1 // -*- mode:C++; tab-width:8; c-basic-offset:2; indent-tabs-mode:t -*-
2 // vim: ts=8 sw=2 smarttab
4 * Ceph - scalable distributed file system
6 * Copyright (C) 2004-2006 Sage Weil <sage@newdream.net>
8 * This is free software; you can redistribute it and/or
9 * modify it under the terms of the GNU Lesser General Public
10 * License version 2.1, as published by the Free Software
11 * Foundation. See file COPYING.
20 #include "msg/Dispatcher.h"
22 #include "common/Mutex.h"
23 #include "common/RWLock.h"
24 #include "common/Timer.h"
25 #include "common/WorkQueue.h"
26 #include "common/AsyncReserver.h"
27 #include "common/ceph_context.h"
28 #include "common/zipkin_trace.h"
30 #include "mgr/MgrClient.h"
32 #include "os/ObjectStore.h"
35 #include "auth/KeyRing.h"
36 #include "osd/ClassHandler.h"
38 #include "include/CompatSet.h"
40 #include "OpRequest.h"
43 #include "osd/PGQueueable.h"
48 #include "include/memory.h"
51 #include "include/unordered_map.h"
53 #include "common/shared_cache.hpp"
54 #include "common/simple_cache.hpp"
55 #include "common/sharedptr_registry.hpp"
56 #include "common/WeightedPriorityQueue.h"
57 #include "common/PrioritizedQueue.h"
58 #include "osd/mClockOpClassQueue.h"
59 #include "osd/mClockClientQueue.h"
60 #include "messages/MOSDOp.h"
61 #include "include/Spinlock.h"
62 #include "common/EventTrace.h"
64 #define CEPH_OSD_PROTOCOL 10 /* cluster internal */
79 l_osd_op_r_lat_outb_hist
,
80 l_osd_op_r_process_lat
,
81 l_osd_op_r_prepare_lat
,
85 l_osd_op_w_lat_inb_hist
,
86 l_osd_op_w_process_lat
,
87 l_osd_op_w_prepare_lat
,
92 l_osd_op_rw_lat_inb_hist
,
93 l_osd_op_rw_lat_outb_hist
,
94 l_osd_op_rw_process_lat
,
95 l_osd_op_rw_prepare_lat
,
97 l_osd_op_before_queue_op_lat
,
98 l_osd_op_before_dequeue_op_lat
,
120 l_osd_history_alloc_bytes
,
121 l_osd_history_alloc_num
,
123 l_osd_cached_crc_adjusted
,
135 l_osd_waiting_for_map
,
138 l_osd_map_cache_miss
,
139 l_osd_map_cache_miss_low
,
140 l_osd_map_cache_miss_low_avg
,
141 l_osd_map_bl_cache_hit
,
142 l_osd_map_bl_cache_miss
,
145 l_osd_stat_bytes_used
,
146 l_osd_stat_bytes_avail
,
152 l_osd_tier_flush_fail
,
153 l_osd_tier_try_flush
,
154 l_osd_tier_try_flush_fail
,
160 l_osd_tier_proxy_read
,
161 l_osd_tier_proxy_write
,
168 l_osd_object_ctx_cache_hit
,
169 l_osd_object_ctx_cache_total
,
172 l_osd_tier_flush_lat
,
173 l_osd_tier_promote_lat
,
183 // RecoveryState perf counters
192 rs_backfilling_latency
,
193 rs_waitremotebackfillreserved_latency
,
194 rs_waitlocalbackfillreserved_latency
,
195 rs_notbackfilling_latency
,
196 rs_repnotrecovering_latency
,
197 rs_repwaitrecoveryreserved_latency
,
198 rs_repwaitbackfillreserved_latency
,
199 rs_reprecovering_latency
,
200 rs_activating_latency
,
201 rs_waitlocalrecoveryreserved_latency
,
202 rs_waitremoterecoveryreserved_latency
,
203 rs_recovering_latency
,
204 rs_recovered_latency
,
207 rs_replicaactive_latency
,
211 rs_waitactingchange_latency
,
212 rs_incomplete_latency
,
214 rs_getmissing_latency
,
215 rs_waitupthru_latency
,
216 rs_notrecovering_latency
,
233 class AuthAuthorizeHandlerRegistry
;
235 class TestOpsSocketHook
;
236 struct C_CompleteSplits
;
240 typedef ceph::shared_ptr
<ObjectStore::Sequencer
> SequencerRef
;
243 class DeletingState
{
257 const PGRef old_pg_state
;
258 explicit DeletingState(const pair
<spg_t
, PGRef
> &in
) :
259 lock("DeletingState::lock"), status(QUEUED
), stop_deleting(false),
260 pgid(in
.first
), old_pg_state(in
.second
) {
263 /// transition status to CLEARING_WAITING
264 bool pause_clearing() {
265 Mutex::Locker
l(lock
);
266 assert(status
== CLEARING_DIR
);
272 status
= CLEARING_WAITING
;
274 } ///< @return false if we should cancel deletion
276 /// start or resume the clearing - transition the status to CLEARING_DIR
277 bool start_or_resume_clearing() {
278 Mutex::Locker
l(lock
);
281 status
== DELETED_DIR
||
282 status
== CLEARING_WAITING
);
288 status
= CLEARING_DIR
;
290 } ///< @return false if we should cancel the deletion
292 /// transition status to CLEARING_DIR
293 bool resume_clearing() {
294 Mutex::Locker
l(lock
);
295 assert(status
== CLEARING_WAITING
);
301 status
= CLEARING_DIR
;
303 } ///< @return false if we should cancel deletion
305 /// transition status to deleting
306 bool start_deleting() {
307 Mutex::Locker
l(lock
);
308 assert(status
== CLEARING_DIR
);
314 status
= DELETING_DIR
;
316 } ///< @return false if we should cancel deletion
318 /// signal collection removal queued
319 void finish_deleting() {
320 Mutex::Locker
l(lock
);
321 assert(status
== DELETING_DIR
);
322 status
= DELETED_DIR
;
326 /// try to halt the deletion
327 bool try_stop_deletion() {
328 Mutex::Locker
l(lock
);
329 stop_deleting
= true;
331 * If we are in DELETING_DIR or CLEARING_DIR, there are in progress
332 * operations we have to wait for before continuing on. States
333 * CLEARING_WAITING and QUEUED indicate that the remover will check
334 * stop_deleting before queueing any further operations. CANCELED
335 * indicates that the remover has already halted. DELETED_DIR
336 * indicates that the deletion has been fully queued.
338 while (status
== DELETING_DIR
|| status
== CLEARING_DIR
)
340 return status
!= DELETED_DIR
;
341 } ///< @return true if we don't need to recreate the collection
343 typedef ceph::shared_ptr
<DeletingState
> DeletingStateRef
;
351 SharedPtrRegistry
<spg_t
, ObjectStore::Sequencer
> osr_registry
;
352 ceph::shared_ptr
<ObjectStore::Sequencer
> meta_osr
;
353 SharedPtrRegistry
<spg_t
, DeletingState
> deleting_pgs
;
356 LogClient
&log_client
;
358 PGRecoveryStats
&pg_recovery_stats
;
360 Messenger
*&cluster_messenger
;
361 Messenger
*&client_messenger
;
363 PerfCounters
*&logger
;
364 PerfCounters
*&recoverystate_perf
;
366 ThreadPool::BatchWorkQueue
<PG
> &peering_wq
;
367 GenContextWQ recovery_gen_wq
;
368 ClassHandler
*&class_handler
;
370 void enqueue_back(spg_t pgid
, PGQueueable qi
);
371 void enqueue_front(spg_t pgid
, PGQueueable qi
);
373 void maybe_inject_dispatch_delay() {
374 if (g_conf
->osd_debug_inject_dispatch_delay_probability
> 0) {
376 g_conf
->osd_debug_inject_dispatch_delay_probability
* 10000) {
378 t
.set_from_double(g_conf
->osd_debug_inject_dispatch_delay_duration
);
385 // -- map epoch lower bound --
387 multiset
<epoch_t
> pg_epochs
;
388 map
<spg_t
,epoch_t
> pg_epoch
;
391 void pg_add_epoch(spg_t pgid
, epoch_t epoch
) {
392 Mutex::Locker
l(pg_epoch_lock
);
393 map
<spg_t
,epoch_t
>::iterator t
= pg_epoch
.find(pgid
);
394 assert(t
== pg_epoch
.end());
395 pg_epoch
[pgid
] = epoch
;
396 pg_epochs
.insert(epoch
);
398 void pg_update_epoch(spg_t pgid
, epoch_t epoch
) {
399 Mutex::Locker
l(pg_epoch_lock
);
400 map
<spg_t
,epoch_t
>::iterator t
= pg_epoch
.find(pgid
);
401 assert(t
!= pg_epoch
.end());
402 pg_epochs
.erase(pg_epochs
.find(t
->second
));
404 pg_epochs
.insert(epoch
);
406 void pg_remove_epoch(spg_t pgid
) {
407 Mutex::Locker
l(pg_epoch_lock
);
408 map
<spg_t
,epoch_t
>::iterator t
= pg_epoch
.find(pgid
);
409 if (t
!= pg_epoch
.end()) {
410 pg_epochs
.erase(pg_epochs
.find(t
->second
));
414 epoch_t
get_min_pg_epoch() {
415 Mutex::Locker
l(pg_epoch_lock
);
416 if (pg_epochs
.empty())
419 return *pg_epochs
.begin();
424 Mutex publish_lock
, pre_publish_lock
; // pre-publish orders before publish
425 OSDSuperblock superblock
;
428 OSDSuperblock
get_superblock() {
429 Mutex::Locker
l(publish_lock
);
432 void publish_superblock(const OSDSuperblock
&block
) {
433 Mutex::Locker
l(publish_lock
);
437 int get_nodeid() const { return whoami
; }
439 std::atomic
<epoch_t
> max_oldest_map
;
444 OSDMapRef
get_osdmap() {
445 Mutex::Locker
l(publish_lock
);
448 epoch_t
get_osdmap_epoch() {
449 Mutex::Locker
l(publish_lock
);
450 return osdmap
? osdmap
->get_epoch() : 0;
452 void publish_map(OSDMapRef map
) {
453 Mutex::Locker
l(publish_lock
);
458 * osdmap - current published map
459 * next_osdmap - pre_published map that is about to be published.
461 * We use the next_osdmap to send messages and initiate connections,
462 * but only if the target is the same instance as the one in the map
463 * epoch the current user is working from (i.e., the result is
464 * equivalent to what is in next_osdmap).
466 * This allows the helpers to start ignoring osds that are about to
467 * go down, and let OSD::handle_osd_map()/note_down_osd() mark them
468 * down, without worrying about reopening connections from threads
469 * working from old maps.
472 OSDMapRef next_osdmap
;
473 Cond pre_publish_cond
;
476 void pre_publish_map(OSDMapRef map
) {
477 Mutex::Locker
l(pre_publish_lock
);
478 next_osdmap
= std::move(map
);
482 /// map epochs reserved below
483 map
<epoch_t
, unsigned> map_reservations
;
485 /// gets ref to next_osdmap and registers the epoch as reserved
486 OSDMapRef
get_nextmap_reserved() {
487 Mutex::Locker
l(pre_publish_lock
);
490 epoch_t e
= next_osdmap
->get_epoch();
491 map
<epoch_t
, unsigned>::iterator i
=
492 map_reservations
.insert(make_pair(e
, 0)).first
;
496 /// releases reservation on map
497 void release_map(OSDMapRef osdmap
) {
498 Mutex::Locker
l(pre_publish_lock
);
499 map
<epoch_t
, unsigned>::iterator i
=
500 map_reservations
.find(osdmap
->get_epoch());
501 assert(i
!= map_reservations
.end());
502 assert(i
->second
> 0);
503 if (--(i
->second
) == 0) {
504 map_reservations
.erase(i
);
506 pre_publish_cond
.Signal();
508 /// blocks until there are no reserved maps prior to next_osdmap
509 void await_reserved_maps() {
510 Mutex::Locker
l(pre_publish_lock
);
513 map
<epoch_t
, unsigned>::const_iterator i
= map_reservations
.cbegin();
514 if (i
== map_reservations
.cend() || i
->first
>= next_osdmap
->get_epoch()) {
517 pre_publish_cond
.Wait(pre_publish_lock
);
523 Mutex peer_map_epoch_lock
;
524 map
<int, epoch_t
> peer_map_epoch
;
526 epoch_t
get_peer_epoch(int p
);
527 epoch_t
note_peer_epoch(int p
, epoch_t e
);
528 void forget_peer_epoch(int p
, epoch_t e
);
530 void send_map(class MOSDMap
*m
, Connection
*con
);
531 void send_incremental_map(epoch_t since
, Connection
*con
, OSDMapRef
& osdmap
);
532 MOSDMap
*build_incremental_map_msg(epoch_t from
, epoch_t to
,
533 OSDSuperblock
& superblock
);
534 bool should_share_map(entity_name_t name
, Connection
*con
, epoch_t epoch
,
535 const OSDMapRef
& osdmap
, const epoch_t
*sent_epoch_p
);
536 void share_map(entity_name_t name
, Connection
*con
, epoch_t epoch
,
537 OSDMapRef
& osdmap
, epoch_t
*sent_epoch_p
);
538 void share_map_peer(int peer
, Connection
*con
,
539 OSDMapRef map
= OSDMapRef());
541 ConnectionRef
get_con_osd_cluster(int peer
, epoch_t from_epoch
);
542 pair
<ConnectionRef
,ConnectionRef
> get_con_osd_hb(int peer
, epoch_t from_epoch
); // (back, front)
543 void send_message_osd_cluster(int peer
, Message
*m
, epoch_t from_epoch
);
544 void send_message_osd_cluster(Message
*m
, Connection
*con
) {
545 con
->send_message(m
);
547 void send_message_osd_cluster(Message
*m
, const ConnectionRef
& con
) {
548 con
->send_message(m
);
550 void send_message_osd_client(Message
*m
, Connection
*con
) {
551 con
->send_message(m
);
553 void send_message_osd_client(Message
*m
, const ConnectionRef
& con
) {
554 con
->send_message(m
);
556 entity_name_t
get_cluster_msgr_name() {
557 return cluster_messenger
->get_myname();
561 // -- scrub scheduling --
562 Mutex sched_scrub_lock
;
569 /// pg to be scrubbed
571 /// a time scheduled for scrub. but the scrub could be delayed if system
572 /// load is too high or it fails to fall in the scrub hours
574 /// the hard upper bound of scrub time
576 ScrubJob() : cct(nullptr) {}
577 explicit ScrubJob(CephContext
* cct
, const spg_t
& pg
,
578 const utime_t
& timestamp
,
579 double pool_scrub_min_interval
= 0,
580 double pool_scrub_max_interval
= 0, bool must
= true);
581 /// order the jobs by sched_time
582 bool operator<(const ScrubJob
& rhs
) const;
584 set
<ScrubJob
> sched_scrub_pg
;
586 /// @returns the scrub_reg_stamp used for unregister the scrub job
587 utime_t
reg_pg_scrub(spg_t pgid
, utime_t t
, double pool_scrub_min_interval
,
588 double pool_scrub_max_interval
, bool must
) {
589 ScrubJob
scrub(cct
, pgid
, t
, pool_scrub_min_interval
, pool_scrub_max_interval
,
591 Mutex::Locker
l(sched_scrub_lock
);
592 sched_scrub_pg
.insert(scrub
);
593 return scrub
.sched_time
;
595 void unreg_pg_scrub(spg_t pgid
, utime_t t
) {
596 Mutex::Locker
l(sched_scrub_lock
);
597 size_t removed
= sched_scrub_pg
.erase(ScrubJob(cct
, pgid
, t
));
600 bool first_scrub_stamp(ScrubJob
*out
) {
601 Mutex::Locker
l(sched_scrub_lock
);
602 if (sched_scrub_pg
.empty())
604 set
<ScrubJob
>::iterator iter
= sched_scrub_pg
.begin();
608 bool next_scrub_stamp(const ScrubJob
& next
,
610 Mutex::Locker
l(sched_scrub_lock
);
611 if (sched_scrub_pg
.empty())
613 set
<ScrubJob
>::const_iterator iter
= sched_scrub_pg
.lower_bound(next
);
614 if (iter
== sched_scrub_pg
.cend())
617 if (iter
== sched_scrub_pg
.cend())
623 void dumps_scrub(Formatter
*f
) {
624 assert(f
!= nullptr);
625 Mutex::Locker
l(sched_scrub_lock
);
627 f
->open_array_section("scrubs");
628 for (const auto &i
: sched_scrub_pg
) {
629 f
->open_object_section("scrub");
630 f
->dump_stream("pgid") << i
.pgid
;
631 f
->dump_stream("sched_time") << i
.sched_time
;
632 f
->dump_stream("deadline") << i
.deadline
;
633 f
->dump_bool("forced", i
.sched_time
== i
.deadline
);
639 bool can_inc_scrubs_pending();
640 bool inc_scrubs_pending();
641 void inc_scrubs_active(bool reserved
);
642 void dec_scrubs_pending();
643 void dec_scrubs_active();
645 void reply_op_error(OpRequestRef op
, int err
);
646 void reply_op_error(OpRequestRef op
, int err
, eversion_t v
, version_t uv
);
647 void handle_misdirected_op(PG
*pg
, OpRequestRef op
);
651 // -- agent shared state --
654 map
<uint64_t, set
<PGRef
> > agent_queue
;
655 set
<PGRef
>::iterator agent_queue_pos
;
656 bool agent_valid_iterator
;
658 int flush_mode_high_count
; //once have one pg with FLUSH_MODE_HIGH then flush objects with high speed
659 set
<hobject_t
> agent_oids
;
661 struct AgentThread
: public Thread
{
663 explicit AgentThread(OSDService
*o
) : osd(o
) {}
664 void *entry() override
{
669 bool agent_stop_flag
;
670 Mutex agent_timer_lock
;
671 SafeTimer agent_timer
;
677 void _enqueue(PG
*pg
, uint64_t priority
) {
678 if (!agent_queue
.empty() &&
679 agent_queue
.rbegin()->first
< priority
)
680 agent_valid_iterator
= false; // inserting higher-priority queue
681 set
<PGRef
>& nq
= agent_queue
[priority
];
687 void _dequeue(PG
*pg
, uint64_t old_priority
) {
688 set
<PGRef
>& oq
= agent_queue
[old_priority
];
689 set
<PGRef
>::iterator p
= oq
.find(pg
);
690 assert(p
!= oq
.end());
691 if (p
== agent_queue_pos
)
695 if (agent_queue
.rbegin()->first
== old_priority
)
696 agent_valid_iterator
= false;
697 agent_queue
.erase(old_priority
);
701 /// enable agent for a pg
702 void agent_enable_pg(PG
*pg
, uint64_t priority
) {
703 Mutex::Locker
l(agent_lock
);
704 _enqueue(pg
, priority
);
707 /// adjust priority for an enagled pg
708 void agent_adjust_pg(PG
*pg
, uint64_t old_priority
, uint64_t new_priority
) {
709 Mutex::Locker
l(agent_lock
);
710 assert(new_priority
!= old_priority
);
711 _enqueue(pg
, new_priority
);
712 _dequeue(pg
, old_priority
);
715 /// disable agent for a pg
716 void agent_disable_pg(PG
*pg
, uint64_t old_priority
) {
717 Mutex::Locker
l(agent_lock
);
718 _dequeue(pg
, old_priority
);
721 /// note start of an async (evict) op
722 void agent_start_evict_op() {
723 Mutex::Locker
l(agent_lock
);
727 /// note finish or cancellation of an async (evict) op
728 void agent_finish_evict_op() {
729 Mutex::Locker
l(agent_lock
);
730 assert(agent_ops
> 0);
735 /// note start of an async (flush) op
736 void agent_start_op(const hobject_t
& oid
) {
737 Mutex::Locker
l(agent_lock
);
739 assert(agent_oids
.count(oid
) == 0);
740 agent_oids
.insert(oid
);
743 /// note finish or cancellation of an async (flush) op
744 void agent_finish_op(const hobject_t
& oid
) {
745 Mutex::Locker
l(agent_lock
);
746 assert(agent_ops
> 0);
748 assert(agent_oids
.count(oid
) == 1);
749 agent_oids
.erase(oid
);
753 /// check if we are operating on an object
754 bool agent_is_active_oid(const hobject_t
& oid
) {
755 Mutex::Locker
l(agent_lock
);
756 return agent_oids
.count(oid
);
759 /// get count of active agent ops
760 int agent_get_num_ops() {
761 Mutex::Locker
l(agent_lock
);
765 void agent_inc_high_count() {
766 Mutex::Locker
l(agent_lock
);
767 flush_mode_high_count
++;
770 void agent_dec_high_count() {
771 Mutex::Locker
l(agent_lock
);
772 flush_mode_high_count
--;
776 /// throttle promotion attempts
777 std::atomic_uint promote_probability_millis
{1000}; ///< probability thousands. one word.
778 PromoteCounter promote_counter
;
779 utime_t last_recalibrate
;
780 unsigned long promote_max_objects
, promote_max_bytes
;
783 bool promote_throttle() {
784 // NOTE: lockless! we rely on the probability being a single word.
785 promote_counter
.attempt();
786 if ((unsigned)rand() % 1000 > promote_probability_millis
)
787 return true; // yes throttle (no promote)
788 if (promote_max_objects
&&
789 promote_counter
.objects
> promote_max_objects
)
790 return true; // yes throttle
791 if (promote_max_bytes
&&
792 promote_counter
.bytes
> promote_max_bytes
)
793 return true; // yes throttle
794 return false; // no throttle (promote)
796 void promote_finish(uint64_t bytes
) {
797 promote_counter
.finish(bytes
);
799 void promote_throttle_recalibrate();
801 // -- Objecter, for tiering reads/writes from/to other OSDs --
803 Finisher objecter_finisher
;
807 SafeTimer watch_timer
;
808 uint64_t next_notif_id
;
809 uint64_t get_next_id(epoch_t cur_epoch
) {
810 Mutex::Locker
l(watch_lock
);
811 return (((uint64_t)cur_epoch
) << 32) | ((uint64_t)(next_notif_id
++));
814 // -- Recovery/Backfill Request Scheduling --
815 Mutex recovery_request_lock
;
816 SafeTimer recovery_request_timer
;
818 // For async recovery sleep
819 bool recovery_needs_sleep
= true;
820 utime_t recovery_schedule_time
= utime_t();
822 Mutex recovery_sleep_lock
;
823 SafeTimer recovery_sleep_timer
;
827 std::atomic_uint last_tid
{0};
828 ceph_tid_t
get_tid() {
829 return (ceph_tid_t
)last_tid
++;
832 // -- backfill_reservation --
833 Finisher reserver_finisher
;
834 AsyncReserver
<spg_t
> local_reserver
;
835 AsyncReserver
<spg_t
> remote_reserver
;
840 map
<pg_t
, vector
<int> > pg_temp_wanted
;
841 map
<pg_t
, vector
<int> > pg_temp_pending
;
842 void _sent_pg_temp();
844 void queue_want_pg_temp(pg_t pgid
, vector
<int>& want
);
845 void remove_want_pg_temp(pg_t pgid
);
846 void requeue_pg_temp();
849 void send_pg_created(pg_t pgid
);
851 void queue_for_peering(PG
*pg
);
853 Mutex snap_sleep_lock
;
854 SafeTimer snap_sleep_timer
;
856 Mutex scrub_sleep_lock
;
857 SafeTimer scrub_sleep_timer
;
859 AsyncReserver
<spg_t
> snap_reserver
;
860 void queue_for_snap_trim(PG
*pg
);
862 void queue_for_scrub(PG
*pg
, bool with_high_priority
) {
863 unsigned scrub_queue_priority
= pg
->scrubber
.priority
;
864 if (with_high_priority
&& scrub_queue_priority
< cct
->_conf
->osd_client_op_priority
) {
865 scrub_queue_priority
= cct
->_conf
->osd_client_op_priority
;
870 PGScrub(pg
->get_osdmap()->get_epoch()),
871 cct
->_conf
->osd_scrub_cost
,
872 scrub_queue_priority
,
875 pg
->get_osdmap()->get_epoch()));
879 // -- pg recovery and associated throttling --
881 list
<pair
<epoch_t
, PGRef
> > awaiting_throttle
;
883 utime_t defer_recovery_until
;
884 uint64_t recovery_ops_active
;
885 uint64_t recovery_ops_reserved
;
886 bool recovery_paused
;
887 #ifdef DEBUG_RECOVERY_OIDS
888 map
<spg_t
, set
<hobject_t
> > recovery_oids
;
890 bool _recover_now(uint64_t *available_pushes
);
891 void _maybe_queue_recovery();
892 void _queue_for_recovery(
893 pair
<epoch_t
, PGRef
> p
, uint64_t reserved_pushes
) {
894 assert(recovery_lock
.is_locked_by_me());
898 PGRecovery(p
.first
, reserved_pushes
),
899 cct
->_conf
->osd_recovery_cost
,
900 cct
->_conf
->osd_recovery_priority
,
906 void start_recovery_op(PG
*pg
, const hobject_t
& soid
);
907 void finish_recovery_op(PG
*pg
, const hobject_t
& soid
, bool dequeue
);
908 bool is_recovery_active();
909 void release_reserved_pushes(uint64_t pushes
) {
910 Mutex::Locker
l(recovery_lock
);
911 assert(recovery_ops_reserved
>= pushes
);
912 recovery_ops_reserved
-= pushes
;
913 _maybe_queue_recovery();
915 void defer_recovery(float defer_for
) {
916 defer_recovery_until
= ceph_clock_now();
917 defer_recovery_until
+= defer_for
;
919 void pause_recovery() {
920 Mutex::Locker
l(recovery_lock
);
921 recovery_paused
= true;
923 bool recovery_is_paused() {
924 Mutex::Locker
l(recovery_lock
);
925 return recovery_paused
;
927 void unpause_recovery() {
928 Mutex::Locker
l(recovery_lock
);
929 recovery_paused
= false;
930 _maybe_queue_recovery();
932 void kick_recovery_queue() {
933 Mutex::Locker
l(recovery_lock
);
934 _maybe_queue_recovery();
936 void clear_queued_recovery(PG
*pg
) {
937 Mutex::Locker
l(recovery_lock
);
938 for (list
<pair
<epoch_t
, PGRef
> >::iterator i
= awaiting_throttle
.begin();
939 i
!= awaiting_throttle
.end();
941 if (i
->second
.get() == pg
) {
942 awaiting_throttle
.erase(i
);
949 // delayed pg activation
950 void queue_for_recovery(PG
*pg
) {
951 Mutex::Locker
l(recovery_lock
);
953 if (pg
->get_state() & (PG_STATE_FORCED_RECOVERY
| PG_STATE_FORCED_BACKFILL
)) {
954 awaiting_throttle
.push_front(make_pair(pg
->get_osdmap()->get_epoch(), pg
));
956 awaiting_throttle
.push_back(make_pair(pg
->get_osdmap()->get_epoch(), pg
));
958 _maybe_queue_recovery();
960 void queue_recovery_after_sleep(PG
*pg
, epoch_t queued
, uint64_t reserved_pushes
) {
961 Mutex::Locker
l(recovery_lock
);
962 _queue_for_recovery(make_pair(queued
, pg
), reserved_pushes
);
965 void adjust_pg_priorities(const vector
<PGRef
>& pgs
, int newflags
);
967 // osd map cache (past osd maps)
968 Mutex map_cache_lock
;
969 SharedLRU
<epoch_t
, const OSDMap
> map_cache
;
970 SimpleLRU
<epoch_t
, bufferlist
> map_bl_cache
;
971 SimpleLRU
<epoch_t
, bufferlist
> map_bl_inc_cache
;
973 OSDMapRef
try_get_map(epoch_t e
);
974 OSDMapRef
get_map(epoch_t e
) {
975 OSDMapRef
ret(try_get_map(e
));
979 OSDMapRef
add_map(OSDMap
*o
) {
980 Mutex::Locker
l(map_cache_lock
);
983 OSDMapRef
_add_map(OSDMap
*o
);
985 void add_map_bl(epoch_t e
, bufferlist
& bl
) {
986 Mutex::Locker
l(map_cache_lock
);
987 return _add_map_bl(e
, bl
);
989 void pin_map_bl(epoch_t e
, bufferlist
&bl
);
990 void _add_map_bl(epoch_t e
, bufferlist
& bl
);
991 bool get_map_bl(epoch_t e
, bufferlist
& bl
) {
992 Mutex::Locker
l(map_cache_lock
);
993 return _get_map_bl(e
, bl
);
995 bool _get_map_bl(epoch_t e
, bufferlist
& bl
);
997 void add_map_inc_bl(epoch_t e
, bufferlist
& bl
) {
998 Mutex::Locker
l(map_cache_lock
);
999 return _add_map_inc_bl(e
, bl
);
1001 void pin_map_inc_bl(epoch_t e
, bufferlist
&bl
);
1002 void _add_map_inc_bl(epoch_t e
, bufferlist
& bl
);
1003 bool get_inc_map_bl(epoch_t e
, bufferlist
& bl
);
1005 void clear_map_bl_cache_pins(epoch_t e
);
1007 void need_heartbeat_peer_update();
1009 void pg_stat_queue_enqueue(PG
*pg
);
1010 void pg_stat_queue_dequeue(PG
*pg
);
1014 void start_shutdown();
1015 void shutdown_reserver();
1020 Mutex in_progress_split_lock
;
1021 map
<spg_t
, spg_t
> pending_splits
; // child -> parent
1022 map
<spg_t
, set
<spg_t
> > rev_pending_splits
; // parent -> [children]
1023 set
<spg_t
> in_progress_splits
; // child
1026 void _start_split(spg_t parent
, const set
<spg_t
> &children
);
1027 void start_split(spg_t parent
, const set
<spg_t
> &children
) {
1028 Mutex::Locker
l(in_progress_split_lock
);
1029 return _start_split(parent
, children
);
1031 void mark_split_in_progress(spg_t parent
, const set
<spg_t
> &pgs
);
1032 void complete_split(const set
<spg_t
> &pgs
);
1033 void cancel_pending_splits_for_parent(spg_t parent
);
1034 void _cancel_pending_splits_for_parent(spg_t parent
);
1035 bool splitting(spg_t pgid
);
1036 void expand_pg_num(OSDMapRef old_map
,
1038 void _maybe_split_pgid(OSDMapRef old_map
,
1041 void init_splits_between(spg_t pgid
, OSDMapRef frommap
, OSDMapRef tomap
);
1045 osd_stat_t osd_stat
;
1048 void update_osd_stat(vector
<int>& hb_peers
);
1049 osd_stat_t
set_osd_stat(const struct store_statfs_t
&stbuf
,
1050 vector
<int>& hb_peers
,
1052 osd_stat_t
get_osd_stat() {
1053 Mutex::Locker
l(stat_lock
);
1055 osd_stat
.up_from
= up_epoch
;
1056 osd_stat
.seq
= ((uint64_t)osd_stat
.up_from
<< 32) + seq
;
1059 uint64_t get_osd_stat_seq() {
1060 Mutex::Locker
l(stat_lock
);
1061 return osd_stat
.seq
;
1064 // -- OSD Full Status --
1066 friend TestOpsSocketHook
;
1067 mutable Mutex full_status_lock
;
1068 enum s_names
{ INVALID
= -1, NONE
, NEARFULL
, BACKFILLFULL
, FULL
, FAILSAFE
} cur_state
; // ascending
1069 const char *get_full_state_name(s_names s
) const {
1071 case NONE
: return "none";
1072 case NEARFULL
: return "nearfull";
1073 case BACKFILLFULL
: return "backfillfull";
1074 case FULL
: return "full";
1075 case FAILSAFE
: return "failsafe";
1076 default: return "???";
1079 s_names
get_full_state(string type
) const {
1082 else if (type
== "failsafe")
1084 else if (type
== "full")
1086 else if (type
== "backfillfull")
1087 return BACKFILLFULL
;
1088 else if (type
== "nearfull")
1093 double cur_ratio
; ///< current utilization
1094 mutable int64_t injectfull
= 0;
1095 s_names injectfull_state
= NONE
;
1096 float get_failsafe_full_ratio();
1097 void check_full_status(float ratio
);
1098 bool _check_full(s_names type
, ostream
&ss
) const;
1100 bool check_failsafe_full(ostream
&ss
) const;
1101 bool check_full(ostream
&ss
) const;
1102 bool check_backfill_full(ostream
&ss
) const;
1103 bool check_nearfull(ostream
&ss
) const;
1104 bool is_failsafe_full() const;
1105 bool is_full() const;
1106 bool is_backfillfull() const;
1107 bool is_nearfull() const;
1108 bool need_fullness_update(); ///< osdmap state needs update
1109 void set_injectfull(s_names type
, int64_t count
);
1110 bool check_osdmap_full(const set
<pg_shard_t
> &missing_on
);
1115 mutable Mutex epoch_lock
; // protects access to boot_epoch, up_epoch, bind_epoch
1116 epoch_t boot_epoch
; // _first_ epoch we were marked up (after this process started)
1117 epoch_t up_epoch
; // _most_recent_ epoch we were marked up
1118 epoch_t bind_epoch
; // epoch we last did a bind to new ip:ports
1121 * Retrieve the boot_, up_, and bind_ epochs the OSD has set. The params
1122 * can be NULL if you don't care about them.
1124 void retrieve_epochs(epoch_t
*_boot_epoch
, epoch_t
*_up_epoch
,
1125 epoch_t
*_bind_epoch
) const;
1127 * Set the boot, up, and bind epochs. Any NULL params will not be set.
1129 void set_epochs(const epoch_t
*_boot_epoch
, const epoch_t
*_up_epoch
,
1130 const epoch_t
*_bind_epoch
);
1131 epoch_t
get_boot_epoch() const {
1133 retrieve_epochs(&ret
, NULL
, NULL
);
1136 epoch_t
get_up_epoch() const {
1138 retrieve_epochs(NULL
, &ret
, NULL
);
1141 epoch_t
get_bind_epoch() const {
1143 retrieve_epochs(NULL
, NULL
, &ret
);
1147 void request_osdmap_update(epoch_t e
);
1150 Mutex is_stopping_lock
;
1151 Cond is_stopping_cond
;
1156 std::atomic_int state
{NOT_STOPPING
};
1160 void set_state(int s
) {
1163 bool is_stopping() const {
1164 return state
== STOPPING
;
1166 bool is_preparing_to_stop() const {
1167 return state
== PREPARING_TO_STOP
;
1169 bool prepare_to_stop();
1170 void got_stop_ack();
1173 #ifdef PG_DEBUG_REFS
1175 map
<spg_t
, int> pgid_tracker
;
1176 map
<spg_t
, PG
*> live_pgs
;
1177 void add_pgid(spg_t pgid
, PG
*pg
);
1178 void remove_pgid(spg_t pgid
, PG
*pg
);
1179 void dump_live_pgids();
1182 explicit OSDService(OSD
*osd
);
1186 class OSD
: public Dispatcher
,
1187 public md_config_obs_t
{
1189 Mutex osd_lock
; // global lock
1190 SafeTimer tick_timer
; // safe timer (osd_lock)
1192 // Tick timer for those stuff that do not need osd_lock
1193 Mutex tick_timer_lock
;
1194 SafeTimer tick_timer_without_osd_lock
;
1196 // config observer bits
1197 const char** get_tracked_conf_keys() const override
;
1198 void handle_conf_change(const struct md_config_t
*conf
,
1199 const std::set
<std::string
> &changed
) override
;
1200 void update_log_config();
1201 void check_config();
1205 static const double OSD_TICK_INTERVAL
; // tick interval for tick_timer and tick_timer_without_osd_lock
1207 AuthAuthorizeHandlerRegistry
*authorize_handler_cluster_registry
;
1208 AuthAuthorizeHandlerRegistry
*authorize_handler_service_registry
;
1210 Messenger
*cluster_messenger
;
1211 Messenger
*client_messenger
;
1212 Messenger
*objecter_messenger
;
1213 MonClient
*monc
; // check the "monc helpers" list before accessing directly
1215 PerfCounters
*logger
;
1216 PerfCounters
*recoverystate_perf
;
1219 FuseStore
*fuse_store
= nullptr;
1221 LogClient log_client
;
1225 std::string dev_path
, journal_path
;
1227 bool store_is_rotational
= true;
1228 bool journal_is_rotational
= true;
1230 ZTracer::Endpoint trace_endpoint
;
1231 void create_logger();
1232 void create_recoverystate_perf();
1234 void tick_without_osd_lock();
1235 void _dispatch(Message
*m
);
1236 void dispatch_op(OpRequestRef op
);
1238 void check_osdmap_features(ObjectStore
*store
);
1241 friend class OSDSocketHook
;
1242 class OSDSocketHook
*asok_hook
;
1243 bool asok_command(string admin_command
, cmdmap_t
& cmdmap
, string format
, ostream
& ss
);
1246 ClassHandler
*class_handler
= nullptr;
1247 int get_nodeid() { return whoami
; }
1249 static ghobject_t
get_osdmap_pobject_name(epoch_t epoch
) {
1251 snprintf(foo
, sizeof(foo
), "osdmap.%d", epoch
);
1252 return ghobject_t(hobject_t(sobject_t(object_t(foo
), 0)));
1254 static ghobject_t
get_inc_osdmap_pobject_name(epoch_t epoch
) {
1256 snprintf(foo
, sizeof(foo
), "inc_osdmap.%d", epoch
);
1257 return ghobject_t(hobject_t(sobject_t(object_t(foo
), 0)));
1260 static ghobject_t
make_snapmapper_oid() {
1261 return ghobject_t(hobject_t(
1263 object_t("snapmapper"),
1267 static ghobject_t
make_pg_log_oid(spg_t pg
) {
1269 ss
<< "pglog_" << pg
;
1272 return ghobject_t(hobject_t(sobject_t(object_t(s
.c_str()), 0)));
1275 static ghobject_t
make_pg_biginfo_oid(spg_t pg
) {
1277 ss
<< "pginfo_" << pg
;
1280 return ghobject_t(hobject_t(sobject_t(object_t(s
.c_str()), 0)));
1282 static ghobject_t
make_infos_oid() {
1283 hobject_t
oid(sobject_t("infos", CEPH_NOSNAP
));
1284 return ghobject_t(oid
);
1286 static void recursive_remove_collection(CephContext
* cct
,
1292 * get_osd_initial_compat_set()
1294 * Get the initial feature set for this OSD. Features
1295 * here are automatically upgraded.
1297 * Return value: Initial osd CompatSet
1299 static CompatSet
get_osd_initial_compat_set();
1302 * get_osd_compat_set()
1304 * Get all features supported by this OSD
1306 * Return value: CompatSet of all supported features
1308 static CompatSet
get_osd_compat_set();
1313 class C_Tick_WithoutOSDLock
;
1316 OSDSuperblock superblock
;
1318 void write_superblock();
1319 void write_superblock(ObjectStore::Transaction
& t
);
1320 int read_superblock();
1322 void clear_temp_objects();
1324 CompatSet osd_compat
;
1329 STATE_INITIALIZING
= 1,
1334 STATE_WAITING_FOR_HEALTHY
1337 static const char *get_state_name(int s
) {
1339 case STATE_INITIALIZING
: return "initializing";
1340 case STATE_PREBOOT
: return "preboot";
1341 case STATE_BOOTING
: return "booting";
1342 case STATE_ACTIVE
: return "active";
1343 case STATE_STOPPING
: return "stopping";
1344 case STATE_WAITING_FOR_HEALTHY
: return "waiting_for_healthy";
1345 default: return "???";
1350 std::atomic_int state
{STATE_INITIALIZING
};
1351 bool waiting_for_luminous_mons
= false;
1354 int get_state() const {
1357 void set_state(int s
) {
1360 bool is_initializing() const {
1361 return state
== STATE_INITIALIZING
;
1363 bool is_preboot() const {
1364 return state
== STATE_PREBOOT
;
1366 bool is_booting() const {
1367 return state
== STATE_BOOTING
;
1369 bool is_active() const {
1370 return state
== STATE_ACTIVE
;
1372 bool is_stopping() const {
1373 return state
== STATE_STOPPING
;
1375 bool is_waiting_for_healthy() const {
1376 return state
== STATE_WAITING_FOR_HEALTHY
;
1381 ThreadPool peering_tp
;
1382 ShardedThreadPool osd_op_tp
;
1384 ThreadPool command_tp
;
1386 void set_disk_tp_priority();
1387 void get_latest_osdmap();
1391 void dispatch_session_waiting(Session
*session
, OSDMapRef osdmap
);
1392 void maybe_share_map(Session
*session
, OpRequestRef op
, OSDMapRef osdmap
);
1394 Mutex session_waiting_lock
;
1395 set
<Session
*> session_waiting_for_map
;
1397 /// Caller assumes refs for included Sessions
1398 void get_sessions_waiting_for_map(set
<Session
*> *out
) {
1399 Mutex::Locker
l(session_waiting_lock
);
1400 out
->swap(session_waiting_for_map
);
1402 void register_session_waiting_on_map(Session
*session
) {
1403 Mutex::Locker
l(session_waiting_lock
);
1404 if (session_waiting_for_map
.insert(session
).second
) {
1408 void clear_session_waiting_on_map(Session
*session
) {
1409 Mutex::Locker
l(session_waiting_lock
);
1410 set
<Session
*>::iterator i
= session_waiting_for_map
.find(session
);
1411 if (i
!= session_waiting_for_map
.end()) {
1413 session_waiting_for_map
.erase(i
);
1416 void dispatch_sessions_waiting_on_map() {
1417 set
<Session
*> sessions_to_check
;
1418 get_sessions_waiting_for_map(&sessions_to_check
);
1419 for (set
<Session
*>::iterator i
= sessions_to_check
.begin();
1420 i
!= sessions_to_check
.end();
1421 sessions_to_check
.erase(i
++)) {
1422 (*i
)->session_dispatch_lock
.Lock();
1423 dispatch_session_waiting(*i
, osdmap
);
1424 (*i
)->session_dispatch_lock
.Unlock();
1428 void session_handle_reset(Session
*session
) {
1429 Mutex::Locker
l(session
->session_dispatch_lock
);
1430 clear_session_waiting_on_map(session
);
1432 session
->clear_backoffs();
1434 /* Messages have connection refs, we need to clear the
1435 * connection->session->message->connection
1436 * cycles which result.
1439 session
->waiting_on_map
.clear_and_dispose(TrackedOp::Putter());
1444 * @defgroup monc helpers
1446 * Right now we only have the one
1450 * Ask the Monitors for a sequence of OSDMaps.
1452 * @param epoch The epoch to start with when replying
1453 * @param force_request True if this request forces a new subscription to
1454 * the monitors; false if an outstanding request that encompasses it is
1457 void osdmap_subscribe(version_t epoch
, bool force_request
);
1458 /** @} monc helpers */
1460 Mutex osdmap_subscribe_lock
;
1461 epoch_t latest_subscribed_epoch
{0};
1464 /// information about a heartbeat peer
1465 struct HeartbeatInfo
{
1467 ConnectionRef con_front
; ///< peer connection (front)
1468 ConnectionRef con_back
; ///< peer connection (back)
1469 utime_t first_tx
; ///< time we sent our first ping request
1470 utime_t last_tx
; ///< last time we sent a ping request
1471 utime_t last_rx_front
; ///< last time we got a ping reply on the front side
1472 utime_t last_rx_back
; ///< last time we got a ping reply on the back side
1473 epoch_t epoch
; ///< most recent epoch we wanted this peer
1475 bool is_unhealthy(utime_t cutoff
) const {
1477 ! ((last_rx_front
> cutoff
||
1478 (last_rx_front
== utime_t() && (last_tx
== utime_t() ||
1479 first_tx
> cutoff
))) &&
1480 (last_rx_back
> cutoff
||
1481 (last_rx_back
== utime_t() && (last_tx
== utime_t() ||
1482 first_tx
> cutoff
))));
1484 bool is_healthy(utime_t cutoff
) const {
1485 return last_rx_front
> cutoff
&& last_rx_back
> cutoff
;
1489 /// state attached to outgoing heartbeat connections
1490 struct HeartbeatSession
: public RefCountedObject
{
1492 explicit HeartbeatSession(int p
) : peer(p
) {}
1494 Mutex heartbeat_lock
;
1495 map
<int, int> debug_heartbeat_drops_remaining
;
1496 Cond heartbeat_cond
;
1497 bool heartbeat_stop
;
1498 std::atomic_bool heartbeat_need_update
;
1499 map
<int,HeartbeatInfo
> heartbeat_peers
; ///< map of osd id to HeartbeatInfo
1500 utime_t last_mon_heartbeat
;
1501 Messenger
*hb_front_client_messenger
;
1502 Messenger
*hb_back_client_messenger
;
1503 Messenger
*hb_front_server_messenger
;
1504 Messenger
*hb_back_server_messenger
;
1505 utime_t last_heartbeat_resample
; ///< last time we chose random peers in waiting-for-healthy state
1506 double daily_loadavg
;
1508 void _add_heartbeat_peer(int p
);
1509 void _remove_heartbeat_peer(int p
);
1510 bool heartbeat_reset(Connection
*con
);
1511 void maybe_update_heartbeat_peers();
1512 void reset_heartbeat_peers();
1513 bool heartbeat_peers_need_update() {
1514 return heartbeat_need_update
.load();
1516 void heartbeat_set_peers_need_update() {
1517 heartbeat_need_update
.store(true);
1519 void heartbeat_clear_peers_need_update() {
1520 heartbeat_need_update
.store(false);
1523 void heartbeat_check();
1524 void heartbeat_entry();
1525 void need_heartbeat_peer_update();
1527 void heartbeat_kick() {
1528 Mutex::Locker
l(heartbeat_lock
);
1529 heartbeat_cond
.Signal();
1532 struct T_Heartbeat
: public Thread
{
1534 explicit T_Heartbeat(OSD
*o
) : osd(o
) {}
1535 void *entry() override
{
1536 osd
->heartbeat_entry();
1542 bool heartbeat_dispatch(Message
*m
);
1544 struct HeartbeatDispatcher
: public Dispatcher
{
1546 explicit HeartbeatDispatcher(OSD
*o
) : Dispatcher(o
->cct
), osd(o
) {}
1548 bool ms_can_fast_dispatch_any() const override
{ return true; }
1549 bool ms_can_fast_dispatch(const Message
*m
) const override
{
1550 switch (m
->get_type()) {
1558 void ms_fast_dispatch(Message
*m
) override
{
1559 osd
->heartbeat_dispatch(m
);
1561 bool ms_dispatch(Message
*m
) override
{
1562 return osd
->heartbeat_dispatch(m
);
1564 bool ms_handle_reset(Connection
*con
) override
{
1565 return osd
->heartbeat_reset(con
);
1567 void ms_handle_remote_reset(Connection
*con
) override
{}
1568 bool ms_handle_refused(Connection
*con
) override
{
1569 return osd
->ms_handle_refused(con
);
1571 bool ms_verify_authorizer(Connection
*con
, int peer_type
,
1572 int protocol
, bufferlist
& authorizer_data
, bufferlist
& authorizer_reply
,
1573 bool& isvalid
, CryptoKey
& session_key
) override
{
1577 } heartbeat_dispatcher
;
1581 list
<OpRequestRef
> finished
;
1583 void take_waiters(list
<OpRequestRef
>& ls
) {
1584 assert(osd_lock
.is_locked());
1585 finished
.splice(finished
.end(), ls
);
1589 // -- op tracking --
1590 OpTracker op_tracker
;
1591 void check_ops_in_flight();
1592 void test_ops(std::string command
, std::string args
, ostream
& ss
);
1593 friend class TestOpsSocketHook
;
1594 TestOpsSocketHook
*test_ops_hook
;
1595 friend struct C_CompleteSplits
;
1596 friend struct C_OpenPGs
;
1599 enum class io_queue
{
1605 friend std::ostream
& operator<<(std::ostream
& out
, const OSD::io_queue
& q
);
1607 const io_queue op_queue
;
1608 const unsigned int op_prio_cutoff
;
1611 * The ordered op delivery chain is:
1613 * fast dispatch -> pqueue back
1614 * pqueue front <-> to_process back
1615 * to_process front -> RunVis(item)
1618 * The pqueue is per-shard, and to_process is per pg_slot. Items can be
1619 * pushed back up into to_process and/or pqueue while order is preserved.
1621 * Multiple worker threads can operate on each shard.
1623 * Under normal circumstances, num_running == to_proces.size(). There are
1624 * two times when that is not true: (1) when waiting_for_pg == true and
1625 * to_process is accumulating requests that are waiting for the pg to be
1626 * instantiated; in that case they will all get requeued together by
1627 * wake_pg_waiters, and (2) when wake_pg_waiters just ran, waiting_for_pg
1628 * and already requeued the items.
1630 friend class PGQueueable
;
1633 : public ShardedThreadPool::ShardedWQ
<pair
<spg_t
,PGQueueable
>>
1639 Mutex sdata_op_ordering_lock
; ///< protects all members below
1641 OSDMapRef waiting_for_pg_osdmap
;
1643 PGRef pg
; ///< cached pg reference [optional]
1644 list
<PGQueueable
> to_process
; ///< order items for this slot
1645 int num_running
= 0; ///< _process threads doing pg lookup/lock
1647 /// true if pg does/did not exist. if so all new items go directly to
1648 /// to_process. cleared by prune_pg_waiters.
1649 bool waiting_for_pg
= false;
1651 /// incremented by wake_pg_waiters; indicates racing _process threads
1652 /// should bail out (their op has been requeued)
1653 uint64_t requeue_seq
= 0;
1656 /// map of slots for each spg_t. maintains ordering of items dequeued
1657 /// from pqueue while _process thread drops shard lock to acquire the
1658 /// pg lock. slots are removed only by prune_pg_waiters.
1659 unordered_map
<spg_t
,pg_slot
> pg_slots
;
1662 std::unique_ptr
<OpQueue
< pair
<spg_t
, PGQueueable
>, entity_inst_t
>> pqueue
;
1664 void _enqueue_front(pair
<spg_t
, PGQueueable
> item
, unsigned cutoff
) {
1665 unsigned priority
= item
.second
.get_priority();
1666 unsigned cost
= item
.second
.get_cost();
1667 if (priority
>= cutoff
)
1668 pqueue
->enqueue_strict_front(
1669 item
.second
.get_owner(),
1672 pqueue
->enqueue_front(
1673 item
.second
.get_owner(),
1674 priority
, cost
, item
);
1678 string lock_name
, string ordering_lock
,
1679 uint64_t max_tok_per_prio
, uint64_t min_cost
, CephContext
*cct
,
1681 : sdata_lock(lock_name
.c_str(), false, true, false, cct
),
1682 sdata_op_ordering_lock(ordering_lock
.c_str(), false, true,
1684 if (opqueue
== io_queue::weightedpriority
) {
1685 pqueue
= std::unique_ptr
1686 <WeightedPriorityQueue
<pair
<spg_t
,PGQueueable
>,entity_inst_t
>>(
1687 new WeightedPriorityQueue
<pair
<spg_t
,PGQueueable
>,entity_inst_t
>(
1688 max_tok_per_prio
, min_cost
));
1689 } else if (opqueue
== io_queue::prioritized
) {
1690 pqueue
= std::unique_ptr
1691 <PrioritizedQueue
<pair
<spg_t
,PGQueueable
>,entity_inst_t
>>(
1692 new PrioritizedQueue
<pair
<spg_t
,PGQueueable
>,entity_inst_t
>(
1693 max_tok_per_prio
, min_cost
));
1694 } else if (opqueue
== io_queue::mclock_opclass
) {
1695 pqueue
= std::unique_ptr
1696 <ceph::mClockOpClassQueue
>(new ceph::mClockOpClassQueue(cct
));
1697 } else if (opqueue
== io_queue::mclock_client
) {
1698 pqueue
= std::unique_ptr
1699 <ceph::mClockClientQueue
>(new ceph::mClockClientQueue(cct
));
1702 }; // struct ShardData
1704 vector
<ShardData
*> shard_list
;
1706 uint32_t num_shards
;
1709 ShardedOpWQ(uint32_t pnum_shards
,
1713 ShardedThreadPool
* tp
)
1714 : ShardedThreadPool::ShardedWQ
<pair
<spg_t
,PGQueueable
>>(ti
, si
, tp
),
1716 num_shards(pnum_shards
) {
1717 for (uint32_t i
= 0; i
< num_shards
; i
++) {
1718 char lock_name
[32] = {0};
1719 snprintf(lock_name
, sizeof(lock_name
), "%s.%d", "OSD:ShardedOpWQ:", i
);
1720 char order_lock
[32] = {0};
1721 snprintf(order_lock
, sizeof(order_lock
), "%s.%d",
1722 "OSD:ShardedOpWQ:order:", i
);
1723 ShardData
* one_shard
= new ShardData(
1724 lock_name
, order_lock
,
1725 osd
->cct
->_conf
->osd_op_pq_max_tokens_per_priority
,
1726 osd
->cct
->_conf
->osd_op_pq_min_cost
, osd
->cct
, osd
->op_queue
);
1727 shard_list
.push_back(one_shard
);
1730 ~ShardedOpWQ() override
{
1731 while (!shard_list
.empty()) {
1732 delete shard_list
.back();
1733 shard_list
.pop_back();
1737 /// wake any pg waiters after a PG is created/instantiated
1738 void wake_pg_waiters(spg_t pgid
);
1740 /// prune ops (and possiblye pg_slots) for pgs that shouldn't be here
1741 void prune_pg_waiters(OSDMapRef osdmap
, int whoami
);
1743 /// clear cached PGRef on pg deletion
1744 void clear_pg_pointer(spg_t pgid
);
1746 /// clear pg_slots on shutdown
1747 void clear_pg_slots();
1749 /// try to do some work
1750 void _process(uint32_t thread_index
, heartbeat_handle_d
*hb
) override
;
1752 /// enqueue a new item
1753 void _enqueue(pair
<spg_t
, PGQueueable
> item
) override
;
1755 /// requeue an old item (at the front of the line)
1756 void _enqueue_front(pair
<spg_t
, PGQueueable
> item
) override
;
1758 void return_waiting_threads() override
{
1759 for(uint32_t i
= 0; i
< num_shards
; i
++) {
1760 ShardData
* sdata
= shard_list
[i
];
1761 assert (NULL
!= sdata
);
1762 sdata
->sdata_lock
.Lock();
1763 sdata
->sdata_cond
.Signal();
1764 sdata
->sdata_lock
.Unlock();
1768 void dump(Formatter
*f
) {
1769 for(uint32_t i
= 0; i
< num_shards
; i
++) {
1770 ShardData
* sdata
= shard_list
[i
];
1771 char lock_name
[32] = {0};
1772 snprintf(lock_name
, sizeof(lock_name
), "%s%d", "OSD:ShardedOpWQ:", i
);
1773 assert (NULL
!= sdata
);
1774 sdata
->sdata_op_ordering_lock
.Lock();
1775 f
->open_object_section(lock_name
);
1776 sdata
->pqueue
->dump(f
);
1778 sdata
->sdata_op_ordering_lock
.Unlock();
1782 /// Must be called on ops queued back to front
1785 list
<OpRequestRef
> *out_ops
;
1786 uint64_t reserved_pushes_to_free
;
1787 Pred(spg_t pg
, list
<OpRequestRef
> *out_ops
= 0)
1788 : pgid(pg
), out_ops(out_ops
), reserved_pushes_to_free(0) {}
1789 void accumulate(const PGQueueable
&op
) {
1790 reserved_pushes_to_free
+= op
.get_reserved_pushes();
1792 boost::optional
<OpRequestRef
> mop
= op
.maybe_get_op();
1794 out_ops
->push_front(*mop
);
1797 bool operator()(const pair
<spg_t
, PGQueueable
> &op
) {
1798 if (op
.first
== pgid
) {
1799 accumulate(op
.second
);
1805 uint64_t get_reserved_pushes_to_free() const {
1806 return reserved_pushes_to_free
;
1810 bool is_shard_empty(uint32_t thread_index
) override
{
1811 uint32_t shard_index
= thread_index
% num_shards
;
1812 ShardData
* sdata
= shard_list
[shard_index
];
1813 assert(NULL
!= sdata
);
1814 Mutex::Locker
l(sdata
->sdata_op_ordering_lock
);
1815 return sdata
->pqueue
->empty();
1820 void enqueue_op(spg_t pg
, OpRequestRef
& op
, epoch_t epoch
);
1822 PGRef pg
, OpRequestRef op
,
1823 ThreadPool::TPHandle
&handle
);
1825 // -- peering queue --
1826 struct PeeringWQ
: public ThreadPool::BatchWorkQueue
<PG
> {
1827 list
<PG
*> peering_queue
;
1830 PeeringWQ(OSD
*o
, time_t ti
, time_t si
, ThreadPool
*tp
)
1831 : ThreadPool::BatchWorkQueue
<PG
>(
1832 "OSD::PeeringWQ", ti
, si
, tp
), osd(o
) {}
1834 void _dequeue(PG
*pg
) override
{
1835 for (list
<PG
*>::iterator i
= peering_queue
.begin();
1836 i
!= peering_queue
.end();
1839 peering_queue
.erase(i
++);
1840 pg
->put("PeeringWQ");
1846 bool _enqueue(PG
*pg
) override
{
1847 pg
->get("PeeringWQ");
1848 peering_queue
.push_back(pg
);
1851 bool _empty() override
{
1852 return peering_queue
.empty();
1854 void _dequeue(list
<PG
*> *out
) override
;
1856 const list
<PG
*> &pgs
,
1857 ThreadPool::TPHandle
&handle
) override
{
1858 assert(!pgs
.empty());
1859 osd
->process_peering_events(pgs
, handle
);
1860 for (list
<PG
*>::const_iterator i
= pgs
.begin();
1863 (*i
)->put("PeeringWQ");
1866 void _process_finish(const list
<PG
*> &pgs
) override
{
1867 for (list
<PG
*>::const_iterator i
= pgs
.begin();
1873 void _clear() override
{
1874 assert(peering_queue
.empty());
1878 void process_peering_events(
1879 const list
<PG
*> &pg
,
1880 ThreadPool::TPHandle
&handle
);
1883 friend class PrimaryLogPG
;
1890 OSDMapRef
get_osdmap() {
1893 epoch_t
get_osdmap_epoch() const {
1894 return osdmap
? osdmap
->get_epoch() : 0;
1897 utime_t had_map_since
;
1899 list
<OpRequestRef
> waiting_for_osdmap
;
1900 deque
<utime_t
> osd_markdown_log
;
1902 friend struct send_map_on_destruct
;
1904 void wait_for_new_map(OpRequestRef op
);
1905 void handle_osd_map(class MOSDMap
*m
);
1906 void _committed_osd_maps(epoch_t first
, epoch_t last
, class MOSDMap
*m
);
1907 void trim_maps(epoch_t oldest
, int nreceived
, bool skip_maps
);
1908 void note_down_osd(int osd
);
1909 void note_up_osd(int osd
);
1910 friend class C_OnMapCommit
;
1913 epoch_t advance_to
, PG
*pg
,
1914 ThreadPool::TPHandle
&handle
,
1915 PG::RecoveryCtx
*rctx
,
1916 set
<PGRef
> *split_pgs
1919 void activate_map();
1921 // osd map cache (past osd maps)
1922 OSDMapRef
get_map(epoch_t e
) {
1923 return service
.get_map(e
);
1925 OSDMapRef
add_map(OSDMap
*o
) {
1926 return service
.add_map(o
);
1928 void add_map_bl(epoch_t e
, bufferlist
& bl
) {
1929 return service
.add_map_bl(e
, bl
);
1931 void pin_map_bl(epoch_t e
, bufferlist
&bl
) {
1932 return service
.pin_map_bl(e
, bl
);
1934 bool get_map_bl(epoch_t e
, bufferlist
& bl
) {
1935 return service
.get_map_bl(e
, bl
);
1937 void add_map_inc_bl(epoch_t e
, bufferlist
& bl
) {
1938 return service
.add_map_inc_bl(e
, bl
);
1940 void pin_map_inc_bl(epoch_t e
, bufferlist
&bl
) {
1941 return service
.pin_map_inc_bl(e
, bl
);
1945 // -- placement groups --
1946 RWLock pg_map_lock
; // this lock orders *above* individual PG _locks
1947 ceph::unordered_map
<spg_t
, PG
*> pg_map
; // protected by pg_map lock
1949 std::mutex pending_creates_lock
;
1950 std::set
<pg_t
> pending_creates_from_osd
;
1951 unsigned pending_creates_from_mon
= 0;
1953 map
<spg_t
, list
<PG::CephPeeringEvtRef
> > peering_wait_for_split
;
1954 PGRecoveryStats pg_recovery_stats
;
1956 PGPool
_get_pool(int id
, OSDMapRef createmap
);
1958 PG
*_lookup_lock_pg_with_map_lock_held(spg_t pgid
);
1959 PG
*_lookup_lock_pg(spg_t pgid
);
1962 PG
*lookup_lock_pg(spg_t pgid
);
1965 RWLock::RLocker
l(pg_map_lock
);
1966 return pg_map
.size();
1970 PG
*_open_lock_pg(OSDMapRef createmap
,
1971 spg_t pg
, bool no_lockdep_check
=false);
1973 RES_PARENT
, // resurrected a parent
1974 RES_SELF
, // resurrected self
1975 RES_NONE
// nothing relevant deleting
1977 res_result
_try_resurrect_pg(
1978 OSDMapRef curmap
, spg_t pgid
, spg_t
*resurrected
, PGRef
*old_pg_state
);
1980 PG
*_create_lock_pg(
1981 OSDMapRef createmap
,
1986 vector
<int>& up
, int up_primary
,
1987 vector
<int>& acting
, int acting_primary
,
1988 pg_history_t history
,
1989 const PastIntervals
& pi
,
1990 ObjectStore::Transaction
& t
);
1992 PG
* _make_pg(OSDMapRef createmap
, spg_t pgid
);
1993 void add_newly_split_pg(PG
*pg
,
1994 PG::RecoveryCtx
*rctx
);
1996 int handle_pg_peering_evt(
1998 const pg_history_t
& orig_history
,
1999 const PastIntervals
& pi
,
2001 PG::CephPeeringEvtRef evt
);
2002 bool maybe_wait_for_max_pg(spg_t pgid
, bool is_mon_create
);
2003 void resume_creating_pg();
2006 void build_past_intervals_parallel();
2008 /// build initial pg history and intervals on create
2009 void build_initial_pg_history(
2012 utime_t created_stamp
,
2016 /// project pg history from from to now
2017 bool project_pg_history(
2018 spg_t pgid
, pg_history_t
& h
, epoch_t from
,
2019 const vector
<int>& lastup
,
2021 const vector
<int>& lastacting
,
2022 int lastactingprimary
2023 ); ///< @return false if there was a map gap between from and now
2025 // this must be called with pg->lock held on any pg addition to pg_map
2026 void wake_pg_waiters(PGRef pg
) {
2027 assert(pg
->is_locked());
2028 op_shardedwq
.wake_pg_waiters(pg
->info
.pgid
);
2030 epoch_t last_pg_create_epoch
;
2032 void handle_pg_create(OpRequestRef op
);
2036 const set
<spg_t
> &childpgids
, set
<PGRef
> *out_pgs
,
2039 PG::RecoveryCtx
*rctx
);
2041 // == monitor interaction ==
2042 Mutex mon_report_lock
;
2043 utime_t last_mon_report
;
2044 utime_t last_pg_stats_sent
;
2046 /* if our monitor dies, we want to notice it and reconnect.
2047 * So we keep track of when it last acked our stat updates,
2048 * and if too much time passes (and we've been sending
2049 * more updates) then we can call it dead and reconnect
2052 utime_t last_pg_stats_ack
;
2053 float stats_ack_timeout
;
2054 set
<uint64_t> outstanding_pg_stats
; // how many stat updates haven't been acked yet
2058 void _got_mon_epochs(epoch_t oldest
, epoch_t newest
);
2059 void _preboot(epoch_t oldest
, epoch_t newest
);
2061 void _collect_metadata(map
<string
,string
> *pmeta
);
2063 void start_waiting_for_healthy();
2066 void send_full_update();
2068 friend struct C_OSD_GetVersion
;
2071 epoch_t up_thru_wanted
;
2073 void queue_want_up_thru(epoch_t want
);
2076 // -- full map requests --
2077 epoch_t requested_full_first
, requested_full_last
;
2079 void request_full_map(epoch_t first
, epoch_t last
);
2080 void rerequest_full_maps() {
2081 epoch_t first
= requested_full_first
;
2082 epoch_t last
= requested_full_last
;
2083 requested_full_first
= 0;
2084 requested_full_last
= 0;
2085 request_full_map(first
, last
);
2087 void got_full_map(epoch_t e
);
2090 map
<int,utime_t
> failure_queue
;
2091 map
<int,pair
<utime_t
,entity_inst_t
> > failure_pending
;
2093 void requeue_failures();
2094 void send_failures();
2095 void send_still_alive(epoch_t epoch
, const entity_inst_t
&i
);
2098 Mutex pg_stat_queue_lock
;
2099 Cond pg_stat_queue_cond
;
2100 xlist
<PG
*> pg_stat_queue
;
2101 bool osd_stat_updated
;
2102 uint64_t pg_stat_tid
, pg_stat_tid_flushed
;
2104 void send_pg_stats(const utime_t
&now
);
2105 void handle_pg_stats_ack(class MPGStatsAck
*ack
);
2106 void flush_pg_stats();
2108 ceph::coarse_mono_clock::time_point last_sent_beacon
;
2109 Mutex min_last_epoch_clean_lock
{"OSD::min_last_epoch_clean_lock"};
2110 epoch_t min_last_epoch_clean
= 0;
2111 // which pgs were scanned for min_lec
2112 std::vector
<pg_t
> min_last_epoch_clean_pgs
;
2113 void send_beacon(const ceph::coarse_mono_clock::time_point
& now
);
2115 void pg_stat_queue_enqueue(PG
*pg
) {
2116 pg_stat_queue_lock
.Lock();
2117 if (pg
->is_primary() && !pg
->stat_queue_item
.is_on_list()) {
2118 pg
->get("pg_stat_queue");
2119 pg_stat_queue
.push_back(&pg
->stat_queue_item
);
2121 osd_stat_updated
= true;
2122 pg_stat_queue_lock
.Unlock();
2124 void pg_stat_queue_dequeue(PG
*pg
) {
2125 pg_stat_queue_lock
.Lock();
2126 if (pg
->stat_queue_item
.remove_myself())
2127 pg
->put("pg_stat_queue");
2128 pg_stat_queue_lock
.Unlock();
2130 void clear_pg_stat_queue() {
2131 pg_stat_queue_lock
.Lock();
2132 while (!pg_stat_queue
.empty()) {
2133 PG
*pg
= pg_stat_queue
.front();
2134 pg_stat_queue
.pop_front();
2135 pg
->put("pg_stat_queue");
2137 pg_stat_queue_lock
.Unlock();
2139 void clear_outstanding_pg_stats(){
2140 Mutex::Locker
l(pg_stat_queue_lock
);
2141 outstanding_pg_stats
.clear();
2144 ceph_tid_t
get_tid() {
2145 return service
.get_tid();
2148 // -- generic pg peering --
2149 PG::RecoveryCtx
create_context();
2150 void dispatch_context(PG::RecoveryCtx
&ctx
, PG
*pg
, OSDMapRef curmap
,
2151 ThreadPool::TPHandle
*handle
= NULL
);
2152 void dispatch_context_transaction(PG::RecoveryCtx
&ctx
, PG
*pg
,
2153 ThreadPool::TPHandle
*handle
= NULL
);
2154 void do_notifies(map
<int,
2155 vector
<pair
<pg_notify_t
, PastIntervals
> > >&
2158 void do_queries(map
<int, map
<spg_t
,pg_query_t
> >& query_map
,
2160 void do_infos(map
<int,
2161 vector
<pair
<pg_notify_t
, PastIntervals
> > >& info_map
,
2164 bool require_mon_peer(const Message
*m
);
2165 bool require_mon_or_mgr_peer(const Message
*m
);
2166 bool require_osd_peer(const Message
*m
);
2168 * Verifies that we were alive in the given epoch, and that
2171 bool require_self_aliveness(const Message
*m
, epoch_t alive_since
);
2173 * Verifies that the OSD who sent the given op has the same
2174 * address as in the given map.
2175 * @pre op was sent by an OSD using the cluster messenger
2177 bool require_same_peer_instance(const Message
*m
, OSDMapRef
& map
,
2178 bool is_fast_dispatch
);
2180 bool require_same_or_newer_map(OpRequestRef
& op
, epoch_t e
,
2181 bool is_fast_dispatch
);
2183 void handle_pg_query(OpRequestRef op
);
2184 void handle_pg_notify(OpRequestRef op
);
2185 void handle_pg_log(OpRequestRef op
);
2186 void handle_pg_info(OpRequestRef op
);
2187 void handle_pg_trim(OpRequestRef op
);
2189 void handle_pg_backfill_reserve(OpRequestRef op
);
2190 void handle_pg_recovery_reserve(OpRequestRef op
);
2192 void handle_force_recovery(Message
*m
);
2194 void handle_pg_remove(OpRequestRef op
);
2195 void _remove_pg(PG
*pg
);
2204 Command(vector
<string
>& c
, ceph_tid_t t
, bufferlist
& bl
, Connection
*co
)
2205 : cmd(c
), tid(t
), indata(bl
), con(co
) {}
2207 list
<Command
*> command_queue
;
2208 struct CommandWQ
: public ThreadPool::WorkQueue
<Command
> {
2210 CommandWQ(OSD
*o
, time_t ti
, time_t si
, ThreadPool
*tp
)
2211 : ThreadPool::WorkQueue
<Command
>("OSD::CommandWQ", ti
, si
, tp
), osd(o
) {}
2213 bool _empty() override
{
2214 return osd
->command_queue
.empty();
2216 bool _enqueue(Command
*c
) override
{
2217 osd
->command_queue
.push_back(c
);
2220 void _dequeue(Command
*pg
) override
{
2223 Command
*_dequeue() override
{
2224 if (osd
->command_queue
.empty())
2226 Command
*c
= osd
->command_queue
.front();
2227 osd
->command_queue
.pop_front();
2230 void _process(Command
*c
, ThreadPool::TPHandle
&) override
{
2231 osd
->osd_lock
.Lock();
2232 if (osd
->is_stopping()) {
2233 osd
->osd_lock
.Unlock();
2237 osd
->do_command(c
->con
.get(), c
->tid
, c
->cmd
, c
->indata
);
2238 osd
->osd_lock
.Unlock();
2241 void _clear() override
{
2242 while (!osd
->command_queue
.empty()) {
2243 Command
*c
= osd
->command_queue
.front();
2244 osd
->command_queue
.pop_front();
2250 void handle_command(class MMonCommand
*m
);
2251 void handle_command(class MCommand
*m
);
2252 void do_command(Connection
*con
, ceph_tid_t tid
, vector
<string
>& cmd
, bufferlist
& data
);
2254 // -- pg recovery --
2255 void do_recovery(PG
*pg
, epoch_t epoch_queued
, uint64_t pushes_reserved
,
2256 ThreadPool::TPHandle
&handle
);
2261 bool scrub_random_backoff();
2262 bool scrub_load_below_threshold();
2263 bool scrub_time_permit(utime_t now
);
2267 public ThreadPool::WorkQueueVal
<pair
<PGRef
, DeletingStateRef
> > {
2269 ObjectStore
*&store
;
2270 list
<pair
<PGRef
, DeletingStateRef
> > remove_queue
;
2271 RemoveWQ(CephContext
* cct
, ObjectStore
*&o
, time_t ti
, time_t si
,
2273 : ThreadPool::WorkQueueVal
<pair
<PGRef
, DeletingStateRef
> >(
2274 "OSD::RemoveWQ", ti
, si
, tp
), cct(cct
), store(o
) {}
2276 bool _empty() override
{
2277 return remove_queue
.empty();
2279 void _enqueue(pair
<PGRef
, DeletingStateRef
> item
) override
{
2280 remove_queue
.push_back(item
);
2282 void _enqueue_front(pair
<PGRef
, DeletingStateRef
> item
) override
{
2283 remove_queue
.push_front(item
);
2285 bool _dequeue(pair
<PGRef
, DeletingStateRef
> item
) {
2288 pair
<PGRef
, DeletingStateRef
> _dequeue() override
{
2289 assert(!remove_queue
.empty());
2290 pair
<PGRef
, DeletingStateRef
> item
= remove_queue
.front();
2291 remove_queue
.pop_front();
2294 void _process(pair
<PGRef
, DeletingStateRef
>,
2295 ThreadPool::TPHandle
&) override
;
2296 void _clear() override
{
2297 remove_queue
.clear();
2302 bool ms_can_fast_dispatch_any() const override
{ return true; }
2303 bool ms_can_fast_dispatch(const Message
*m
) const override
{
2304 switch (m
->get_type()) {
2305 case CEPH_MSG_OSD_OP
:
2306 case CEPH_MSG_OSD_BACKOFF
:
2309 case MSG_OSD_SUBOPREPLY
:
2310 case MSG_OSD_REPOPREPLY
:
2311 case MSG_OSD_PG_PUSH
:
2312 case MSG_OSD_PG_PULL
:
2313 case MSG_OSD_PG_PUSH_REPLY
:
2314 case MSG_OSD_PG_SCAN
:
2315 case MSG_OSD_PG_BACKFILL
:
2316 case MSG_OSD_PG_BACKFILL_REMOVE
:
2317 case MSG_OSD_EC_WRITE
:
2318 case MSG_OSD_EC_WRITE_REPLY
:
2319 case MSG_OSD_EC_READ
:
2320 case MSG_OSD_EC_READ_REPLY
:
2321 case MSG_OSD_SCRUB_RESERVE
:
2322 case MSG_OSD_REP_SCRUB
:
2323 case MSG_OSD_REP_SCRUBMAP
:
2324 case MSG_OSD_PG_UPDATE_LOG_MISSING
:
2325 case MSG_OSD_PG_UPDATE_LOG_MISSING_REPLY
:
2326 case MSG_OSD_PG_RECOVERY_DELETE
:
2327 case MSG_OSD_PG_RECOVERY_DELETE_REPLY
:
2333 void ms_fast_dispatch(Message
*m
) override
;
2334 void ms_fast_preprocess(Message
*m
) override
;
2335 bool ms_dispatch(Message
*m
) override
;
2336 bool ms_get_authorizer(int dest_type
, AuthAuthorizer
**authorizer
, bool force_new
) override
;
2337 bool ms_verify_authorizer(Connection
*con
, int peer_type
,
2338 int protocol
, bufferlist
& authorizer
, bufferlist
& authorizer_reply
,
2339 bool& isvalid
, CryptoKey
& session_key
) override
;
2340 void ms_handle_connect(Connection
*con
) override
;
2341 void ms_handle_fast_connect(Connection
*con
) override
;
2342 void ms_handle_fast_accept(Connection
*con
) override
;
2343 bool ms_handle_reset(Connection
*con
) override
;
2344 void ms_handle_remote_reset(Connection
*con
) override
{}
2345 bool ms_handle_refused(Connection
*con
) override
;
2347 io_queue
get_io_queue() const {
2348 if (cct
->_conf
->osd_op_queue
== "debug_random") {
2349 static io_queue index_lookup
[] = { io_queue::prioritized
,
2350 io_queue::weightedpriority
,
2351 io_queue::mclock_opclass
,
2352 io_queue::mclock_client
};
2354 unsigned which
= rand() % (sizeof(index_lookup
) / sizeof(index_lookup
[0]));
2355 return index_lookup
[which
];
2356 } else if (cct
->_conf
->osd_op_queue
== "prioritized") {
2357 return io_queue::prioritized
;
2358 } else if (cct
->_conf
->osd_op_queue
== "mclock_opclass") {
2359 return io_queue::mclock_opclass
;
2360 } else if (cct
->_conf
->osd_op_queue
== "mclock_client") {
2361 return io_queue::mclock_client
;
2363 // default / catch-all is 'wpq'
2364 return io_queue::weightedpriority
;
2368 unsigned int get_io_prio_cut() const {
2369 if (cct
->_conf
->osd_op_queue_cut_off
== "debug_random") {
2371 return (rand() % 2 < 1) ? CEPH_MSG_PRIO_HIGH
: CEPH_MSG_PRIO_LOW
;
2372 } else if (cct
->_conf
->osd_op_queue_cut_off
== "high") {
2373 return CEPH_MSG_PRIO_HIGH
;
2375 // default / catch-all is 'low'
2376 return CEPH_MSG_PRIO_LOW
;
2381 /* internal and external can point to the same messenger, they will still
2382 * be cleaned up properly*/
2383 OSD(CephContext
*cct_
,
2384 ObjectStore
*store_
,
2386 Messenger
*internal
,
2387 Messenger
*external
,
2388 Messenger
*hb_front_client
,
2389 Messenger
*hb_back_client
,
2390 Messenger
*hb_front_server
,
2391 Messenger
*hb_back_server
,
2392 Messenger
*osdc_messenger
,
2393 MonClient
*mc
, const std::string
&dev
, const std::string
&jdev
);
2397 static int mkfs(CephContext
*cct
, ObjectStore
*store
,
2399 uuid_d fsid
, int whoami
);
2400 /* remove any non-user xattrs from a map of them */
2401 void filter_xattrs(map
<string
, bufferptr
>& attrs
) {
2402 for (map
<string
, bufferptr
>::iterator iter
= attrs
.begin();
2403 iter
!= attrs
.end();
2405 if (('_' != iter
->first
.at(0)) || (iter
->first
.size() == 1))
2406 attrs
.erase(iter
++);
2412 int mon_cmd_maybe_osd_create(string
&cmd
);
2413 int update_crush_device_class();
2414 int update_crush_location();
2416 static int write_meta(CephContext
*cct
,
2418 uuid_d
& cluster_fsid
, uuid_d
& osd_fsid
, int whoami
);
2420 void handle_pg_scrub(struct MOSDScrub
*m
, PG
* pg
);
2421 void handle_scrub(struct MOSDScrub
*m
);
2422 void handle_osd_ping(class MOSDPing
*m
);
2424 int init_op_flags(OpRequestRef
& op
);
2426 int get_num_op_shards();
2427 int get_num_op_threads();
2429 float get_osd_recovery_sleep();
2432 static int peek_meta(ObjectStore
*store
, string
& magic
,
2433 uuid_d
& cluster_fsid
, uuid_d
& osd_fsid
, int& whoami
);
2441 int enable_disable_fuse(bool stop
);
2443 void suicide(int exitcode
);
2446 void handle_signal(int signum
);
2448 /// check if we can throw out op from a disconnected client
2449 static bool op_is_discardable(const MOSDOp
*m
);
2453 friend class OSDService
;
2457 std::ostream
& operator<<(std::ostream
& out
, const OSD::io_queue
& q
);
2460 //compatibility of the executable
2461 extern const CompatSet::Feature ceph_osd_feature_compat
[];
2462 extern const CompatSet::Feature ceph_osd_feature_ro_compat
[];
2463 extern const CompatSet::Feature ceph_osd_feature_incompat
[];
2465 #endif // CEPH_OSD_H