/
rollback_impl.cpp
1407 lines (1256 loc) · 66.9 KB
/
rollback_impl.cpp
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
/**
* Copyright (C) 2018-present MongoDB, Inc.
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the Server Side Public License, version 1,
* as published by MongoDB, Inc.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* Server Side Public License for more details.
*
* You should have received a copy of the Server Side Public License
* along with this program. If not, see
* <http://www.mongodb.com/licensing/server-side-public-license>.
*
* As a special exception, the copyright holders give permission to link the
* code of portions of this program with the OpenSSL library under certain
* conditions as described in each individual source file and distribute
* linked combinations including the program with the OpenSSL library. You
* must comply with the Server Side Public License in all respects for
* all of the code used other than as permitted herein. If you modify file(s)
* with this exception, you may extend this exception to your version of the
* file(s), but you are not obligated to do so. If you do not wish to do so,
* delete this exception statement from your version. If you delete this
* exception statement from all source files in the program, then also delete
* it in the license file.
*/
#define MONGO_LOGV2_DEFAULT_COMPONENT ::mongo::logv2::LogComponent::kReplicationRollback
#include "mongo/platform/basic.h"
#include "mongo/db/repl/rollback_impl.h"
#include "mongo/db/repl/rollback_impl_gen.h"
#include <fmt/format.h>
#include "mongo/bson/util/bson_extract.h"
#include "mongo/db/catalog/collection_catalog.h"
#include "mongo/db/catalog/database_holder.h"
#include "mongo/db/catalog/import_collection_oplog_entry_gen.h"
#include "mongo/db/commands.h"
#include "mongo/db/concurrency/d_concurrency.h"
#include "mongo/db/concurrency/replication_state_transition_lock_guard.h"
#include "mongo/db/concurrency/write_conflict_exception.h"
#include "mongo/db/db_raii.h"
#include "mongo/db/dbdirectclient.h"
#include "mongo/db/dbhelpers.h"
#include "mongo/db/index_builds_coordinator.h"
#include "mongo/db/kill_sessions_local.h"
#include "mongo/db/logical_time_validator.h"
#include "mongo/db/operation_context.h"
#include "mongo/db/repl/apply_ops.h"
#include "mongo/db/repl/drop_pending_collection_reaper.h"
#include "mongo/db/repl/replication_coordinator.h"
#include "mongo/db/repl/replication_process.h"
#include "mongo/db/repl/roll_back_local_operations.h"
#include "mongo/db/repl/storage_interface.h"
#include "mongo/db/repl/tenant_migration_access_blocker_util.h"
#include "mongo/db/repl/transaction_oplog_application.h"
#include "mongo/db/s/type_shard_identity.h"
#include "mongo/db/server_recovery.h"
#include "mongo/db/session_catalog_mongod.h"
#include "mongo/db/session_txn_record_gen.h"
#include "mongo/db/storage/remove_saver.h"
#include "mongo/db/transaction_history_iterator.h"
#include "mongo/logv2/log.h"
#include "mongo/s/catalog/type_config_version.h"
#include "mongo/util/scopeguard.h"
namespace mongo {
namespace repl {
using namespace fmt::literals;
MONGO_FAIL_POINT_DEFINE(rollbackHangAfterTransitionToRollback);
namespace {
// Used to set RollbackImpl::_newCounts to force a collection scan to fix count.
constexpr long long kCollectionScanRequired = -1;
RollbackImpl::Listener kNoopListener;
// The name of the insert, update and delete commands as found in oplog command entries.
constexpr auto kInsertCmdName = "insert"_sd;
constexpr auto kUpdateCmdName = "update"_sd;
constexpr auto kDeleteCmdName = "delete"_sd;
constexpr auto kNumRecordsFieldName = "numRecords"_sd;
constexpr auto kToFieldName = "to"_sd;
constexpr auto kDropTargetFieldName = "dropTarget"_sd;
/**
* Parses the o2 field of a drop or rename oplog entry for the count of the collection that was
* dropped.
*/
boost::optional<long long> _parseDroppedCollectionCount(const OplogEntry& oplogEntry) {
auto commandType = oplogEntry.getCommandType();
auto desc = OplogEntry::CommandType::kDrop == commandType ? "drop"_sd : "rename"_sd;
auto obj2 = oplogEntry.getObject2();
if (!obj2) {
LOGV2_WARNING(21634,
"Unable to get collection count from oplog entry without the o2 field",
"type"_attr = desc,
"oplogEntry"_attr = redact(oplogEntry.toBSONForLogging()));
return boost::none;
}
long long count = 0;
// TODO: Use IDL to parse o2 object. See txn_cmds.idl for example.
auto status = bsonExtractIntegerField(*obj2, kNumRecordsFieldName, &count);
if (!status.isOK()) {
LOGV2_WARNING(21635,
"Failed to parse oplog entry for collection count",
"type"_attr = desc,
"error"_attr = status,
"oplogEntry"_attr = redact(oplogEntry.toBSONForLogging()));
return boost::none;
}
if (count < 0) {
LOGV2_WARNING(21636,
"Invalid collection count found in oplog entry",
"type"_attr = desc,
"count"_attr = count,
"oplogEntry"_attr = redact(oplogEntry.toBSONForLogging()));
return boost::none;
}
LOGV2_DEBUG(21590,
2,
"Parsed collection count of oplog entry",
"count"_attr = count,
"type"_attr = desc,
"oplogEntry"_attr = redact(oplogEntry.toBSONForLogging()));
return count;
}
} // namespace
constexpr const char* RollbackImpl::kRollbackRemoveSaverType;
constexpr const char* RollbackImpl::kRollbackRemoveSaverWhy;
bool RollbackImpl::shouldCreateDataFiles() {
return gCreateRollbackDataFiles.load();
}
RollbackImpl::RollbackImpl(OplogInterface* localOplog,
OplogInterface* remoteOplog,
StorageInterface* storageInterface,
ReplicationProcess* replicationProcess,
ReplicationCoordinator* replicationCoordinator,
Listener* listener)
: _listener(listener),
_localOplog(localOplog),
_remoteOplog(remoteOplog),
_storageInterface(storageInterface),
_replicationProcess(replicationProcess),
_replicationCoordinator(replicationCoordinator) {
invariant(localOplog);
invariant(remoteOplog);
invariant(storageInterface);
invariant(replicationProcess);
invariant(replicationCoordinator);
invariant(listener);
}
RollbackImpl::RollbackImpl(OplogInterface* localOplog,
OplogInterface* remoteOplog,
StorageInterface* storageInterface,
ReplicationProcess* replicationProcess,
ReplicationCoordinator* replicationCoordinator)
: RollbackImpl(localOplog,
remoteOplog,
storageInterface,
replicationProcess,
replicationCoordinator,
&kNoopListener) {}
RollbackImpl::~RollbackImpl() {
shutdown();
}
Status RollbackImpl::runRollback(OperationContext* opCtx) {
_rollbackStats.startTime = opCtx->getServiceContext()->getFastClockSource()->now();
auto status = _transitionToRollback(opCtx);
if (!status.isOK()) {
return status;
}
_listener->onTransitionToRollback();
if (MONGO_unlikely(rollbackHangAfterTransitionToRollback.shouldFail())) {
LOGV2(21591,
"rollbackHangAfterTransitionToRollback fail point enabled. Blocking until fail "
"point is disabled (rollback_impl)");
rollbackHangAfterTransitionToRollback.pauseWhileSet(opCtx);
}
// We clear the SizeRecoveryState before we recover to a stable timestamp. This ensures that we
// only use size adjustment markings from the storage and replication recovery processes in this
// rollback.
sizeRecoveryState(opCtx->getServiceContext()).clearStateBeforeRecovery();
// After successfully transitioning to the ROLLBACK state, we must always transition back to
// SECONDARY, even if we fail at any point during the rollback process.
ON_BLOCK_EXIT([this, opCtx] { _transitionFromRollbackToSecondary(opCtx); });
ON_BLOCK_EXIT([this, opCtx] { _summarizeRollback(opCtx); });
auto commonPointSW = _findCommonPoint(opCtx);
if (!commonPointSW.isOK()) {
return commonPointSW.getStatus();
}
const auto commonPoint = commonPointSW.getValue();
const OpTime commonPointOpTime = commonPoint.getOpTime();
_rollbackStats.commonPoint = commonPointOpTime;
_listener->onCommonPointFound(commonPointOpTime.getTimestamp());
// Now that we have found the common point, we make sure to proceed only if the rollback
// period is not too long.
status = _checkAgainstTimeLimit(commonPoint);
if (!status.isOK()) {
return status;
}
// Increment the Rollback ID of this node. The Rollback ID is a natural number that it is
// incremented by 1 every time a rollback occurs. Note that the Rollback ID must be incremented
// before modifying any local data.
status = _replicationProcess->incrementRollbackID(opCtx);
if (!status.isOK()) {
return status;
}
_rollbackStats.rollbackId = _replicationProcess->getRollbackID();
_listener->onRollbackIDIncremented();
// This function cannot fail without terminating the process.
_runPhaseFromAbortToReconstructPreparedTxns(opCtx, commonPoint);
_listener->onPreparedTransactionsReconstructed();
// We can now accept interruptions again.
if (_isInShutdown()) {
return Status(ErrorCodes::ShutdownInProgress, "rollback shutting down");
}
// At this point, the last applied and durable optimes on this node still point to ops on
// the divergent branch of history. We therefore update the last optimes to the top of the
// oplog, which should now be at the common point.
_replicationCoordinator->resetLastOpTimesFromOplog(opCtx);
status = _triggerOpObserver(opCtx);
if (!status.isOK()) {
return status;
}
_listener->onRollbackOpObserver(_observerInfo);
LOGV2(21592, "Rollback complete");
return Status::OK();
}
void RollbackImpl::shutdown() {
stdx::lock_guard<Latch> lock(_mutex);
_inShutdown = true;
}
bool RollbackImpl::_isInShutdown() const {
stdx::lock_guard<Latch> lock(_mutex);
return _inShutdown;
}
void RollbackImpl::_killAllUserOperations(OperationContext* opCtx) {
invariant(opCtx);
ServiceContext* serviceCtx = opCtx->getServiceContext();
invariant(serviceCtx);
int numOpsKilled = 0;
int numOpsRunning = 0;
for (ServiceContext::LockedClientsCursor cursor(serviceCtx); Client* client = cursor.next();) {
stdx::lock_guard<Client> lk(*client);
if (client->isFromSystemConnection() && !client->canKillSystemOperationInStepdown(lk)) {
continue;
}
OperationContext* toKill = client->getOperationContext();
if (toKill && toKill->getOpID() == opCtx->getOpID()) {
// Don't kill the rollback thread.
continue;
}
if (toKill && !toKill->isKillPending()) {
serviceCtx->killOperation(lk, toKill, ErrorCodes::InterruptedDueToReplStateChange);
numOpsKilled++;
} else {
numOpsRunning++;
}
}
// Update the metrics for tracking user operations during state transitions.
_replicationCoordinator->updateAndLogStateTransitionMetrics(
ReplicationCoordinator::OpsKillingStateTransitionEnum::kRollback,
numOpsKilled,
numOpsRunning);
}
Status RollbackImpl::_transitionToRollback(OperationContext* opCtx) {
invariant(opCtx);
if (_isInShutdown()) {
return Status(ErrorCodes::ShutdownInProgress, "rollback shutting down");
}
LOGV2(21593, "Transition to ROLLBACK");
{
ReplicationStateTransitionLockGuard rstlLock(
opCtx, MODE_X, ReplicationStateTransitionLockGuard::EnqueueOnly());
// Kill all user operations to ensure we can successfully acquire the RSTL. Since the node
// must be a secondary, this is only killing readers, whose connections will be closed
// shortly regardless.
_killAllUserOperations(opCtx);
rstlLock.waitForLockUntil(Date_t::max());
auto status = _replicationCoordinator->setFollowerModeRollback(opCtx);
if (!status.isOK()) {
static constexpr char message[] = "Cannot perform replica set state transition";
LOGV2(21594,
message,
"currentState"_attr = _replicationCoordinator->getMemberState().toString(),
"targetState"_attr = MemberState(MemberState::RS_ROLLBACK).toString(),
"error"_attr = status);
status.addContext(str::stream() << message << ", current state: "
<< _replicationCoordinator->getMemberState().toString()
<< ", target state: "
<< MemberState(MemberState::RS_ROLLBACK).toString());
return status;
}
}
return Status::OK();
}
void RollbackImpl::_stopAndWaitForIndexBuilds(OperationContext* opCtx) {
invariant(opCtx);
// Aborts all active, two-phase index builds.
[[maybe_unused]] auto stoppedIndexBuilds =
IndexBuildsCoordinator::get(opCtx)->stopIndexBuildsForRollback(opCtx);
// Get a list of all databases.
StorageEngine* storageEngine = opCtx->getServiceContext()->getStorageEngine();
std::vector<std::string> dbs;
{
Lock::GlobalLock lk(opCtx, MODE_IS);
dbs = storageEngine->listDatabases();
}
// Wait for all background operations to complete by waiting on each database. Single-phase
// index builds are not stopped before rollback, so we must wait for these index builds to
// complete.
std::vector<StringData> dbNames(dbs.begin(), dbs.end());
LOGV2(21595, "Waiting for all background operations to complete before starting rollback");
for (auto db : dbNames) {
auto numInProg = IndexBuildsCoordinator::get(opCtx)->numInProgForDb(db);
if (numInProg > 0) {
LOGV2_DEBUG(21596,
1,
"Waiting for {numBackgroundOperationsInProgress} "
"background operations to complete on database '{db}'",
"Waiting for background operations to complete",
"numBackgroundOperationsInProgress"_attr = numInProg,
"db"_attr = db);
IndexBuildsCoordinator::get(opCtx)->awaitNoBgOpInProgForDb(opCtx, db);
}
}
LOGV2(21597, "Finished waiting for background operations to complete before rollback");
}
StatusWith<std::set<NamespaceString>> RollbackImpl::_namespacesForOp(const OplogEntry& oplogEntry) {
NamespaceString opNss = oplogEntry.getNss();
OpTypeEnum opType = oplogEntry.getOpType();
std::set<NamespaceString> namespaces;
// No namespaces for a no-op.
if (opType == OpTypeEnum::kNoop) {
return std::set<NamespaceString>();
}
// CRUD ops have the proper namespace in the operation 'ns' field.
if (opType == OpTypeEnum::kInsert || opType == OpTypeEnum::kUpdate ||
opType == OpTypeEnum::kDelete) {
return std::set<NamespaceString>({opNss});
}
// If the operation is a command, then we need to extract the appropriate namespaces from the
// command object, as opposed to just using the 'ns' field of the oplog entry itself.
if (opType == OpTypeEnum::kCommand) {
auto obj = oplogEntry.getObject();
auto firstElem = obj.firstElement();
// Does not handle 'applyOps' entries.
invariant(oplogEntry.getCommandType() != OplogEntry::CommandType::kApplyOps,
"_namespacesForOp does not handle 'applyOps' oplog entries.");
switch (oplogEntry.getCommandType()) {
case OplogEntry::CommandType::kRenameCollection: {
// Add both the 'from' and 'to' namespaces.
namespaces.insert(NamespaceString(firstElem.valuestrsafe()));
namespaces.insert(NamespaceString(obj.getStringField("to")));
break;
}
case OplogEntry::CommandType::kDropDatabase: {
// There is no specific namespace to save for a drop database operation.
break;
}
case OplogEntry::CommandType::kDbCheck:
case OplogEntry::CommandType::kEmptyCapped: {
// These commands do not need to be supported by rollback. 'convertToCapped' should
// always be converted to lower level DDL operations, and 'emptycapped' is a
// testing-only command.
std::string message = str::stream()
<< "Encountered unsupported command type '" << firstElem.fieldName()
<< "' during rollback.";
return Status(ErrorCodes::UnrecoverableRollbackError, message);
}
case OplogEntry::CommandType::kCreate:
case OplogEntry::CommandType::kDrop:
case OplogEntry::CommandType::kImportCollection:
case OplogEntry::CommandType::kCreateIndexes:
case OplogEntry::CommandType::kDropIndexes:
case OplogEntry::CommandType::kStartIndexBuild:
case OplogEntry::CommandType::kAbortIndexBuild:
case OplogEntry::CommandType::kCommitIndexBuild:
case OplogEntry::CommandType::kCollMod: {
// For all other command types, we should be able to parse the collection name from
// the first command argument.
try {
auto cmdNss = CommandHelpers::parseNsCollectionRequired(opNss.db(), obj);
namespaces.insert(cmdNss);
} catch (const DBException& ex) {
return ex.toStatus();
}
break;
}
case OplogEntry::CommandType::kCommitTransaction:
case OplogEntry::CommandType::kAbortTransaction: {
break;
}
case OplogEntry::CommandType::kApplyOps:
default:
// Every possible command type should be handled above.
MONGO_UNREACHABLE
}
}
return namespaces;
}
void RollbackImpl::_restoreTxnsTableEntryFromRetryableWrites(OperationContext* opCtx,
Timestamp stableTimestamp) {
auto client = std::make_unique<DBDirectClient>(opCtx);
// Query for retryable writes oplog entries with a non-null 'prevWriteOpTime' value
// less than or equal to the 'stableTimestamp'. This query intends to include no-op
// retryable writes oplog entries that have been applied through a migration process.
const auto filter = BSON("op" << BSON("$in" << BSON_ARRAY("i"
<< "u"
<< "d")));
// We use the 'fromMigrate' field to differentiate migrated retryable writes entries from
// transactions entries.
const auto filterFromMigration = BSON("op"
<< "n"
<< "fromMigrate" << true);
auto cursor = client->query(
NamespaceString::kRsOplogNamespace,
QUERY("ts" << BSON("$gt" << stableTimestamp) << "txnNumber" << BSON("$exists" << true)
<< "stmtId" << BSON("$exists" << true) << "prevOpTime.ts"
<< BSON("$gte" << Timestamp(1, 0) << "$lte" << stableTimestamp) << "$or"
<< BSON_ARRAY(filter << filterFromMigration)));
while (cursor->more()) {
auto doc = cursor->next();
auto swEntry = OplogEntry::parse(doc);
fassert(5530502, swEntry.isOK());
auto entry = swEntry.getValue();
auto prevWriteOpTime = *entry.getPrevWriteOpTimeInTransaction();
OperationSessionInfo opSessionInfo = entry.getOperationSessionInfo();
const auto sessionId = *opSessionInfo.getSessionId();
const auto txnNumber = *opSessionInfo.getTxnNumber();
const auto wallClockTime = entry.getWallClockTime();
invariant(!prevWriteOpTime.isNull() && prevWriteOpTime.getTimestamp() <= stableTimestamp);
// This is a retryable writes oplog entry with a non-null 'prevWriteOpTime' value that
// is less than or equal to the 'stableTimestamp'.
LOGV2(5530501,
"Restoring sessions entry to be consistent with 'stableTimestamp'",
"stableTimestamp"_attr = stableTimestamp,
"sessionId"_attr = sessionId,
"txnNumber"_attr = txnNumber,
"lastWriteOpTime"_attr = prevWriteOpTime);
SessionTxnRecord sessionTxnRecord;
sessionTxnRecord.setSessionId(sessionId);
sessionTxnRecord.setTxnNum(txnNumber);
try {
TransactionHistoryIterator iter(prevWriteOpTime);
auto nextOplogEntry = iter.next(opCtx);
sessionTxnRecord.setLastWriteOpTime(nextOplogEntry.getOpTime());
sessionTxnRecord.setLastWriteDate(nextOplogEntry.getWallClockTime());
} catch (ExceptionFor<ErrorCodes::IncompleteTransactionHistory>&) {
// It's possible that the next entry in the oplog chain has been truncated due to
// oplog cap maintenance.
sessionTxnRecord.setLastWriteOpTime(prevWriteOpTime);
sessionTxnRecord.setLastWriteDate(wallClockTime);
}
const auto nss = NamespaceString::kSessionTransactionsTableNamespace;
writeConflictRetry(opCtx, "updateSessionTransactionsTableInRollback", nss.ns(), [&] {
AutoGetCollection collection(opCtx, nss, MODE_IX);
auto filter = BSON(SessionTxnRecord::kSessionIdFieldName << sessionId.toBSON());
UnreplicatedWritesBlock uwb(opCtx);
// Perform an untimestamped write so that it will not be rolled back on recovering
// to the 'stableTimestamp' if we were to crash. This is safe because this update is
// meant to be consistent with the 'stableTimestamp' and not the common point.
Helpers::upsert(
opCtx, nss.ns(), filter, sessionTxnRecord.toBSON(), /*fromMigrate=*/false);
});
}
// Take a stable checkpoint so that writes to the 'config.transactions' table are
// persisted to disk before truncating the oplog. If we were to take an unstable checkpoint, we
// would have to update replication metadata like 'minValid.appliedThrough' to be consistent
// with the oplog.
opCtx->recoveryUnit()->waitUntilUnjournaledWritesDurable(opCtx,
/*stableCheckpoint=*/true);
}
void RollbackImpl::_runPhaseFromAbortToReconstructPreparedTxns(
OperationContext* opCtx, RollBackLocalOperations::RollbackCommonPoint commonPoint) noexcept {
// Stop and wait for all background index builds to complete before starting the rollback
// process.
_stopAndWaitForIndexBuilds(opCtx);
_listener->onBgIndexesComplete();
// Before computing record store counts, abort all active transactions. This ensures that
// the count adjustments are based on correct values where no prepared transactions are
// active and all in-memory counts have been rolled-back.
// Before calling recoverToStableTimestamp, we must abort the storage transaction of any
// prepared transaction. This will require us to scan all sessions and call
// abortPreparedTransactionForRollback() on any txnParticipant with a prepared transaction.
killSessionsAbortAllPreparedTransactions(opCtx);
// Ask the record store for the pre-rollback counts of any collections whose counts will
// change and create a map with the adjusted counts for post-rollback. While finding the
// common point, we keep track of how much each collection's count will change during the
// rollback. Note: these numbers are relative to the common point, not the stable timestamp,
// and thus must be set after recovering from the oplog.
auto status = _findRecordStoreCounts(opCtx);
fassert(31227, status);
if (shouldCreateDataFiles()) {
// Write a rollback file for each namespace that has documents that would be deleted by
// rollback. We need to do this after aborting prepared transactions. Otherwise, we risk
// unecessary prepare conflicts when trying to read documents that were modified by
// those prepared transactions, which we know we will abort anyway.
status = _writeRollbackFiles(opCtx);
fassert(31228, status);
} else {
LOGV2(21598, "Not writing rollback files. 'createRollbackDataFiles' set to false");
}
// If there were rolled back operations on any session, invalidate all sessions.
// We invalidate sessions before we recover so that we avoid invalidating sessions that had
// just recovered prepared transactions.
if (!_observerInfo.rollbackSessionIds.empty()) {
MongoDSessionCatalog::invalidateAllSessions(opCtx);
}
// Recover to the stable timestamp.
auto stableTimestamp = _recoverToStableTimestamp(opCtx);
_rollbackStats.stableTimestamp = stableTimestamp;
_listener->onRecoverToStableTimestamp(stableTimestamp);
// Log the total number of insert and update operations that have been rolled back as a
// result of recovering to the stable timestamp.
auto getCommandCount = [&](StringData key) {
const auto& m = _observerInfo.rollbackCommandCounts;
auto it = m.find(key);
return (it == m.end()) ? 0 : it->second;
};
LOGV2(21599,
"Rollback reverted {insert} insert operations, {update} update operations and {delete} "
"delete operations.",
"Rollback reverted command counts",
"insert"_attr = getCommandCount(kInsertCmdName),
"update"_attr = getCommandCount(kUpdateCmdName),
"delete"_attr = getCommandCount(kDeleteCmdName));
// Retryable writes create derived updates to the transactions table which can be coalesced into
// one operation, so certain session operations history may be lost after restoring to the
// 'stableTimestamp'. We must scan the oplog and restore the transactions table entries to
// detail the last executed writes.
_restoreTxnsTableEntryFromRetryableWrites(opCtx, stableTimestamp);
// During replication recovery, we truncate all oplog entries with timestamps greater than the
// oplog truncate after point. If we entered rollback, we are guaranteed to have at least one
// oplog entry after the common point.
LOGV2(21600,
"Marking to truncate all oplog entries with timestamps greater than "
"{commonPoint}",
"Marking to truncate all oplog entries with timestamps greater than common point",
"commonPoint"_attr = commonPoint.getOpTime().getTimestamp());
Timestamp truncatePoint = commonPoint.getOpTime().getTimestamp();
// Persist the truncate point to the 'oplogTruncateAfterPoint' document. We save this value so
// that the replication recovery logic knows where to truncate the oplog. We save this value
// durably to match the behavior during startup recovery. This must occur after we successfully
// recover to a stable timestamp. If recovering to a stable timestamp fails and we still
// truncate the oplog then the oplog will not match the data files. If we crash at any earlier
// point, we will recover, find a new sync source, and restart roll back (if necessary on the
// new sync source). This is safe because a crash before this point would recover to a stable
// checkpoint anyways at or earlier than the stable timestamp.
//
// Note that storage engine timestamp recovery only restores the database *data* to a stable
// timestamp, but does not revert the oplog, which must be done as part of the rollback process.
_replicationProcess->getConsistencyMarkers()->setOplogTruncateAfterPoint(opCtx, truncatePoint);
_rollbackStats.truncateTimestamp = truncatePoint;
_listener->onSetOplogTruncateAfterPoint(truncatePoint);
// Align the drop pending reaper state with what's on disk. Oplog recovery depends on those
// being consistent.
_resetDropPendingState(opCtx);
// Run the recovery process.
_replicationProcess->getReplicationRecovery()->recoverFromOplog(opCtx, stableTimestamp);
_listener->onRecoverFromOplog();
// Sets the correct post-rollback counts on any collections whose counts changed during the
// rollback.
_correctRecordStoreCounts(opCtx);
tenant_migration_access_blocker::recoverTenantMigrationAccessBlockers(opCtx);
// Reconstruct prepared transactions after counts have been adjusted. Since prepared
// transactions were aborted (i.e. the in-memory counts were rolled-back) before computing
// collection counts, reconstruct the prepared transactions now, adding on any additional counts
// to the now corrected record store.
reconstructPreparedTransactions(opCtx, OplogApplication::Mode::kRecovering);
}
void RollbackImpl::_correctRecordStoreCounts(OperationContext* opCtx) {
// This function explicitly does not check for shutdown since a clean shutdown post oplog
// truncation is not allowed to occur until the record store counts are corrected.
auto catalog = CollectionCatalog::get(opCtx);
for (const auto& uiCount : _newCounts) {
const auto uuid = uiCount.first;
const auto coll = catalog->lookupCollectionByUUID(opCtx, uuid);
invariant(coll,
str::stream() << "The collection with UUID " << uuid
<< " is unexpectedly missing in the CollectionCatalog");
const auto nss = coll->ns();
invariant(!nss.isEmpty(),
str::stream() << "The collection with UUID " << uuid << " has no namespace.");
const auto ident = coll->getRecordStore()->getIdent();
invariant(!ident.empty(),
str::stream() << "The collection with UUID " << uuid << " has no ident.");
auto newCount = uiCount.second;
// If the collection is marked for size adjustment, then we made sure the collection size
// was accurate at the stable timestamp and we can trust replication recovery to keep it
// correct. This is necessary for capped collections whose deletions will be untracked
// if we just set the collection count here.
if (sizeRecoveryState(opCtx->getServiceContext())
.collectionAlwaysNeedsSizeAdjustment(ident)) {
LOGV2_DEBUG(
21601,
2,
"Not setting collection count to {newCount} for {namespace} ({uuid}) "
"[{ident}] because it is marked for size adjustment.",
"Not setting collection count because namespace is marked for size adjustment",
"newCount"_attr = newCount,
"namespace"_attr = nss.ns(),
"uuid"_attr = uuid.toString(),
"ident"_attr = ident);
continue;
}
// If _findRecordStoreCounts() is unable to determine the correct count from the oplog
// (most likely due to a 4.0 drop oplog entry without the count information), we will
// determine the correct count here post-recovery using a collection scan.
if (kCollectionScanRequired == newCount) {
LOGV2(21602,
"Scanning collection {namespace} ({uuid}) to fix collection count.",
"Scanning collection to fix collection count",
"namespace"_attr = nss.ns(),
"uuid"_attr = uuid.toString());
AutoGetCollectionForRead collToScan(opCtx, nss);
invariant(coll == collToScan.getCollection(),
str::stream() << "Catalog returned invalid collection: " << nss.ns() << " ("
<< uuid.toString() << ")");
auto exec = collToScan->makePlanExecutor(opCtx,
collToScan.getCollection(),
PlanYieldPolicy::YieldPolicy::INTERRUPT_ONLY,
Collection::ScanDirection::kForward);
long long countFromScan = 0;
PlanExecutor::ExecState state;
while (PlanExecutor::ADVANCED ==
(state = exec->getNext(static_cast<BSONObj*>(nullptr), nullptr))) {
++countFromScan;
}
if (PlanExecutor::IS_EOF != state) {
// We ignore errors here because crashing or leaving rollback would only leave
// collection counts more inaccurate.
LOGV2_WARNING(21637,
"Failed to set count of {namespace} ({uuid}) [{ident}] due to failed "
"collection scan: {error}",
"Failed to set count of namespace due to failed collection scan",
"namespace"_attr = nss.ns(),
"uuid"_attr = uuid.toString(),
"ident"_attr = ident,
"error"_attr = exec->stateToStr(state));
continue;
}
newCount = countFromScan;
}
auto status =
_storageInterface->setCollectionCount(opCtx, {nss.db().toString(), uuid}, newCount);
if (!status.isOK()) {
// We ignore errors here because crashing or leaving rollback would only leave
// collection counts more inaccurate.
LOGV2_WARNING(21638,
"Failed to set count of {namespace} ({uuid}) [{ident}] to {newCount}. "
"Received: {error}",
"Failed to set count of namespace",
"namespace"_attr = nss.ns(),
"uuid"_attr = uuid.toString(),
"ident"_attr = ident,
"newCount"_attr = newCount,
"error"_attr = status);
} else {
LOGV2_DEBUG(21603,
2,
"Set collection count of {namespace} ({uuid}) [{ident}] to {newCount}.",
"Set collection count of namespace",
"namespace"_attr = nss.ns(),
"uuid"_attr = uuid.toString(),
"ident"_attr = ident,
"newCount"_attr = newCount);
}
}
}
Status RollbackImpl::_findRecordStoreCounts(OperationContext* opCtx) {
auto catalog = CollectionCatalog::get(opCtx);
auto storageEngine = opCtx->getServiceContext()->getStorageEngine();
LOGV2(21604, "Finding record store counts");
for (const auto& uiCount : _countDiffs) {
auto uuid = uiCount.first;
auto countDiff = uiCount.second;
if (countDiff == 0) {
continue;
}
auto nss = catalog->lookupNSSByUUID(opCtx, uuid);
StorageInterface::CollectionCount oldCount = 0;
// Drop-pending collections are not visible to rollback via the catalog when they are
// managed by the storage engine. See StorageEngine::supportsPendingDrops().
if (!nss) {
invariant(storageEngine->supportsPendingDrops(),
str::stream() << "The collection with UUID " << uuid
<< " is unexpectedly missing in the CollectionCatalog");
auto it = _pendingDrops.find(uuid);
if (it == _pendingDrops.end()) {
_newCounts[uuid] = kCollectionScanRequired;
continue;
}
const auto& dropPendingInfo = it->second;
nss = dropPendingInfo.nss;
invariant(dropPendingInfo.count >= 0,
str::stream() << "The collection with UUID " << uuid
<< " was dropped with a negative collection count of "
<< dropPendingInfo.count
<< " in the drop or rename oplog entry. Unable to reset "
"collection count during rollback.");
oldCount = static_cast<StorageInterface::CollectionCount>(dropPendingInfo.count);
} else {
auto countSW = _storageInterface->getCollectionCount(opCtx, *nss);
if (!countSW.isOK()) {
return countSW.getStatus();
}
oldCount = countSW.getValue();
}
if (oldCount > static_cast<uint64_t>(std::numeric_limits<long long>::max())) {
LOGV2_WARNING(21639,
"Count for {namespace} ({uuid}) was {oldCount} which is larger than the "
"maximum int64_t value. Not attempting to fix "
"count during rollback.",
"Count for namespace was larger than the maximum int64_t value. Not "
"attempting to fix count during rollback",
"namespace"_attr = nss->ns(),
"uuid"_attr = uuid.toString(),
"oldCount"_attr = oldCount);
continue;
}
long long oldCountSigned = static_cast<long long>(oldCount);
auto newCount = oldCountSigned + countDiff;
if (newCount < 0) {
LOGV2_WARNING(
21640,
"Attempted to set count for {namespace} ({uuid}) to {newCount} but set it to 0 "
"instead. This is likely due to the count previously "
"becoming inconsistent from an unclean shutdown or a rollback that could "
"not fix the count correctly. Old count: {oldCount}. Count change: {countDiff}",
"Attempted to set count for namespace but set it to 0 instead. This is likely due "
"to the count previously becoming inconsistent from an unclean shutdown or a "
"rollback that could not fix the count correctly",
"namespace"_attr = nss->ns(),
"uuid"_attr = uuid.toString(),
"newCount"_attr = newCount,
"oldCount"_attr = oldCount,
"countDiff"_attr = countDiff);
newCount = 0;
}
LOGV2_DEBUG(
21605,
2,
"Record count of {namespace} ({uuid}) before rollback is {oldCount}. Setting it "
"to {newCount}, due to change of {countDiff}",
"Setting record count for namespace after rollback",
"namespace"_attr = nss->ns(),
"uuid"_attr = uuid.toString(),
"oldCount"_attr = oldCount,
"newCount"_attr = newCount,
"countDiff"_attr = countDiff);
_newCounts[uuid] = newCount;
}
return Status::OK();
}
/**
* Process a single oplog entry that is getting rolled back and update the necessary rollback info
* structures.
*/
Status RollbackImpl::_processRollbackOp(OperationContext* opCtx, const OplogEntry& oplogEntry) {
++_observerInfo.numberOfEntriesObserved;
NamespaceString opNss = oplogEntry.getNss();
OpTypeEnum opType = oplogEntry.getOpType();
// For applyOps entries, we process each sub-operation individually.
if (oplogEntry.getCommandType() == OplogEntry::CommandType::kApplyOps) {
if (oplogEntry.shouldPrepare()) {
// Uncommitted prepared transactions are always aborted before rollback begins, which
// rolls back collection counts. Processing the operation here would result in
// double-counting the sub-operations when correcting collection counts later.
// Additionally, this logic makes an assumption that transactions are only ever
// committed when the prepare operation is majority committed. This implies that when a
// prepare oplog entry is rolled-back, it is guaranteed that it has never committed.
return Status::OK();
}
if (oplogEntry.isPartialTransaction()) {
// This oplog entry will be processed when we rollback the implicit commit for the
// unprepared transaction (applyOps without partialTxn field).
return Status::OK();
}
// Follow chain on applyOps oplog entries to process entire unprepared transaction.
// The beginning of the applyOps chain may precede the common point.
auto status = _processRollbackOpForApplyOps(opCtx, oplogEntry);
if (const auto prevOpTime = oplogEntry.getPrevWriteOpTimeInTransaction()) {
for (TransactionHistoryIterator iter(*prevOpTime); status.isOK() && iter.hasNext();) {
status = _processRollbackOpForApplyOps(opCtx, iter.next(opCtx));
}
}
return status;
}
// No information to record for a no-op.
if (opType == OpTypeEnum::kNoop) {
return Status::OK();
}
// Extract the appropriate namespaces from the oplog operation.
auto namespacesSW = _namespacesForOp(oplogEntry);
if (!namespacesSW.isOK()) {
return namespacesSW.getStatus();
} else {
_observerInfo.rollbackNamespaces.insert(namespacesSW.getValue().begin(),
namespacesSW.getValue().end());
}
// If the operation being rolled back has a session id, then we add it to the set of
// sessions that had operations rolled back.
OperationSessionInfo opSessionInfo = oplogEntry.getOperationSessionInfo();
auto sessionId = opSessionInfo.getSessionId();
if (sessionId) {
_observerInfo.rollbackSessionIds.insert(sessionId->getId());
}
// Keep track of the _ids of inserted and updated documents, as we may need to write them out to
// a rollback file.
if (opType == OpTypeEnum::kInsert || opType == OpTypeEnum::kUpdate) {
const auto uuid = oplogEntry.getUuid();
invariant(uuid,
str::stream() << "Oplog entry to roll back is unexpectedly missing a UUID: "
<< redact(oplogEntry.toBSONForLogging()));
const auto idElem = oplogEntry.getIdElement();
if (!idElem.eoo()) {
// We call BSONElement::wrap() on each _id element to create a new BSONObj with an owned
// buffer, as the underlying storage may be gone when we access this map to write
// rollback files.
_observerInfo.rollbackDeletedIdsMap[uuid.get()].insert(idElem.wrap());
const auto cmdName = opType == OpTypeEnum::kInsert ? kInsertCmdName : kUpdateCmdName;
++_observerInfo.rollbackCommandCounts[cmdName];
}
}
if (opType == OpTypeEnum::kInsert) {
auto idVal = oplogEntry.getObject().getStringField("_id");
if (serverGlobalParams.clusterRole == ClusterRole::ShardServer &&
opNss == NamespaceString::kServerConfigurationNamespace &&
idVal == ShardIdentityType::IdName) {
// Check if the creation of the shard identity document is being rolled back.
_observerInfo.shardIdentityRolledBack = true;
LOGV2_WARNING(21641,
"Shard identity document rollback detected. oplog op: {oplogEntry}",
"Shard identity document rollback detected",
"oplogEntry"_attr = redact(oplogEntry.toBSONForLogging()));
} else if (serverGlobalParams.clusterRole == ClusterRole::ConfigServer &&
opNss == VersionType::ConfigNS) {
// Check if the creation of the config server config version document is being rolled
// back.
_observerInfo.configServerConfigVersionRolledBack = true;
LOGV2_WARNING(21642,
"Config version document rollback detected. oplog op: {oplogEntry}",
"Config version document rollback detected",
"oplogEntry"_attr = redact(oplogEntry.toBSONForLogging()));
}
// Rolling back an insert must decrement the count by 1.
_countDiffs[oplogEntry.getUuid().get()] -= 1;
} else if (opType == OpTypeEnum::kDelete) {
// Rolling back a delete must increment the count by 1.
_countDiffs[oplogEntry.getUuid().get()] += 1;
} else if (opType == OpTypeEnum::kCommand) {
if (oplogEntry.getCommandType() == OplogEntry::CommandType::kCreate) {
// If we roll back a create, then we do not need to change the size of that uuid.
_countDiffs.erase(oplogEntry.getUuid().get());
_pendingDrops.erase(oplogEntry.getUuid().get());
_newCounts.erase(oplogEntry.getUuid().get());
} else if (oplogEntry.getCommandType() == OplogEntry::CommandType::kImportCollection) {
auto importEntry = mongo::ImportCollectionOplogEntry::parse(
IDLParserErrorContext("importCollectionOplogEntry"), oplogEntry.getObject());
// Nothing to roll back if this is a dryRun.
if (!importEntry.getDryRun()) {
const auto& catalogEntry = importEntry.getCatalogEntry();
auto importTargetUUID =
invariant(UUID::parse(catalogEntry["md"]["options"]["uuid"]),
str::stream() << "Oplog entry to roll back is unexpectedly missing "
"import collection UUID: "
<< redact(oplogEntry.toBSONForLogging()));
// If we roll back an import, then we do not need to change the size of that uuid.
_countDiffs.erase(importTargetUUID);
_pendingDrops.erase(importTargetUUID);
_newCounts.erase(importTargetUUID);
}
} else if (oplogEntry.getCommandType() == OplogEntry::CommandType::kDrop) {
// If we roll back a collection drop, parse the o2 field for the collection count for
// use later by _findRecordStoreCounts().
// This will be used to reconcile collection counts in the case where the drop-pending
// collection is managed by the storage engine and is not accessible through the UUID
// catalog.
// Adding a _newCounts entry ensures that the count will be set after the rollback.
const auto uuid = oplogEntry.getUuid().get();
invariant(_countDiffs.find(uuid) == _countDiffs.end(),
str::stream() << "Unexpected existing count diff for " << uuid.toString()
<< " op: " << redact(oplogEntry.toBSONForLogging()));
if (auto countResult = _parseDroppedCollectionCount(oplogEntry)) {
PendingDropInfo info;
info.count = *countResult;
const auto& opNss = oplogEntry.getNss();
info.nss =
CommandHelpers::parseNsCollectionRequired(opNss.db(), oplogEntry.getObject());
_pendingDrops[uuid] = info;
_newCounts[uuid] = info.count;
} else {
_newCounts[uuid] = kCollectionScanRequired;
}
} else if (oplogEntry.getCommandType() == OplogEntry::CommandType::kRenameCollection &&
oplogEntry.getObject()[kDropTargetFieldName].trueValue()) {
// If we roll back a rename with a dropped target collection, parse the o2 field for the
// target collection count for use later by _findRecordStoreCounts().
// This will be used to reconcile collection counts in the case where the drop-pending
// collection is managed by the storage engine and is not accessible through the UUID