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
|
// collection.h
/**
* Copyright (C) 2012-2014 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.
*/
#pragma once
#include <cstdint>
#include <memory>
#include <string>
#include "mongo/base/status.h"
#include "mongo/base/status_with.h"
#include "mongo/base/string_data.h"
#include "mongo/bson/mutable/damage_vector.h"
#include "mongo/db/catalog/collection_info_cache.h"
#include "mongo/db/catalog/cursor_manager.h"
#include "mongo/db/catalog/index_catalog.h"
#include "mongo/db/exec/collection_scan_common.h"
#include "mongo/db/namespace_string.h"
#include "mongo/db/op_observer.h"
#include "mongo/db/record_id.h"
#include "mongo/db/storage/capped_callback.h"
#include "mongo/db/storage/record_store.h"
#include "mongo/db/storage/snapshot.h"
#include "mongo/stdx/condition_variable.h"
#include "mongo/stdx/mutex.h"
namespace mongo {
class CollectionCatalogEntry;
class DatabaseCatalogEntry;
class ExtentManager;
class IndexCatalog;
class MatchExpression;
class MultiIndexBlock;
class OpDebug;
class OperationContext;
class RecordCursor;
class RecordFetcher;
class UpdateDriver;
class UpdateRequest;
struct CompactOptions {
CompactOptions() {
paddingMode = NONE;
validateDocuments = true;
paddingFactor = 1;
paddingBytes = 0;
}
// padding
enum PaddingMode { PRESERVE, NONE, MANUAL } paddingMode;
// only used if _paddingMode == MANUAL
double paddingFactor; // what to multiple document size by
int paddingBytes; // what to add to ducment size after multiplication
unsigned computeRecordSize(unsigned recordSize) const {
recordSize = static_cast<unsigned>(paddingFactor * recordSize);
recordSize += paddingBytes;
return recordSize;
}
// other
bool validateDocuments;
std::string toString() const;
};
struct CompactStats {
CompactStats() {
corruptDocuments = 0;
}
long long corruptDocuments;
};
/**
* Queries with the awaitData option use this notifier object to wait for more data to be
* inserted into the capped collection.
*/
class CappedInsertNotifier {
public:
CappedInsertNotifier();
/**
* Wakes up all threads waiting.
*/
void notifyAll();
/**
* Waits for 'timeout' microseconds, or until notifyAll() is called to indicate that new
* data is available in the capped collection.
*
* NOTE: Waiting threads can be signaled by calling kill or notify* methods.
*/
void wait(Microseconds timeout) const;
/**
* Same as above but also ensures that if the version has changed, it also returns.
*/
void wait(uint64_t prevVersion, Microseconds timeout) const;
/**
* Returns the version for use as an additional wake condition when used above.
*/
uint64_t getVersion() const {
return _version;
}
/**
* Same as above but without a timeout.
*/
void wait() const;
/**
* Cancels the notifier if the collection is dropped/invalidated, and wakes all waiting.
*/
void kill();
/**
* Returns true if no new insert notification will occur.
*/
bool isDead();
private:
// Helper for wait impls.
void _wait(stdx::unique_lock<stdx::mutex>& lk,
uint64_t prevVersion,
Microseconds timeout) const;
// Signalled when a successful insert is made into a capped collection.
mutable stdx::condition_variable _notifier;
// Mutex used with '_notifier'. Protects access to '_version'.
mutable stdx::mutex _mutex;
// A counter, incremented on insertion of new data into the capped collection.
//
// The condition which '_cappedNewDataNotifier' is being notified of is an increment of this
// counter. Access to this counter is synchronized with '_mutex'.
uint64_t _version;
// True once the notifier is dead.
bool _dead;
};
/**
* this is NOT safe through a yield right now
* not sure if it will be, or what yet
*/
class Collection final : CappedCallback, UpdateNotifier {
public:
Collection(OperationContext* txn,
StringData fullNS,
CollectionCatalogEntry* details, // does not own
RecordStore* recordStore, // does not own
DatabaseCatalogEntry* dbce); // does not own
~Collection();
bool ok() const {
return _magic == 1357924;
}
CollectionCatalogEntry* getCatalogEntry() {
return _details;
}
const CollectionCatalogEntry* getCatalogEntry() const {
return _details;
}
CollectionInfoCache* infoCache() {
return &_infoCache;
}
const CollectionInfoCache* infoCache() const {
return &_infoCache;
}
const NamespaceString& ns() const {
return _ns;
}
const IndexCatalog* getIndexCatalog() const {
return &_indexCatalog;
}
IndexCatalog* getIndexCatalog() {
return &_indexCatalog;
}
const RecordStore* getRecordStore() const {
return _recordStore;
}
RecordStore* getRecordStore() {
return _recordStore;
}
CursorManager* getCursorManager() const {
return &_cursorManager;
}
bool requiresIdIndex() const;
Snapshotted<BSONObj> docFor(OperationContext* txn, const RecordId& loc) const;
/**
* @param out - contents set to the right docs if exists, or nothing.
* @return true iff loc exists
*/
bool findDoc(OperationContext* txn, const RecordId& loc, Snapshotted<BSONObj>* out) const;
std::unique_ptr<SeekableRecordCursor> getCursor(OperationContext* txn,
bool forward = true) const;
/**
* Returns many cursors that partition the Collection into many disjoint sets. Iterating
* all returned cursors is equivalent to iterating the full collection.
*/
std::vector<std::unique_ptr<RecordCursor>> getManyCursors(OperationContext* txn) const;
/**
* Deletes the document with the given RecordId from the collection.
*
* 'fromMigrate' indicates whether the delete was induced by a chunk migration, and
* so should be ignored by the user as an internal maintenance operation and not a
* real delete.
* 'loc' key to uniquely identify a record in a collection.
* 'cappedOK' if true, allows deletes on capped collections (Cloner::copyDB uses this).
* 'noWarn' if unindexing the record causes an error, if noWarn is true the error
* will not be logged.
*/
void deleteDocument(OperationContext* txn,
const RecordId& loc,
bool fromMigrate = false,
bool noWarn = false);
/*
* Inserts all documents inside one WUOW.
* Caller should ensure vector is appropriately sized for this.
* If any errors occur (including WCE), caller should retry documents individually.
*/
Status insertDocuments(OperationContext* txn,
std::vector<BSONObj>::const_iterator begin,
std::vector<BSONObj>::const_iterator end,
bool enforceQuota,
bool fromMigrate = false);
/**
* this does NOT modify the doc before inserting
* i.e. will not add an _id field for documents that are missing it
*
* If enforceQuota is false, quotas will be ignored.
*/
Status insertDocument(OperationContext* txn,
const BSONObj& doc,
bool enforceQuota,
bool fromMigrate = false);
/**
* Callers must ensure no document validation is performed for this collection when calling
* this method.
*/
Status insertDocument(OperationContext* txn, const DocWriter* doc, bool enforceQuota);
Status insertDocument(OperationContext* txn,
const BSONObj& doc,
MultiIndexBlock* indexBlock,
bool enforceQuota);
/**
* Updates the document @ oldLocation with newDoc.
*
* If the document fits in the old space, it is put there; if not, it is moved.
* Sets 'args.updatedDoc' to the updated version of the document with damages applied, on
* success.
* @return the post update location of the doc (may or may not be the same as oldLocation)
*/
StatusWith<RecordId> updateDocument(OperationContext* txn,
const RecordId& oldLocation,
const Snapshotted<BSONObj>& oldDoc,
const BSONObj& newDoc,
bool enforceQuota,
bool indexesAffected,
OpDebug* debug,
OplogUpdateEntryArgs* args);
bool updateWithDamagesSupported() const;
/**
* Not allowed to modify indexes.
* Illegal to call if updateWithDamagesSupported() returns false.
* Sets 'args.updatedDoc' to the updated version of the document with damages applied, on
* success.
* @return the contents of the updated record.
*/
StatusWith<RecordData> updateDocumentWithDamages(OperationContext* txn,
const RecordId& loc,
const Snapshotted<RecordData>& oldRec,
const char* damageSource,
const mutablebson::DamageVector& damages,
OplogUpdateEntryArgs* args);
// -----------
StatusWith<CompactStats> compact(OperationContext* txn, const CompactOptions* options);
/**
* removes all documents as fast as possible
* indexes before and after will be the same
* as will other characteristics
*/
Status truncate(OperationContext* txn);
/**
* @param full - does more checks
* @param scanData - scans each document
* @return OK if the validate run successfully
* OK will be returned even if corruption is found
* deatils will be in result
*/
Status validate(OperationContext* txn,
bool full,
bool scanData,
ValidateResults* results,
BSONObjBuilder* output);
/**
* forces data into cache
*/
Status touch(OperationContext* txn,
bool touchData,
bool touchIndexes,
BSONObjBuilder* output) const;
/**
* Truncate documents newer than the document at 'end' from the capped
* collection. The collection cannot be completely emptied using this
* function. An assertion will be thrown if that is attempted.
* @param inclusive - Truncate 'end' as well iff true
* XXX: this will go away soon, just needed to move for now
*/
void temp_cappedTruncateAfter(OperationContext* txn, RecordId end, bool inclusive);
/**
* Sets the validator for this collection.
*
* An empty validator removes all validation.
* Requires an exclusive lock on the collection.
*/
Status setValidator(OperationContext* txn, BSONObj validator);
Status setValidationLevel(OperationContext* txn, StringData newLevel);
Status setValidationAction(OperationContext* txn, StringData newAction);
StringData getValidationLevel() const;
StringData getValidationAction() const;
// -----------
//
// Stats
//
bool isCapped() const;
/**
* Get a pointer to a capped insert notifier object. The caller can wait on this object
* until it is notified of a new insert into the capped collection.
*
* It is invalid to call this method unless the collection is capped.
*/
std::shared_ptr<CappedInsertNotifier> getCappedInsertNotifier() const;
uint64_t numRecords(OperationContext* txn) const;
uint64_t dataSize(OperationContext* txn) const;
int averageObjectSize(OperationContext* txn) const {
uint64_t n = numRecords(txn);
if (n == 0)
return 5;
return static_cast<int>(dataSize(txn) / n);
}
uint64_t getIndexSize(OperationContext* opCtx, BSONObjBuilder* details = NULL, int scale = 1);
/**
* If return value is not boost::none, reads with majority read concern using an older snapshot
* must error.
*/
boost::optional<SnapshotName> getMinimumVisibleSnapshot() {
return _minVisibleSnapshot;
}
void setMinimumVisibleSnapshot(SnapshotName name) {
_minVisibleSnapshot = name;
}
/**
* Notify (capped collection) waiters of data changes, like an insert.
*/
void notifyCappedWaitersIfNeeded();
private:
/**
* Returns a non-ok Status if document does not pass this collection's validator.
*/
Status checkValidation(OperationContext* txn, const BSONObj& document) const;
/**
* Returns a non-ok Status if validator is not legal for this collection.
*/
StatusWithMatchExpression parseValidator(const BSONObj& validator) const;
Status recordStoreGoingToMove(OperationContext* txn,
const RecordId& oldLocation,
const char* oldBuffer,
size_t oldSize);
Status recordStoreGoingToUpdateInPlace(OperationContext* txn, const RecordId& loc);
Status aboutToDeleteCapped(OperationContext* txn, const RecordId& loc, RecordData data);
/**
* same semantics as insertDocument, but doesn't do:
* - some user error checks
* - adjust padding
*/
Status _insertDocument(OperationContext* txn, const BSONObj& doc, bool enforceQuota);
Status _insertDocuments(OperationContext* txn,
std::vector<BSONObj>::const_iterator begin,
std::vector<BSONObj>::const_iterator end,
bool enforceQuota);
bool _enforceQuota(bool userEnforeQuota) const;
int _magic;
const NamespaceString _ns;
CollectionCatalogEntry* const _details;
RecordStore* const _recordStore;
DatabaseCatalogEntry* const _dbce;
const bool _needCappedLock;
CollectionInfoCache _infoCache;
IndexCatalog _indexCatalog;
// Empty means no filter.
BSONObj _validatorDoc;
// Points into _validatorDoc. Null means no filter.
std::unique_ptr<MatchExpression> _validator;
enum ValidationAction { WARN, ERROR_V } _validationAction;
enum ValidationLevel { OFF, MODERATE, STRICT_V } _validationLevel;
static StatusWith<ValidationLevel> _parseValidationLevel(StringData);
static StatusWith<ValidationAction> _parseValidationAction(StringData);
// this is mutable because read only users of the Collection class
// use it keep state. This seems valid as const correctness of Collection
// should be about the data.
mutable CursorManager _cursorManager;
// Notifier object for awaitData. Threads polling a capped collection for new data can wait
// on this object until notified of the arrival of new data.
//
// This is non-null if and only if the collection is a capped collection.
const std::shared_ptr<CappedInsertNotifier> _cappedNotifier;
const bool _mustTakeCappedLockOnInsert;
// The earliest snapshot that is allowed to use this collection.
boost::optional<SnapshotName> _minVisibleSnapshot;
friend class Database;
friend class IndexCatalog;
friend class NamespaceDetails;
};
}
|