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
|
// collection.cpp
/**
* Copyright (C) 2013 10gen 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.
*/
#include "mongo/db/catalog/collection.h"
#include "mongo/base/counter.h"
#include "mongo/base/owned_pointer_map.h"
#include "mongo/db/clientcursor.h"
#include "mongo/db/commands/server_status_metric.h"
#include "mongo/db/curop.h"
#include "mongo/db/catalog/database.h"
#include "mongo/db/catalog/index_create.h"
#include "mongo/db/dbhelpers.h"
#include "mongo/db/index/index_access_method.h"
#include "mongo/db/ops/update.h"
#include "mongo/db/operation_context.h"
#include "mongo/db/structure/catalog/namespace_details.h"
#include "mongo/db/structure/catalog/namespace_details_rsv1_metadata.h"
#include "mongo/db/structure/record_store_v1_capped.h"
#include "mongo/db/structure/record_store_v1_simple.h"
#include "mongo/db/repl/rs.h"
#include "mongo/db/storage/extent.h"
#include "mongo/db/storage/extent_manager.h"
#include "mongo/db/storage/mmap_v1/mmap_v1_extent_manager.h"
#include "mongo/db/storage/record.h"
#include "mongo/db/auth/user_document_parser.h" // XXX-ANDY
namespace mongo {
std::string CompactOptions::toString() const {
std::stringstream ss;
ss << "paddingMode: ";
switch ( paddingMode ) {
case NONE:
ss << "NONE";
break;
case PRESERVE:
ss << "PRESERVE";
break;
case MANUAL:
ss << "MANUAL (" << paddingBytes << " + ( doc * " << paddingFactor <<") )";
}
ss << " validateDocuments: " << validateDocuments;
return ss.str();
}
// ----
Collection::Collection( OperationContext* txn,
const StringData& fullNS,
NamespaceDetails* details,
Database* database )
: _ns( fullNS ),
_infoCache( this ),
_indexCatalog( this, details ),
_cursorCache( fullNS ) {
_details = details;
_database = database;
if ( details->isCapped() ) {
_recordStore.reset( new CappedRecordStoreV1( txn,
this,
_ns.ns(),
new NamespaceDetailsRSV1MetaData( details ),
database->getExtentManager(),
_ns.coll() == "system.indexes" ) );
}
else {
_recordStore.reset( new SimpleRecordStoreV1( txn,
_ns.ns(),
new NamespaceDetailsRSV1MetaData( details ),
database->getExtentManager(),
_ns.coll() == "system.indexes" ) );
}
_magic = 1357924;
_indexCatalog.init(txn);
}
Collection::~Collection() {
verify( ok() );
_magic = 0;
}
bool Collection::requiresIdIndex() const {
if ( _ns.ns().find( '$' ) != string::npos ) {
// no indexes on indexes
return false;
}
if ( _ns == _database->_namespacesName ||
_ns == _database->_indexesName ||
_ns == _database->_profileName ) {
return false;
}
if ( _ns.db() == "local" ) {
if ( _ns.coll().startsWith( "oplog." ) )
return false;
}
if ( !_ns.isSystem() ) {
// non system collections definitely have an _id index
return true;
}
return true;
}
RecordIterator* Collection::getIterator( const DiskLoc& start, bool tailable,
const CollectionScanParams::Direction& dir) const {
invariant( ok() );
return _recordStore->getIterator( start, tailable, dir );
}
vector<RecordIterator*> Collection::getManyIterators() const {
return _recordStore->getManyIterators();
}
int64_t Collection::countTableScan( const MatchExpression* expression ) {
scoped_ptr<RecordIterator> iterator( getIterator( DiskLoc(),
false,
CollectionScanParams::FORWARD ) );
int64_t count = 0;
while ( !iterator->isEOF() ) {
DiskLoc loc = iterator->getNext();
BSONObj obj = docFor( loc );
if ( expression->matchesBSON( obj ) )
count++;
}
return count;
}
BSONObj Collection::docFor(const DiskLoc& loc) const {
Record* rec = _recordStore->recordFor( loc );
return BSONObj( rec->data() );
}
StatusWith<DiskLoc> Collection::insertDocument( OperationContext* txn,
const DocWriter* doc,
bool enforceQuota ) {
verify( _indexCatalog.numIndexesTotal() == 0 ); // eventually can implement, just not done
StatusWith<DiskLoc> loc = _recordStore->insertRecord( txn,
doc,
enforceQuota
? largestFileNumberInQuota()
: 0 );
if ( !loc.isOK() )
return loc;
return StatusWith<DiskLoc>( loc );
}
StatusWith<DiskLoc> Collection::insertDocument( OperationContext* txn,
const BSONObj& docToInsert,
bool enforceQuota ) {
if ( _indexCatalog.findIdIndex() ) {
if ( docToInsert["_id"].eoo() ) {
return StatusWith<DiskLoc>( ErrorCodes::InternalError,
str::stream() << "Collection::insertDocument got "
"document without _id for ns:" << _ns.ns() );
}
}
if ( isCapped() ) {
// TOOD: old god not done
Status ret = _indexCatalog.checkNoIndexConflicts( docToInsert );
if ( !ret.isOK() )
return StatusWith<DiskLoc>( ret );
}
StatusWith<DiskLoc> status = _insertDocument( txn, docToInsert, enforceQuota );
if ( status.isOK() ) {
_details->paddingFits( txn );
}
return status;
}
StatusWith<DiskLoc> Collection::insertDocument( OperationContext* txn,
const BSONObj& doc,
MultiIndexBlock& indexBlock ) {
StatusWith<DiskLoc> loc = _recordStore->insertRecord( txn,
doc.objdata(),
doc.objsize(),
0 );
if ( !loc.isOK() )
return loc;
InsertDeleteOptions indexOptions;
indexOptions.logIfError = false;
indexOptions.dupsAllowed = true; // in repair we should be doing no checking
Status status = indexBlock.insert( doc, loc.getValue(), indexOptions );
if ( !status.isOK() )
return StatusWith<DiskLoc>( status );
return loc;
}
StatusWith<DiskLoc> Collection::_insertDocument( OperationContext* txn,
const BSONObj& docToInsert,
bool enforceQuota ) {
// TODO: for now, capped logic lives inside NamespaceDetails, which is hidden
// under the RecordStore, this feels broken since that should be a
// collection access method probably
StatusWith<DiskLoc> loc = _recordStore->insertRecord( txn,
docToInsert.objdata(),
docToInsert.objsize(),
enforceQuota ? largestFileNumberInQuota() : 0 );
if ( !loc.isOK() )
return loc;
_infoCache.notifyOfWriteOp();
try {
_indexCatalog.indexRecord(txn, docToInsert, loc.getValue());
}
catch ( AssertionException& e ) {
if ( isCapped() ) {
return StatusWith<DiskLoc>( ErrorCodes::InternalError,
str::stream() << "unexpected index insertion failure on"
<< " capped collection" << e.toString()
<< " - collection and its index will not match" );
}
// indexRecord takes care of rolling back indexes
// so we just have to delete the main storage
_recordStore->deleteRecord( txn, loc.getValue() );
return StatusWith<DiskLoc>( e.toStatus( "insertDocument" ) );
}
return loc;
}
Status Collection::aboutToDeleteCapped( OperationContext* txn, const DiskLoc& loc ) {
BSONObj doc = docFor( loc );
/* check if any cursors point to us. if so, advance them. */
_cursorCache.invalidateDocument(loc, INVALIDATION_DELETION);
_indexCatalog.unindexRecord(txn, doc, loc, false);
return Status::OK();
}
void Collection::deleteDocument( OperationContext* txn,
const DiskLoc& loc,
bool cappedOK,
bool noWarn,
BSONObj* deletedId ) {
if ( isCapped() && !cappedOK ) {
log() << "failing remove on a capped ns " << _ns << endl;
uasserted( 10089, "cannot remove from a capped collection" );
return;
}
BSONObj doc = docFor( loc );
if ( deletedId ) {
BSONElement e = doc["_id"];
if ( e.type() ) {
*deletedId = e.wrap();
}
}
/* check if any cursors point to us. if so, advance them. */
_cursorCache.invalidateDocument(loc, INVALIDATION_DELETION);
_indexCatalog.unindexRecord(txn, doc, loc, noWarn);
_recordStore->deleteRecord( txn, loc );
_infoCache.notifyOfWriteOp();
}
Counter64 moveCounter;
ServerStatusMetricField<Counter64> moveCounterDisplay( "record.moves", &moveCounter );
StatusWith<DiskLoc> Collection::updateDocument( OperationContext* txn,
const DiskLoc& oldLocation,
const BSONObj& objNew,
bool enforceQuota,
OpDebug* debug ) {
Record* oldRecord = _recordStore->recordFor( oldLocation );
BSONObj objOld( oldRecord->data() );
if ( objOld.hasElement( "_id" ) ) {
BSONElement oldId = objOld["_id"];
BSONElement newId = objNew["_id"];
if ( oldId != newId )
return StatusWith<DiskLoc>( ErrorCodes::InternalError,
"in Collection::updateDocument _id mismatch",
13596 );
}
if ( ns().coll() == "system.users" ) {
// XXX - andy and spencer think this should go away now
V2UserDocumentParser parser;
Status s = parser.checkValidUserDocument(objNew);
if ( !s.isOK() )
return StatusWith<DiskLoc>( s );
}
/* duplicate key check. we descend the btree twice - once for this check, and once for the actual inserts, further
below. that is suboptimal, but it's pretty complicated to do it the other way without rollbacks...
*/
OwnedPointerMap<IndexDescriptor*,UpdateTicket> updateTickets;
IndexCatalog::IndexIterator ii = _indexCatalog.getIndexIterator( true );
while ( ii.more() ) {
IndexDescriptor* descriptor = ii.next();
IndexAccessMethod* iam = _indexCatalog.getIndex( descriptor );
InsertDeleteOptions options;
options.logIfError = false;
options.dupsAllowed =
!(KeyPattern::isIdKeyPattern(descriptor->keyPattern()) || descriptor->unique())
|| ignoreUniqueIndex(descriptor);
UpdateTicket* updateTicket = new UpdateTicket();
updateTickets.mutableMap()[descriptor] = updateTicket;
Status ret = iam->validateUpdate(objOld, objNew, oldLocation, options, updateTicket );
if ( !ret.isOK() ) {
return StatusWith<DiskLoc>( ret );
}
}
if ( oldRecord->netLength() < objNew.objsize() ) {
// doesn't fit, have to move to new location
if ( isCapped() )
return StatusWith<DiskLoc>( ErrorCodes::InternalError,
"failing update: objects in a capped ns cannot grow",
10003 );
moveCounter.increment();
_details->paddingTooSmall( txn );
// unindex old record, don't delete
// this way, if inserting new doc fails, we can re-index this one
_cursorCache.invalidateDocument(oldLocation, INVALIDATION_DELETION);
_indexCatalog.unindexRecord(txn, objOld, oldLocation, true);
if ( debug ) {
if (debug->nmoved == -1) // default of -1 rather than 0
debug->nmoved = 1;
else
debug->nmoved += 1;
}
StatusWith<DiskLoc> loc = _insertDocument( txn, objNew, enforceQuota );
if ( loc.isOK() ) {
// insert successful, now lets deallocate the old location
// remember its already unindexed
_recordStore->deleteRecord( txn, oldLocation );
}
else {
// new doc insert failed, so lets re-index the old document and location
_indexCatalog.indexRecord(txn, objOld, oldLocation);
}
return loc;
}
_infoCache.notifyOfWriteOp();
_details->paddingFits( txn );
if ( debug )
debug->keyUpdates = 0;
ii = _indexCatalog.getIndexIterator( true );
while ( ii.more() ) {
IndexDescriptor* descriptor = ii.next();
IndexAccessMethod* iam = _indexCatalog.getIndex( descriptor );
int64_t updatedKeys;
Status ret = iam->update(txn, *updateTickets.mutableMap()[descriptor], &updatedKeys);
if ( !ret.isOK() )
return StatusWith<DiskLoc>( ret );
if ( debug )
debug->keyUpdates += updatedKeys;
}
// Broadcast the mutation so that query results stay correct.
_cursorCache.invalidateDocument(oldLocation, INVALIDATION_MUTATION);
// update in place
int sz = objNew.objsize();
memcpy(txn->writingPtr(oldRecord->data(), sz), objNew.objdata(), sz);
return StatusWith<DiskLoc>( oldLocation );
}
Status Collection::updateDocumentWithDamages( OperationContext* txn,
const DiskLoc& loc,
const char* damangeSource,
const mutablebson::DamageVector& damages ) {
// Broadcast the mutation so that query results stay correct.
_cursorCache.invalidateDocument(loc, INVALIDATION_MUTATION);
_details->paddingFits( txn );
Record* rec = _recordStore->recordFor( loc );
char* root = rec->data();
// All updates were in place. Apply them via durability and writing pointer.
mutablebson::DamageVector::const_iterator where = damages.begin();
const mutablebson::DamageVector::const_iterator end = damages.end();
for( ; where != end; ++where ) {
const char* sourcePtr = damangeSource + where->sourceOffset;
void* targetPtr = txn->writingPtr(root + where->targetOffset, where->size);
std::memcpy(targetPtr, sourcePtr, where->size);
}
return Status::OK();
}
ExtentManager* Collection::getExtentManager() {
verify( ok() );
return _database->getExtentManager();
}
const ExtentManager* Collection::getExtentManager() const {
verify( ok() );
return _database->getExtentManager();
}
void Collection::increaseStorageSize(OperationContext* txn, int size, bool enforceQuota) {
_recordStore->increaseStorageSize(txn, size, enforceQuota ? largestFileNumberInQuota() : 0);
}
int Collection::largestFileNumberInQuota() const {
if ( !storageGlobalParams.quota )
return 0;
if ( _ns.db() == "local" )
return 0;
if ( _ns.isSpecial() )
return 0;
return storageGlobalParams.quotaFiles;
}
void Collection::appendCustomStats( BSONObjBuilder* result, double scale ) const {
result->append( "lastExtentSize", _details->lastExtentSize() / scale );
result->append( "paddingFactor", _details->paddingFactor() );
result->append( "userFlags", _details->userFlags() );
if ( isCapped() ) {
result->appendBool( "capped", true );
result->appendNumber( "max", _details->maxCappedDocs() );
}
}
bool Collection::isCapped() const {
return _details->isCapped();
}
uint64_t Collection::numRecords() const {
return _recordStore->numRecords();
}
uint64_t Collection::dataSize() const {
return _recordStore->dataSize();
}
/**
* order will be:
* 1) store index specs
* 2) drop indexes
* 3) truncate record store
* 4) re-write indexes
*/
Status Collection::truncate(OperationContext* txn) {
massert( 17445, "index build in progress", _indexCatalog.numIndexesInProgress() == 0 );
// 1) store index specs
vector<BSONObj> indexSpecs;
{
IndexCatalog::IndexIterator ii = _indexCatalog.getIndexIterator( false );
while ( ii.more() ) {
const IndexDescriptor* idx = ii.next();
indexSpecs.push_back( idx->infoObj().getOwned() );
}
}
// 2) drop indexes
Status status = _indexCatalog.dropAllIndexes(txn, true);
if ( !status.isOK() )
return status;
_cursorCache.invalidateAll( false );
_infoCache.reset();
// 3) truncate record store
status = _recordStore->truncate(txn);
if ( !status.isOK() )
return status;
// 4) re-create indexes
for ( size_t i = 0; i < indexSpecs.size(); i++ ) {
status = _indexCatalog.createIndex(txn, indexSpecs[i], false);
if ( !status.isOK() )
return status;
}
return Status::OK();
}
void Collection::temp_cappedTruncateAfter(OperationContext* txn,
DiskLoc end,
bool inclusive) {
invariant( isCapped() );
reinterpret_cast<CappedRecordStoreV1*>(
_recordStore.get())->temp_cappedTruncateAfter( txn, end, inclusive );
}
namespace {
class MyValidateAdaptor : public ValidateAdaptor {
public:
virtual ~MyValidateAdaptor(){}
virtual Status validate( Record* record, size_t* dataSize ) {
BSONObj obj = BSONObj( record->data() );
const Status status = validateBSON(obj.objdata(), obj.objsize());
if ( status.isOK() )
*dataSize = obj.objsize();
return Status::OK();
}
};
}
Status Collection::validate( OperationContext* txn,
bool full, bool scanData,
ValidateResults* results, BSONObjBuilder* output ){
MyValidateAdaptor adaptor;
Status status = _recordStore->validate( txn, full, scanData, &adaptor, results, output );
if ( !status.isOK() )
return status;
{ // indexes
output->append("nIndexes", _indexCatalog.numIndexesReady() );
int idxn = 0;
try {
BSONObjBuilder indexes; // not using subObjStart to be exception safe
IndexCatalog::IndexIterator i = _indexCatalog.getIndexIterator(false);
while( i.more() ) {
const IndexDescriptor* descriptor = i.next();
log() << "validating index " << descriptor->indexNamespace() << endl;
IndexAccessMethod* iam = _indexCatalog.getIndex( descriptor );
invariant( iam );
int64_t keys;
iam->validate(&keys);
indexes.appendNumber(descriptor->indexNamespace(),
static_cast<long long>(keys));
idxn++;
}
output->append("keysPerIndex", indexes.done());
}
catch ( DBException& exc ) {
string err = str::stream() <<
"exception during index validate idxn "<<
BSONObjBuilder::numStr(idxn) <<
": " << exc.toString();
results->errors.push_back( err );
results->valid = false;
}
}
return Status::OK();
}
Status Collection::touch( OperationContext* txn,
bool touchData, bool touchIndexes,
BSONObjBuilder* output ) const {
if ( touchData ) {
BSONObjBuilder b;
Status status = _recordStore->touch( txn, &b );
output->append( "data", b.obj() );
if ( !status.isOK() )
return status;
}
if ( touchIndexes ) {
Timer t;
IndexCatalog::IndexIterator ii = _indexCatalog.getIndexIterator( false );
while ( ii.more() ) {
const IndexDescriptor* desc = ii.next();
const IndexAccessMethod* iam = _indexCatalog.getIndex( desc );
Status status = iam->touch( txn );
if ( !status.isOK() )
return status;
}
output->append( "indexes", BSON( "num" << _indexCatalog.numIndexesTotal() <<
"millis" << t.millis() ) );
}
return Status::OK();
}
bool Collection::isUserFlagSet( int flag ) const {
return _details->isUserFlagSet( flag );
}
bool Collection::setUserFlag( OperationContext* txn, int flag ) {
if ( !_details->setUserFlag( txn, flag ) )
return false;
_syncUserFlags(txn);
return true;
}
bool Collection::clearUserFlag( OperationContext* txn, int flag ) {
if ( !_details->clearUserFlag( txn, flag ) )
return false;
_syncUserFlags(txn);
return true;
}
void Collection::_syncUserFlags(OperationContext* txn) {
if ( _ns.coll() == "system.namespaces" )
return;
string system_namespaces = _ns.getSisterNS( "system.namespaces" );
Collection* coll = _database->getCollection( txn, system_namespaces );
DiskLoc oldLocation = Helpers::findOne( coll, BSON( "name" << _ns.ns() ), false );
fassert( 17247, !oldLocation.isNull() );
BSONObj oldEntry = coll->docFor( oldLocation );
BSONObj newEntry = applyUpdateOperators( oldEntry,
BSON( "$set" <<
BSON( "options.flags" <<
_details->userFlags() ) ) );
StatusWith<DiskLoc> loc = coll->updateDocument( txn, oldLocation, newEntry, false, NULL );
if ( !loc.isOK() ) {
// TODO: should this be an fassert?
error() << "syncUserFlags failed! "
<< " ns: " << _ns
<< " error: " << loc.toString();
}
}
void Collection::setMaxCappedDocs( OperationContext* txn, long long max ) {
_details->setMaxCappedDocs( txn, max );
}
}
|