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
|
/**
* Copyright (C) 2015 MongoDB Inc.
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU Affero General Public License, version 3,
* as published by the Free Software Foundation.
*
* 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
* GNU Affero General Public License for more details.
*
* You should have received a copy of the GNU Affero General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*
* 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 GNU Affero General 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_LOG_DEFAULT_COMPONENT ::mongo::logger::LogComponent::kReplication
#include "mongo/platform/basic.h"
#include "mongo/db/repl/database_cloner.h"
#include <algorithm>
#include <iterator>
#include <set>
#include "mongo/client/remote_command_retry_scheduler.h"
#include "mongo/db/catalog/collection_options.h"
#include "mongo/db/commands/list_collections_filter.h"
#include "mongo/db/repl/storage_interface.h"
#include "mongo/db/server_parameters.h"
#include "mongo/stdx/functional.h"
#include "mongo/util/assert_util.h"
#include "mongo/util/destructor_guard.h"
#include "mongo/util/fail_point_service.h"
#include "mongo/util/log.h"
#include "mongo/util/mongoutils/str.h"
namespace mongo {
namespace repl {
// Failpoint which causes the initial sync function to hang before running listCollections.
MONGO_FAIL_POINT_DEFINE(initialSyncHangBeforeListCollections);
namespace {
using LockGuard = stdx::lock_guard<stdx::mutex>;
using UniqueLock = stdx::unique_lock<stdx::mutex>;
using executor::RemoteCommandRequest;
const char* kNameFieldName = "name";
const char* kOptionsFieldName = "options";
const char* kInfoFieldName = "info";
const char* kUUIDFieldName = "uuid";
// The batchSize to use for the find/getMore queries called by the CollectionCloner
constexpr int kUseARMDefaultBatchSize = -1;
MONGO_EXPORT_STARTUP_SERVER_PARAMETER(collectionClonerBatchSize, int, kUseARMDefaultBatchSize);
// The number of attempts for the listCollections commands.
MONGO_EXPORT_SERVER_PARAMETER(numInitialSyncListCollectionsAttempts, int, 3);
// The number of cursors to use in the collection cloning process.
MONGO_EXPORT_SERVER_PARAMETER(maxNumInitialSyncCollectionClonerCursors, int, 1);
// Failpoint which causes initial sync to hang right after listCollections, but before cloning
// any colelctions in the 'database' database.
MONGO_FAIL_POINT_DEFINE(initialSyncHangAfterListCollections);
/**
* Default listCollections predicate.
*/
bool acceptAllPred(const BSONObj&) {
return true;
}
/**
* Creates a listCollections command obj with an optional filter.
*/
BSONObj createListCollectionsCommandObject(const BSONObj& filter) {
BSONObjBuilder output;
output.append("listCollections", 1);
if (!filter.isEmpty()) {
output.append("filter", filter);
}
return output.obj();
}
} // namespace
DatabaseCloner::DatabaseCloner(executor::TaskExecutor* executor,
ThreadPool* dbWorkThreadPool,
const HostAndPort& source,
const std::string& dbname,
const BSONObj& listCollectionsFilter,
const ListCollectionsPredicateFn& listCollectionsPred,
StorageInterface* si,
const CollectionCallbackFn& collWork,
const CallbackFn& onCompletion)
: _executor(executor),
_dbWorkThreadPool(dbWorkThreadPool),
_source(source),
_dbname(dbname),
_listCollectionsFilter(
listCollectionsFilter.isEmpty()
? ListCollectionsFilter::makeTypeCollectionFilter()
: ListCollectionsFilter::addTypeCollectionFilter(listCollectionsFilter)),
_listCollectionsPredicate(listCollectionsPred ? listCollectionsPred : acceptAllPred),
_storageInterface(si),
_collectionWork(collWork),
_onCompletion(onCompletion),
_listCollectionsFetcher(_executor,
_source,
_dbname,
createListCollectionsCommandObject(_listCollectionsFilter),
[=](const StatusWith<Fetcher::QueryResponse>& result,
Fetcher::NextAction * nextAction,
BSONObjBuilder * getMoreBob) {
_listCollectionsCallback(result, nextAction, getMoreBob);
},
ReadPreferenceSetting::secondaryPreferredMetadata(),
RemoteCommandRequest::kNoTimeout /* find network timeout */,
RemoteCommandRequest::kNoTimeout /* getMore network timeout */,
RemoteCommandRetryScheduler::makeRetryPolicy(
numInitialSyncListCollectionsAttempts.load(),
executor::RemoteCommandRequest::kNoTimeout,
RemoteCommandRetryScheduler::kAllRetriableErrors)),
_startCollectionCloner([](CollectionCloner& cloner) { return cloner.startup(); }) {
// Fetcher throws an exception on null executor.
invariant(executor);
uassert(ErrorCodes::BadValue, "db worker thread pool cannot be null", dbWorkThreadPool);
uassert(ErrorCodes::BadValue, "empty database name", !dbname.empty());
uassert(ErrorCodes::BadValue, "storage interface cannot be null", si);
uassert(ErrorCodes::BadValue, "collection callback function cannot be null", collWork);
uassert(ErrorCodes::BadValue, "callback function cannot be null", onCompletion);
_stats.dbname = _dbname;
}
DatabaseCloner::~DatabaseCloner() {
DESTRUCTOR_GUARD(shutdown(); join(););
}
const std::vector<BSONObj>& DatabaseCloner::getCollectionInfos_forTest() const {
LockGuard lk(_mutex);
return _collectionInfos;
}
bool DatabaseCloner::isActive() const {
LockGuard lk(_mutex);
return _isActive_inlock();
}
bool DatabaseCloner::_isActive_inlock() const {
return State::kRunning == _state || State::kShuttingDown == _state;
}
bool DatabaseCloner::_isShuttingDown() const {
LockGuard lk(_mutex);
return State::kShuttingDown == _state;
}
Status DatabaseCloner::startup() noexcept {
UniqueLock lk(_mutex);
switch (_state) {
case State::kPreStart:
_state = State::kRunning;
break;
case State::kRunning:
return Status(ErrorCodes::InternalError, "database cloner already started");
case State::kShuttingDown:
return Status(ErrorCodes::ShutdownInProgress, "database cloner shutting down");
case State::kComplete:
return Status(ErrorCodes::ShutdownInProgress, "database cloner completed");
}
MONGO_FAIL_POINT_BLOCK(initialSyncHangBeforeListCollections, customArgs) {
const auto& data = customArgs.getData();
const auto databaseElem = data["database"];
if (!databaseElem || databaseElem.checkAndGetStringData() == _dbname) {
lk.unlock();
log() << "initial sync - initialSyncHangBeforeListCollections fail point "
"enabled. Blocking until fail point is disabled.";
while (MONGO_FAIL_POINT(initialSyncHangBeforeListCollections) && !_isShuttingDown()) {
mongo::sleepsecs(1);
}
lk.lock();
}
}
_stats.start = _executor->now();
LOG(1) << "Scheduling listCollections call for database: " << _dbname;
Status scheduleResult = _listCollectionsFetcher.schedule();
if (!scheduleResult.isOK()) {
error() << "Error scheduling listCollections for database: " << _dbname
<< ", error:" << scheduleResult;
_state = State::kComplete;
return scheduleResult;
}
return Status::OK();
}
void DatabaseCloner::shutdown() {
stdx::lock_guard<stdx::mutex> lock(_mutex);
switch (_state) {
case State::kPreStart:
// Transition directly from PreStart to Complete if not started yet.
_state = State::kComplete;
return;
case State::kRunning:
_state = State::kShuttingDown;
break;
case State::kShuttingDown:
case State::kComplete:
// Nothing to do if we are already in ShuttingDown or Complete state.
return;
}
for (auto&& collectionCloner : _collectionCloners) {
collectionCloner.shutdown();
}
_listCollectionsFetcher.shutdown();
}
DatabaseCloner::Stats DatabaseCloner::getStats() const {
LockGuard lk(_mutex);
DatabaseCloner::Stats stats = _stats;
for (auto&& collectionCloner : _collectionCloners) {
stats.collectionStats.emplace_back(collectionCloner.getStats());
}
return stats;
}
void DatabaseCloner::join() {
UniqueLock lk(_mutex);
_condition.wait(lk, [this]() { return !_isActive_inlock(); });
}
void DatabaseCloner::setScheduleDbWorkFn_forTest(const CollectionCloner::ScheduleDbWorkFn& work) {
LockGuard lk(_mutex);
_scheduleDbWorkFn = work;
}
void DatabaseCloner::setStartCollectionClonerFn(
const StartCollectionClonerFn& startCollectionCloner) {
_startCollectionCloner = startCollectionCloner;
}
DatabaseCloner::State DatabaseCloner::getState_forTest() const {
stdx::lock_guard<stdx::mutex> lk(_mutex);
return _state;
}
void DatabaseCloner::_listCollectionsCallback(const StatusWith<Fetcher::QueryResponse>& result,
Fetcher::NextAction* nextAction,
BSONObjBuilder* getMoreBob) {
if (!result.isOK()) {
_finishCallback(result.getStatus().withContext(
str::stream() << "Error issuing listCollections on db '" << _dbname << "' (host:"
<< _source.toString()
<< ")"));
return;
}
auto batchData(result.getValue());
auto&& documents = batchData.documents;
UniqueLock lk(_mutex);
// We may be called with multiple batches leading to a need to grow _collectionInfos.
_collectionInfos.reserve(_collectionInfos.size() + documents.size());
std::copy_if(documents.begin(),
documents.end(),
std::back_inserter(_collectionInfos),
_listCollectionsPredicate);
_stats.collections += _collectionInfos.size();
// The fetcher will continue to call with kGetMore until an error or the last batch.
if (*nextAction == Fetcher::NextAction::kGetMore) {
invariant(getMoreBob);
getMoreBob->append("getMore", batchData.cursorId);
getMoreBob->append("collection", batchData.nss.coll());
return;
}
// Nothing to do for an empty database.
if (_collectionInfos.empty()) {
_finishCallback_inlock(lk, Status::OK());
return;
}
MONGO_FAIL_POINT_BLOCK(initialSyncHangAfterListCollections, options) {
const BSONObj& data = options.getData();
if (data["database"].String() == _dbname) {
log() << "initial sync - initialSyncHangAfterListCollections fail point "
"enabled. Blocking until fail point is disabled.";
while (MONGO_FAIL_POINT(initialSyncHangAfterListCollections)) {
mongo::sleepsecs(1);
}
}
}
_collectionNamespaces.reserve(_collectionInfos.size());
std::set<std::string> seen;
for (auto&& info : _collectionInfos) {
BSONElement nameElement = info.getField(kNameFieldName);
if (nameElement.eoo()) {
_finishCallback_inlock(
lk,
{ErrorCodes::FailedToParse,
str::stream() << "collection info must contain '" << kNameFieldName << "' "
<< "field : "
<< info});
return;
}
if (nameElement.type() != mongo::String) {
_finishCallback_inlock(
lk,
{ErrorCodes::TypeMismatch,
str::stream() << "'" << kNameFieldName << "' field must be a string: " << info});
return;
}
const std::string collectionName = nameElement.String();
if (seen.find(collectionName) != seen.end()) {
_finishCallback_inlock(lk,
{ErrorCodes::DuplicateKey,
str::stream()
<< "collection info contains duplicate collection name "
<< "'"
<< collectionName
<< "': "
<< info});
return;
}
BSONElement optionsElement = info.getField(kOptionsFieldName);
if (optionsElement.eoo()) {
_finishCallback_inlock(
lk,
{ErrorCodes::FailedToParse,
str::stream() << "collection info must contain '" << kOptionsFieldName << "' "
<< "field : "
<< info});
return;
}
if (!optionsElement.isABSONObj()) {
_finishCallback_inlock(lk,
Status(ErrorCodes::TypeMismatch,
str::stream() << "'" << kOptionsFieldName
<< "' field must be an object: "
<< info));
return;
}
const BSONObj optionsObj = optionsElement.Obj();
CollectionOptions options;
auto parseStatus = options.parse(optionsObj, CollectionOptions::parseForStorage);
if (!parseStatus.isOK()) {
_finishCallback_inlock(lk, parseStatus);
return;
}
BSONElement infoElement = info.getField(kInfoFieldName);
if (infoElement.isABSONObj()) {
BSONElement uuidElement = infoElement[kUUIDFieldName];
if (!uuidElement.eoo()) {
auto res = CollectionUUID::parse(uuidElement);
if (!res.isOK()) {
_finishCallback_inlock(lk, res.getStatus());
return;
}
options.uuid = res.getValue();
}
}
// TODO(SERVER-27994): Ensure UUID present when FCV >= "3.6".
seen.insert(collectionName);
_collectionNamespaces.emplace_back(_dbname, collectionName);
auto&& nss = *_collectionNamespaces.crbegin();
try {
_collectionCloners.emplace_back(
_executor,
_dbWorkThreadPool,
_source,
nss,
options,
[=](const Status& status) { return _collectionClonerCallback(status, nss); },
_storageInterface,
collectionClonerBatchSize,
maxNumInitialSyncCollectionClonerCursors.load());
} catch (const AssertionException& ex) {
_finishCallback_inlock(lk, ex.toStatus());
return;
}
}
if (_scheduleDbWorkFn) {
for (auto&& collectionCloner : _collectionCloners) {
collectionCloner.setScheduleDbWorkFn_forTest(_scheduleDbWorkFn);
}
}
// Start first collection cloner.
_currentCollectionClonerIter = _collectionCloners.begin();
LOG(1) << " cloning collection " << _currentCollectionClonerIter->getSourceNamespace();
Status startStatus = _startCollectionCloner(*_currentCollectionClonerIter);
if (!startStatus.isOK()) {
LOG(1) << " failed to start collection cloning on "
<< _currentCollectionClonerIter->getSourceNamespace() << ": " << redact(startStatus);
_finishCallback_inlock(lk, startStatus);
return;
}
}
void DatabaseCloner::_collectionClonerCallback(const Status& status, const NamespaceString& nss) {
auto newStatus = status;
UniqueLock lk(_mutex);
if (!status.isOK()) {
newStatus = status.withContext(
str::stream() << "Error cloning collection '" << nss.toString() << "'");
_failedNamespaces.push_back({newStatus, nss});
}
++_stats.clonedCollections;
// Forward collection cloner result to caller.
// Failure to clone a collection does not stop the database cloner
// from cloning the rest of the collections in the listCollections result.
lk.unlock();
_collectionWork(newStatus, nss);
lk.lock();
_currentCollectionClonerIter++;
if (_currentCollectionClonerIter != _collectionCloners.end()) {
Status startStatus = _startCollectionCloner(*_currentCollectionClonerIter);
if (!startStatus.isOK()) {
LOG(1) << " failed to start collection cloning on "
<< _currentCollectionClonerIter->getSourceNamespace() << ": "
<< redact(startStatus);
_finishCallback_inlock(lk, startStatus);
return;
}
return;
}
Status finalStatus(Status::OK());
if (_failedNamespaces.size() > 0) {
finalStatus = {ErrorCodes::InitialSyncFailure,
str::stream() << "Failed to clone " << _failedNamespaces.size()
<< " collection(s) in '"
<< _dbname
<< "' from "
<< _source.toString()};
}
_finishCallback_inlock(lk, finalStatus);
}
void DatabaseCloner::_finishCallback(const Status& status) {
_onCompletion(status);
LockGuard lk(_mutex);
invariant(_state != State::kComplete);
_state = State::kComplete;
_condition.notify_all();
_stats.end = _executor->now();
LOG(1) << " database: " << _dbname << ", stats: " << _stats.toString();
}
void DatabaseCloner::_finishCallback_inlock(UniqueLock& lk, const Status& status) {
if (lk.owns_lock()) {
lk.unlock();
}
_finishCallback(status);
}
std::string DatabaseCloner::getDBName() const {
return _dbname;
}
std::string DatabaseCloner::Stats::toString() const {
return toBSON().toString();
}
BSONObj DatabaseCloner::Stats::toBSON() const {
BSONObjBuilder bob;
bob.append("dbname", dbname);
append(&bob);
return bob.obj();
}
void DatabaseCloner::Stats::append(BSONObjBuilder* builder) const {
builder->appendNumber("collections", collections);
builder->appendNumber("clonedCollections", clonedCollections);
if (start != Date_t()) {
builder->appendDate("start", start);
if (end != Date_t()) {
builder->appendDate("end", end);
auto elapsed = end - start;
long long elapsedMillis = duration_cast<Milliseconds>(elapsed).count();
builder->appendNumber("elapsedMillis", elapsedMillis);
}
}
for (auto&& collection : collectionStats) {
BSONObjBuilder collectionBuilder(builder->subobjStart(collection.ns));
collection.append(&collectionBuilder);
collectionBuilder.doneFast();
}
}
std::ostream& operator<<(std::ostream& os, const DatabaseCloner::State& state) {
switch (state) {
case DatabaseCloner::State::kPreStart:
return os << "PreStart";
case DatabaseCloner::State::kRunning:
return os << "Running";
case DatabaseCloner::State::kShuttingDown:
return os << "ShuttingDown";
case DatabaseCloner::State::kComplete:
return os << "Complete";
}
MONGO_UNREACHABLE;
}
} // namespace repl
} // namespace mongo
|