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
|
// record_store_v1_capped.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/structure/record_store_v1_capped.h"
#include "mongo/db/catalog/collection.h"
#include "mongo/db/storage/extent.h"
#include "mongo/db/storage/extent_manager.h"
#include "mongo/db/storage/mmap_v1/dur_transaction.h"
#include "mongo/db/storage/record.h"
#include "mongo/db/structure/catalog/namespace_details.h"
#include "mongo/db/structure/record_store_v1_capped_iterator.h"
/*
capped collection layout
d's below won't exist if things align perfectly:
extent1 -> extent2 -> extent3
------------------- ----------------------- ---------------------
d r r r r r r r r d d r r r r d r r r r r d d r r r r r r r r r d
^ ^
oldest newest
^cappedFirstDeletedInCurExtent()
^cappedLastDelRecLastExtent()
^cappedListOfAllDeletedRecords()
*/
#define DDD(x)
namespace mongo {
CappedRecordStoreV1::CappedRecordStoreV1( TransactionExperiment* txn,
Collection* collection,
const StringData& ns,
RecordStoreV1MetaData* details,
ExtentManager* em,
bool isSystemIndexes )
: RecordStoreV1Base( ns, details, em, isSystemIndexes ),
_collection( collection ) {
DiskLoc extentLoc = details->firstExtent();
while ( !extentLoc.isNull() ) {
Extent* extent = em->getExtent( extentLoc );
extentLoc = extent->xnext;
MAdvise* m( new MAdvise( reinterpret_cast<void*>( extent ),
extent->length,
MAdvise::Sequential ) );
_extentAdvice.mutableVector().push_back( m );
}
// this is for VERY VERY old versions of capped collections
cappedCheckMigrate(txn);
}
CappedRecordStoreV1::~CappedRecordStoreV1() {
}
StatusWith<DiskLoc> CappedRecordStoreV1::allocRecord( TransactionExperiment* txn,
int lenToAlloc,
int quotaMax ) {
{
// align very slightly.
lenToAlloc = (lenToAlloc + 3) & 0xfffffffc;
}
if ( lenToAlloc > theCapExtent()->length ) {
// the extent check is a way to try and improve performance
// since we have to iterate all the extents (for now) to get
// storage size
if ( lenToAlloc > _collection->storageSize() ) {
return StatusWith<DiskLoc>( ErrorCodes::BadValue,
str::stream() << "document is larger than capped size "
<< lenToAlloc << " > " << _collection->storageSize(),
16328 );
}
}
DiskLoc loc;
{ // do allocation
// signal done allocating new extents.
if ( !cappedLastDelRecLastExtent().isValid() )
setLastDelRecLastExtent( txn, DiskLoc() );
invariant( lenToAlloc < 400000000 );
int passes = 0;
int maxPasses = ( lenToAlloc / 30 ) + 2; // 30 is about the smallest entry that could go in the oplog
if ( maxPasses < 5000 ) {
// this is for bacwards safety since 5000 was the old value
maxPasses = 5000;
}
// delete records until we have room and the max # objects limit achieved.
/* this fails on a rename -- that is ok but must keep commented out */
//invariant( theCapExtent()->ns == ns );
theCapExtent()->assertOk();
DiskLoc firstEmptyExtent;
while ( 1 ) {
if ( _details->numRecords() < _details->maxCappedDocs() ) {
loc = __capAlloc( txn, lenToAlloc );
if ( !loc.isNull() )
break;
}
// If on first iteration through extents, don't delete anything.
if ( !_details->capFirstNewRecord().isValid() ) {
advanceCapExtent( txn, _ns );
if ( _details->capExtent() != _details->firstExtent() )
_details->setCapFirstNewRecord( txn, DiskLoc().setInvalid() );
// else signal done with first iteration through extents.
continue;
}
if ( !_details->capFirstNewRecord().isNull() &&
theCapExtent()->firstRecord == _details->capFirstNewRecord() ) {
// We've deleted all records that were allocated on the previous
// iteration through this extent.
advanceCapExtent( txn, _ns );
continue;
}
if ( theCapExtent()->firstRecord.isNull() ) {
if ( firstEmptyExtent.isNull() )
firstEmptyExtent = _details->capExtent();
advanceCapExtent( txn, _ns );
if ( firstEmptyExtent == _details->capExtent() ) {
_maybeComplain( lenToAlloc );
return StatusWith<DiskLoc>( ErrorCodes::InternalError,
"no space in capped collection" );
}
continue;
}
DiskLoc fr = theCapExtent()->firstRecord;
_collection->deleteDocument( fr, true );
compact(txn);
if( ++passes > maxPasses ) {
StringBuilder sb;
sb << "passes >= maxPasses in CappedRecordStoreV1::cappedAlloc: ns: " << _ns
<< ", lenToAlloc: " << lenToAlloc
<< ", maxPasses: " << maxPasses
<< ", _maxDocsInCapped: " << _details->maxCappedDocs()
<< ", nrecords: " << _details->numRecords()
<< ", datasize: " << _details->dataSize();
return StatusWith<DiskLoc>( ErrorCodes::InternalError, sb.str() );
}
}
// Remember first record allocated on this iteration through capExtent.
if ( _details->capFirstNewRecord().isValid() && _details->capFirstNewRecord().isNull() )
_details->setCapFirstNewRecord( txn, loc );
}
invariant( !loc.isNull() );
// possibly slice up if we've allocated too much space
DeletedRecord *r = drec( loc );
/* note we want to grab from the front so our next pointers on disk tend
to go in a forward direction which is important for performance. */
int regionlen = r->lengthWithHeaders();
invariant( r->extentOfs() < loc.getOfs() );
int left = regionlen - lenToAlloc;
/* split off some for further use. */
txn->writingInt(r->lengthWithHeaders()) = lenToAlloc;
DiskLoc newDelLoc = loc;
newDelLoc.inc(lenToAlloc);
DeletedRecord* newDel = drec( newDelLoc );
DeletedRecord* newDelW = txn->writing(newDel);
newDelW->extentOfs() = r->extentOfs();
newDelW->lengthWithHeaders() = left;
newDelW->nextDeleted().Null();
addDeletedRec(txn, newDelLoc);
return StatusWith<DiskLoc>( loc );
}
Status CappedRecordStoreV1::truncate(TransactionExperiment* txn) {
setLastDelRecLastExtent( txn, DiskLoc() );
setListOfAllDeletedRecords( txn, DiskLoc() );
// preserve firstExtent/lastExtent
_details->setCapExtent( txn, _details->firstExtent() );
_details->setStats( txn, 0, 0 );
// preserve lastExtentSize
// nIndexes preserve 0
// capped preserve true
// max preserve
_details->setPaddingFactor( txn, 1.0 );
_details->clearSystemFlags(txn);
_details->setCapFirstNewRecord( txn, DiskLoc().setInvalid() );
setLastDelRecLastExtent( txn, DiskLoc().setInvalid() );
// dataFileVersion preserve
// indexFileVersion preserve
// Reset all existing extents and recreate the deleted list.
for( DiskLoc ext = _details->firstExtent();
!ext.isNull();
ext = _extentManager->getExtent(ext)->xnext ) {
DiskLoc prev = _extentManager->getExtent(ext)->xprev;
DiskLoc next = _extentManager->getExtent(ext)->xnext;
DiskLoc empty = _extentManager->getExtent(ext)->reuse( _ns, true );
*txn->writing(&_extentManager->getExtent(ext)->xprev) = prev;
*txn->writing(&_extentManager->getExtent(ext)->xnext) = next;
addDeletedRec( txn, empty );
}
return Status::OK();
}
void CappedRecordStoreV1::temp_cappedTruncateAfter( TransactionExperiment* txn,
DiskLoc end,
bool inclusive ) {
cappedTruncateAfter( txn, _ns.c_str(), end, inclusive );
}
/* combine adjacent deleted records *for the current extent* of the capped collection
this is O(n^2) but we call it for capped tables where typically n==1 or 2!
(or 3...there will be a little unused sliver at the end of the extent.)
*/
void CappedRecordStoreV1::compact(TransactionExperiment* txn) {
DDD( "CappedRecordStoreV1::compact enter" );
vector<DiskLoc> drecs;
// Pull out capExtent's DRs from deletedList
DiskLoc i = cappedFirstDeletedInCurExtent();
for (; !i.isNull() && inCapExtent( i ); i = deletedRecordFor( i )->nextDeleted() ) {
DDD( "\t" << i );
drecs.push_back( i );
}
setFirstDeletedInCurExtent( txn, i );
std::sort( drecs.begin(), drecs.end() );
DDD( "\t drecs.size(): " << drecs.size() );
vector<DiskLoc>::const_iterator j = drecs.begin();
invariant( j != drecs.end() );
DiskLoc a = *j;
while ( 1 ) {
j++;
if ( j == drecs.end() ) {
DDD( "\t compact adddelrec" );
addDeletedRec(txn, a);
break;
}
DiskLoc b = *j;
while ( a.a() == b.a() &&
a.getOfs() + drec( a )->lengthWithHeaders() == b.getOfs() ) {
// a & b are adjacent. merge.
txn->writingInt( drec(a)->lengthWithHeaders() ) += drec(b)->lengthWithHeaders();
j++;
if ( j == drecs.end() ) {
DDD( "\t compact adddelrec2" );
addDeletedRec(txn, a);
return;
}
b = *j;
}
DDD( "\t compact adddelrec3" );
addDeletedRec(txn, a);
a = b;
}
}
const DiskLoc &CappedRecordStoreV1::cappedFirstDeletedInCurExtent() const {
if ( cappedLastDelRecLastExtent().isNull() )
return cappedListOfAllDeletedRecords();
else
return drec(cappedLastDelRecLastExtent())->nextDeleted();
}
void CappedRecordStoreV1::setFirstDeletedInCurExtent( TransactionExperiment* txn,
const DiskLoc& loc ) {
if ( cappedLastDelRecLastExtent().isNull() )
setListOfAllDeletedRecords( txn, loc );
else
*txn->writing( &drec(cappedLastDelRecLastExtent())->nextDeleted() ) = loc;
}
void CappedRecordStoreV1::cappedCheckMigrate(TransactionExperiment* txn) {
// migrate old RecordStoreV1MetaData format
if ( _details->capExtent().a() == 0 && _details->capExtent().getOfs() == 0 ) {
_details->setCapFirstNewRecord( txn, DiskLoc().setInvalid() );
// put all the DeletedRecords in cappedListOfAllDeletedRecords()
for ( int i = 1; i < Buckets; ++i ) {
DiskLoc first = _details->deletedListEntry( i );
if ( first.isNull() )
continue;
DiskLoc last = first;
for (; !drec(last)->nextDeleted().isNull(); last = drec(last)->nextDeleted() );
*txn->writing(&drec(last)->nextDeleted()) = cappedListOfAllDeletedRecords();
setListOfAllDeletedRecords( txn, first );
_details->setDeletedListEntry(txn, i, DiskLoc());
}
// NOTE cappedLastDelRecLastExtent() set to DiskLoc() in above
// Last, in case we're killed before getting here
_details->setCapExtent( txn, _details->firstExtent() );
}
}
bool CappedRecordStoreV1::inCapExtent( const DiskLoc &dl ) const {
invariant( !dl.isNull() );
if ( dl.a() != _details->capExtent().a() )
return false;
if ( dl.getOfs() < _details->capExtent().getOfs() )
return false;
const Extent* e = theCapExtent();
int end = _details->capExtent().getOfs() + e->length;
return dl.getOfs() <= end;
}
bool CappedRecordStoreV1::nextIsInCapExtent( const DiskLoc &dl ) const {
invariant( !dl.isNull() );
DiskLoc next = drec(dl)->nextDeleted();
if ( next.isNull() )
return false;
return inCapExtent( next );
}
void CappedRecordStoreV1::advanceCapExtent( TransactionExperiment* txn, const StringData& ns ) {
// We want cappedLastDelRecLastExtent() to be the last DeletedRecord of the prev cap extent
// (or DiskLoc() if new capExtent == firstExtent)
if ( _details->capExtent() == _details->lastExtent() )
setLastDelRecLastExtent( txn, DiskLoc() );
else {
DiskLoc i = cappedFirstDeletedInCurExtent();
for (; !i.isNull() && nextIsInCapExtent( i ); i = drec(i)->nextDeleted() );
setLastDelRecLastExtent( txn, i );
}
_details->setCapExtent( txn,
theCapExtent()->xnext.isNull() ? _details->firstExtent()
: theCapExtent()->xnext );
/* this isn't true if a collection has been renamed...that is ok just used for diagnostics */
//dassert( theCapExtent()->ns == ns );
theCapExtent()->assertOk();
_details->setCapFirstNewRecord( txn, DiskLoc() );
}
DiskLoc CappedRecordStoreV1::__capAlloc( TransactionExperiment* txn, int len ) {
DiskLoc prev = cappedLastDelRecLastExtent();
DiskLoc i = cappedFirstDeletedInCurExtent();
DiskLoc ret;
for (; !i.isNull() && inCapExtent( i ); prev = i, i = drec(i)->nextDeleted() ) {
// We need to keep at least one DR per extent in cappedListOfAllDeletedRecords(),
// so make sure there's space to create a DR at the end.
if ( drec(i)->lengthWithHeaders() >= len + 24 ) {
ret = i;
break;
}
}
/* unlink ourself from the deleted list */
if ( !ret.isNull() ) {
if ( prev.isNull() )
setListOfAllDeletedRecords( txn, drec(ret)->nextDeleted() );
else
*txn->writing(&drec(prev)->nextDeleted()) = drec(ret)->nextDeleted();
*txn->writing(&drec(ret)->nextDeleted()) = DiskLoc().setInvalid(); // defensive.
invariant( drec(ret)->extentOfs() < ret.getOfs() );
}
return ret;
}
void CappedRecordStoreV1::cappedTruncateLastDelUpdate(TransactionExperiment* txn) {
if ( _details->capExtent() == _details->firstExtent() ) {
// Only one extent of the collection is in use, so there
// is no deleted record in a previous extent, so nullify
// cappedLastDelRecLastExtent().
setLastDelRecLastExtent( txn, DiskLoc() );
}
else {
// Scan through all deleted records in the collection
// until the last deleted record for the extent prior
// to the new capExtent is found. Then set
// cappedLastDelRecLastExtent() to that deleted record.
DiskLoc i = cappedListOfAllDeletedRecords();
for( ;
!drec(i)->nextDeleted().isNull() &&
!inCapExtent( drec(i)->nextDeleted() );
i = drec(i)->nextDeleted() );
// In our capped storage model, every extent must have at least one
// deleted record. Here we check that 'i' is not the last deleted
// record. (We expect that there will be deleted records in the new
// capExtent as well.)
invariant( !drec(i)->nextDeleted().isNull() );
setLastDelRecLastExtent( txn, i );
}
}
void CappedRecordStoreV1::cappedTruncateAfter(TransactionExperiment* txn,
const char* ns,
DiskLoc end,
bool inclusive) {
invariant( cappedLastDelRecLastExtent().isValid() );
// We iteratively remove the newest document until the newest document
// is 'end', then we remove 'end' if requested.
bool foundLast = false;
while( 1 ) {
if ( foundLast ) {
// 'end' has been found and removed, so break.
break;
}
txn->commitIfNeeded();
// 'curr' will point to the newest document in the collection.
DiskLoc curr = theCapExtent()->lastRecord;
invariant( !curr.isNull() );
if ( curr == end ) {
if ( inclusive ) {
// 'end' has been found, so break next iteration.
foundLast = true;
}
else {
// 'end' has been found, so break.
break;
}
}
// TODO The algorithm used in this function cannot generate an
// empty collection, but we could call emptyCappedCollection() in
// this case instead of asserting.
uassert( 13415, "emptying the collection is not allowed", _details->numRecords() > 1 );
// Delete the newest record, and coalesce the new deleted
// record with existing deleted records.
_collection->deleteDocument( curr, true );
compact(txn);
// This is the case where we have not yet had to remove any
// documents to make room for other documents, and we are allocating
// documents from free space in fresh extents instead of reusing
// space from familiar extents.
if ( !capLooped() ) {
// We just removed the last record from the 'capExtent', and
// the 'capExtent' can't be empty, so we set 'capExtent' to
// capExtent's prev extent.
if ( theCapExtent()->lastRecord.isNull() ) {
invariant( !theCapExtent()->xprev.isNull() );
// NOTE Because we didn't delete the last document, and
// capLooped() is false, capExtent is not the first extent
// so xprev will be nonnull.
_details->setCapExtent( txn, theCapExtent()->xprev );
theCapExtent()->assertOk();
// update cappedLastDelRecLastExtent()
cappedTruncateLastDelUpdate(txn);
}
continue;
}
// This is the case where capLooped() is true, and we just deleted
// from capExtent, and we just deleted capFirstNewRecord, which was
// the last record on the fresh side of capExtent.
// NOTE In this comparison, curr and potentially capFirstNewRecord
// may point to invalid data, but we can still compare the
// references themselves.
if ( curr == _details->capFirstNewRecord() ) {
// Set 'capExtent' to the first nonempty extent prior to the
// initial capExtent. There must be such an extent because we
// have not deleted the last document in the collection. It is
// possible that all extents other than the capExtent are empty.
// In this case we will keep the initial capExtent and specify
// that all records contained within are on the fresh rather than
// stale side of the extent.
DiskLoc newCapExtent = _details->capExtent();
do {
// Find the previous extent, looping if necessary.
newCapExtent = ( newCapExtent == _details->firstExtent() ) ?
_details->lastExtent() :
_extentManager->getExtent(newCapExtent)->xprev;
_extentManager->getExtent(newCapExtent)->assertOk();
}
while ( _extentManager->getExtent(newCapExtent)->firstRecord.isNull() );
_details->setCapExtent( txn, newCapExtent );
// Place all documents in the new capExtent on the fresh side
// of the capExtent by setting capFirstNewRecord to the first
// document in the new capExtent.
_details->setCapFirstNewRecord( txn, theCapExtent()->firstRecord );
// update cappedLastDelRecLastExtent()
cappedTruncateLastDelUpdate(txn);
}
}
}
const DiskLoc& CappedRecordStoreV1::cappedListOfAllDeletedRecords() const {
return _details->deletedListEntry(0);
}
void CappedRecordStoreV1::setListOfAllDeletedRecords( TransactionExperiment* txn,
const DiskLoc& loc ) {
return _details->setDeletedListEntry(txn, 0, loc);
}
const DiskLoc& CappedRecordStoreV1::cappedLastDelRecLastExtent() const {
return _details->deletedListEntry(1);
}
void CappedRecordStoreV1::setLastDelRecLastExtent( TransactionExperiment* txn,
const DiskLoc& loc ) {
return _details->setDeletedListEntry(txn, 1, loc);
}
bool CappedRecordStoreV1::capLooped() const {
return _details->capFirstNewRecord().isValid();
}
Extent* CappedRecordStoreV1::theCapExtent() const {
return _extentManager->getExtent(_details->capExtent());
}
void CappedRecordStoreV1::addDeletedRec( TransactionExperiment* txn, const DiskLoc& dloc ) {
DeletedRecord* d = drec( dloc );
{
Record *r = (Record *) txn->writingPtr(d, sizeof(Record));
d = &r->asDeleted();
// defensive code: try to make us notice if we reference a deleted record
reinterpret_cast<unsigned*>( r->data() )[0] = 0xeeeeeeee;
}
DEBUGGING log() << "TEMP: add deleted rec " << dloc.toString() << ' ' << hex << d->extentOfs() << endl;
if ( !cappedLastDelRecLastExtent().isValid() ) {
// Initial extent allocation. Insert at end.
d->nextDeleted() = DiskLoc();
if ( cappedListOfAllDeletedRecords().isNull() )
setListOfAllDeletedRecords( txn, dloc );
else {
DiskLoc i = cappedListOfAllDeletedRecords();
for (; !drec(i)->nextDeleted().isNull(); i = drec(i)->nextDeleted() )
;
*txn->writing(&drec(i)->nextDeleted()) = dloc;
}
}
else {
d->nextDeleted() = cappedFirstDeletedInCurExtent();
setFirstDeletedInCurExtent( txn, dloc );
// always compact() after this so order doesn't matter
}
}
RecordIterator* CappedRecordStoreV1::getIterator( const DiskLoc& start, bool tailable,
const CollectionScanParams::Direction& dir) const {
return new CappedRecordStoreV1Iterator( this, start, tailable, dir );
}
vector<RecordIterator*> CappedRecordStoreV1::getManyIterators() const {
OwnedPointerVector<RecordIterator> iterators;
if (!capLooped()) {
// if we haven't looped yet, just spit out all extents (same as non-capped impl)
const Extent* ext;
for (DiskLoc extLoc = details()->firstExtent(); !extLoc.isNull(); extLoc = ext->xnext) {
ext = _getExtent(extLoc);
if (ext->firstRecord.isNull())
continue;
iterators.push_back(new RecordStoreV1Base::IntraExtentIterator(ext->firstRecord,
this));
}
}
else {
// if we've looped we need to iterate the extents, starting and ending with the
// capExtent
const DiskLoc capExtent = details()->capExtent();
invariant(!capExtent.isNull());
invariant(capExtent.isValid());
// First do the "old" portion of capExtent if there is any
DiskLoc extLoc = capExtent;
{
const Extent* ext = _getExtent(extLoc);
if (ext->firstRecord != details()->capFirstNewRecord()) {
// this means there is old data in capExtent
iterators.push_back(new RecordStoreV1Base::IntraExtentIterator(ext->firstRecord,
this));
}
extLoc = ext->xnext.isNull() ? details()->firstExtent() : ext->xnext;
}
// Next handle all the other extents
while (extLoc != capExtent) {
const Extent* ext = _getExtent(extLoc);
iterators.push_back(new RecordStoreV1Base::IntraExtentIterator(ext->firstRecord,
this));
extLoc = ext->xnext.isNull() ? details()->firstExtent() : ext->xnext;
}
// Finally handle the "new" data in the capExtent
iterators.push_back(
new RecordStoreV1Base::IntraExtentIterator(details()->capFirstNewRecord(), this));
}
return iterators.release();
}
Status CappedRecordStoreV1::compact( TransactionExperiment* txn,
RecordStoreCompactAdaptor* adaptor,
const CompactOptions* options,
CompactStats* stats ) {
invariant(false);
}
void CappedRecordStoreV1::_maybeComplain( int len ) const {
RARELY {
std::stringstream buf;
buf << "couldn't make room for record len: " << len << " in capped ns " << _ns << '\n';
int i = 0;
for ( DiskLoc e = _details->firstExtent();
!e.isNull();
e = _extentManager->getExtent( e )->xnext, ++i ) {
buf << " Extent " << i;
if ( e == _details->capExtent() )
buf << " (capExtent)";
buf << '\n';
buf << " magic: " << hex << _extentManager->getExtent( e )->magic << dec
<< " extent->ns: " << _extentManager->getExtent( e )->nsDiagnostic.toString()
<< '\n';
buf << " fr: " << _extentManager->getExtent( e )->firstRecord.toString()
<< " lr: " << _extentManager->getExtent( e )->lastRecord.toString()
<< " extent->len: " << _extentManager->getExtent( e )->length << '\n';
}
warning() << buf.str();
// assume it is unusually large record; if not, something is broken
fassert( 17438, len * 5 > _details->lastExtentSize() );
}
}
DiskLoc CappedRecordStoreV1::firstRecord( const DiskLoc &startExtent ) const {
for (DiskLoc i = startExtent.isNull() ? _details->firstExtent() : startExtent;
!i.isNull();
i = _extentManager->getExtent( i )->xnext ) {
Extent* e = _extentManager->getExtent( i );
if ( !e->firstRecord.isNull() )
return e->firstRecord;
}
return DiskLoc();
}
DiskLoc CappedRecordStoreV1::lastRecord( const DiskLoc &startExtent ) const {
for (DiskLoc i = startExtent.isNull() ? _details->lastExtent() : startExtent;
!i.isNull();
i = _extentManager->getExtent( i )->xprev ) {
Extent* e = _extentManager->getExtent( i );
if ( !e->lastRecord.isNull() )
return e->lastRecord;
}
return DiskLoc();
}
}
|