermark", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
+ .set_default(0.1)
+ .set_description("immutable object cache water mark"),
});
}
std::vector<Option> get_mds_options() {
return std::vector<Option>({
+ Option("mds_numa_node", Option::TYPE_INT, Option::LEVEL_ADVANCED)
+ .set_default(-1)
+ .set_flag(Option::FLAG_STARTUP)
+ .set_description("set mds's cpu affinity to a numa node (-1 for none)"),
+
Option("mds_data", Option::TYPE_STR, Option::LEVEL_ADVANCED)
.set_default("/var/lib/ceph/mds/$cluster-$id")
.set_flag(Option::FLAG_NO_MON_UPDATE)
.set_description("path to MDS data and keyring"),
+ Option("mds_join_fs", Option::TYPE_STR, Option::LEVEL_BASIC)
+ .set_default("")
+ .set_description("file system MDS prefers to join")
+ .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.")
+ .set_flag(Option::FLAG_RUNTIME),
+
Option("mds_max_xattr_pairs_size", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
.set_default(64_K)
.set_description("maximum aggregate size of extended attributes on a file"),
- Option("mds_cache_size", Option::TYPE_INT, Option::LEVEL_ADVANCED)
- .set_default(0)
- .set_description("maximum number of inodes in MDS cache (<=0 is unlimited)")
- .set_long_description("This tunable is no longer recommended. Use mds_cache_memory_limit."),
+ Option("mds_cache_trim_interval", Option::TYPE_SECS, Option::LEVEL_ADVANCED)
+ .set_default(1)
+ .set_description("interval in seconds between cache trimming")
+ .set_flag(Option::FLAG_RUNTIME),
+
+ Option("mds_cache_release_free_interval", Option::TYPE_SECS, Option::LEVEL_DEV)
+ .set_default(10)
+ .set_description("interval in seconds between heap releases")
+ .set_flag(Option::FLAG_RUNTIME),
Option("mds_cache_memory_limit", Option::TYPE_SIZE, Option::LEVEL_BASIC)
- .set_default(1*(1LL<<30))
+ .set_default(4_G)
.set_description("target maximum memory usage of MDS cache")
+ .set_flag(Option::FLAG_RUNTIME)
.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."),
Option("mds_cache_reservation", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
.set_default(.05)
- .set_description("amount of memory to reserve for future cached objects"),
+ .set_description("amount of memory to reserve for future cached objects")
+ .set_flag(Option::FLAG_RUNTIME),
Option("mds_health_cache_threshold", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
.set_default(1.5)
Option("mds_cache_trim_decay_rate", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
.set_default(1)
- .set_description("decay rate for trimming MDS cache throttle"),
+ .set_description("decay rate for trimming MDS cache throttle")
+ .set_flag(Option::FLAG_RUNTIME),
Option("mds_cache_trim_threshold", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
.set_default(64_K)
- .set_description("threshold for number of dentries that can be trimmed"),
+ .set_description("threshold for number of dentries that can be trimmed")
+ .set_flag(Option::FLAG_RUNTIME),
Option("mds_max_file_recover", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
.set_default(32)
Option("mds_recall_max_caps", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
.set_default(5000)
- .set_description("maximum number of caps to recall from client session in single recall"),
+ .set_description("maximum number of caps to recall from client session in single recall")
+ .set_flag(Option::FLAG_RUNTIME),
Option("mds_recall_max_decay_rate", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
.set_default(2.5)
- .set_description("decay rate for throttle on recalled caps on a session"),
+ .set_description("decay rate for throttle on recalled caps on a session")
+ .set_flag(Option::FLAG_RUNTIME),
Option("mds_recall_max_decay_threshold", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
.set_default(16_K)
- .set_description("decay threshold for throttle on recalled caps on a session"),
+ .set_description("decay threshold for throttle on recalled caps on a session")
+ .set_flag(Option::FLAG_RUNTIME),
Option("mds_recall_global_max_decay_threshold", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
.set_default(64_K)
- .set_description("decay threshold for throttle on recalled caps globally"),
+ .set_description("decay threshold for throttle on recalled caps globally")
+ .set_flag(Option::FLAG_RUNTIME),
Option("mds_recall_warning_threshold", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
.set_default(32_K)
- .set_description("decay threshold for warning on slow session cap recall"),
+ .set_description("decay threshold for warning on slow session cap recall")
+ .set_flag(Option::FLAG_RUNTIME),
Option("mds_recall_warning_decay_rate", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
.set_default(60.0)
- .set_description("decay rate for warning on slow session cap recall"),
+ .set_description("decay rate for warning on slow session cap recall")
+ .set_flag(Option::FLAG_RUNTIME),
+
+ Option("mds_session_cache_liveness_decay_rate", Option::TYPE_FLOAT, Option::LEVEL_ADVANCED)
+ .add_see_also("mds_session_cache_liveness_magnitude")
+ .set_default(5_min)
+ .set_description("decay rate for session liveness leading to preemptive cap recall")
+ .set_flag(Option::FLAG_RUNTIME)
+ .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."),
+
+ Option("mds_session_cache_liveness_magnitude", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
+ .add_see_also("mds_session_cache_liveness_decay_rate")
+ .set_default(10)
+ .set_description("decay magnitude for preemptively recalling caps on quiet client")
+ .set_flag(Option::FLAG_RUNTIME)
+ .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."),
Option("mds_freeze_tree_timeout", Option::TYPE_FLOAT, Option::LEVEL_DEV)
.set_default(30)
.set_default(1000)
.set_description("number of unused inodes to pre-allocate to clients for file creation"),
+ Option("mds_client_delegate_inos_pct", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(50)
+ .set_flag(Option::FLAG_RUNTIME)
+ .set_description("percentage of preallocated inos to delegate to client"),
+
Option("mds_early_reply", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(true)
.set_description("additional reply to clients that metadata requests are complete but not yet durable"),
+ Option("mds_replay_unsafe_with_closed_session", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(false)
+ .set_flag(Option::FLAG_RUNTIME)
+ .set_description("complete all the replay request when mds is restarted, no matter the session is closed or not"),
+
Option("mds_default_dir_hash", Option::TYPE_INT, Option::LEVEL_ADVANCED)
.set_default(CEPH_STR_HASH_RJENKINS)
.set_description("hash function to select directory fragment for dentry name"),
.set_default(5)
.set_description("maximum number of scrub operations performed in parallel"),
+ Option("mds_forward_all_requests_to_auth", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
+ .set_default(false)
+ .set_flag(Option::FLAG_RUNTIME)
+ .set_description("always process op on auth mds"),
+
Option("mds_damage_table_max_entries", Option::TYPE_INT, Option::LEVEL_ADVANCED)
.set_default(10000)
.set_description("maximum number of damage table entries"),
.set_default(0)
.set_description("threshold for cache usage to disallow \"dump cache\" operation to file")
.set_long_description("Disallow MDS from dumping caches to file via \"dump cache\" command if cache usage exceeds this threshold."),
+
+ Option("mds_task_status_update_interval", Option::TYPE_FLOAT, Option::LEVEL_DEV)
+ .set_default(2.0)
+ .set_description("task status update interval to manager")
+ .set_long_description("interval (in seconds) for sending mds task status to ceph manager"),
+
+ Option("mds_max_snaps_per_dir", Option::TYPE_UINT, Option::LEVEL_ADVANCED)
+ .set_default(100)
+ .set_min_max(0, 4096)
+ .set_flag(Option::FLAG_RUNTIME)
+ .set_description("max snapshots per directory")
+ .set_long_description("maximum number of snapshots that can be created per directory"),
});
}
Option("fuse_default_permissions", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
.set_default(false)
- .set_description("pass default_permisions to FUSE on mount"),
+ .set_description("pass default_permisions to FUSE on mount")
+ .set_flag(Option::FLAG_STARTUP),
Option("fuse_big_writes", Option::TYPE_BOOL, Option::LEVEL_ADVANCED)
- .set_default(false)
+ .set_default(true)
.set_description("big_writes is deprecated in libfuse 3.0.0"),
Option("fuse_max_write", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
.set_default(false)
.set_description(""),
- Option("client_mds_namespace", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ Option("client_fs", Option::TYPE_STR, Option::LEVEL_ADVANCED)
+ .set_flag(Option::FLAG_STARTUP)
.set_default("")
-
.set_description("CephFS file system name to mount")
.set_long_description("Use this with ceph-fuse, or with any process "
"that uses libcephfs. Programs using libcephfs may also pass "
"If no filesystem name is given in mount() or this setting, the default "
"filesystem will be mounted (usually the first created)."),
+ /* Alias for client_fs. Deprecated */
+ Option("client_mds_namespace", Option::TYPE_STR, Option::LEVEL_DEV)
+ .set_flag(Option::FLAG_STARTUP)
+ .set_default(""),
+
Option("fake_statfs_for_testing", Option::TYPE_INT, Option::LEVEL_DEV)
.set_default(0)
.set_description("Set a value for kb and compute kb_used from total of num_bytes"),
}
+std::vector<Option> get_cephfs_shell_options() {
+ return std::vector<Option>({
+ Option("allow_ansi", Option::TYPE_STR, Option::LEVEL_BASIC)
+ .set_default("Terminal")
+ .set_description("Allow ANSI escape sequences in output. Values: "
+ "Terminal, Always, Never"),
+
+ Option("colors", Option::TYPE_STR, Option::LEVEL_BASIC)
+ .set_default("Terminal")
+ .set_description("Colouring CephFS shell input and output. Values: "
+ "Terminal, Always, Never"),
+
+ Option("continuation_prompt", Option::TYPE_STR, Option::LEVEL_BASIC)
+ .set_default(">")
+ .set_description("Prompt string when a command continue to second line"),
+
+ Option("debug_shell", Option::TYPE_BOOL, Option::LEVEL_BASIC)
+ .set_default(false)
+ .set_description("Allow tracebacks on error for CephFS Shell"),
+
+ Option("echo", Option::TYPE_BOOL, Option::LEVEL_BASIC)
+ .set_default(false)
+ .set_description("Print command issued on prompt before execution"),
+
+ Option("editor", Option::TYPE_STR, Option::LEVEL_BASIC)
+ .set_default("vim")
+ .set_description("Default text editor for shell"),
+
+ Option("feedback_to_output", Option::TYPE_BOOL, Option::LEVEL_BASIC)
+ .set_default(false)
+ .set_description("include '|' and '>' in result"),
+
+ Option("max_completion_items", Option::TYPE_INT, Option::LEVEL_BASIC)
+ .set_default(50)
+ .set_description("Maximum number of items to be displayed by tab "
+ "completion"),
+
+ Option("prompt", Option::TYPE_STR, Option::LEVEL_BASIC)
+ .set_default("\x1b[01;33mCephFS:~\x1b[96m/\x1b[0m\x1b[01;33m>>>\x1b[00m ")
+ .set_description("Whether non-essential feedback should be printed."),
+
+ Option("quiet", Option::TYPE_BOOL, Option::LEVEL_BASIC)
+ .set_default(false)
+ .set_description("Whether non-essential feedback should be printed."),
+
+ Option("timing", Option::TYPE_BOOL, Option::LEVEL_BASIC)
+ .set_default(false)
+ .set_description("Whether execution time should be reported"),
+ });
+}
+
static std::vector<Option> build_options()
{
std::vector<Option> result = get_global_options();
ingest(get_rgw_options(), "rgw");
ingest(get_rbd_options(), "rbd");
ingest(get_rbd_mirror_options(), "rbd-mirror");
+ ingest(get_immutable_object_cache_options(), "immutable-objet-cache");
ingest(get_mds_options(), "mds");
ingest(get_mds_client_options(), "mds_client");
+ ingest(get_cephfs_shell_options(), "cephfs-shell");
return result;
}
projects / ceph.git / blobdiff