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
|
/**
* Copyright (C) 2019-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_LOG_DEFAULT_COMPONENT ::mongo::logger::LogComponent::kStorage
#include "mongo/platform/basic.h"
#include "mongo/db/catalog/validate_adaptor.h"
#include "mongo/bson/bsonobj.h"
#include "mongo/db/catalog/collection.h"
#include "mongo/db/catalog/index_catalog.h"
#include "mongo/db/catalog/index_consistency.h"
#include "mongo/db/catalog/throttle_cursor.h"
#include "mongo/db/curop.h"
#include "mongo/db/index/index_access_method.h"
#include "mongo/db/index/index_descriptor.h"
#include "mongo/db/index/wildcard_access_method.h"
#include "mongo/db/matcher/expression.h"
#include "mongo/db/operation_context.h"
#include "mongo/db/storage/key_string.h"
#include "mongo/db/storage/record_store.h"
#include "mongo/logv2/log.h"
#include "mongo/rpc/object_check.h"
#include "mongo/util/log.h"
namespace mongo {
namespace {
const long long kInterruptIntervalNumRecords = 4096;
const long long kInterruptIntervalNumBytes = 50 * 1024 * 1024; // 50MB.
} // namespace
Status ValidateAdaptor::validateRecord(OperationContext* opCtx,
const RecordId& recordId,
const RecordData& record,
size_t* dataSize) {
BSONObj recordBson;
try {
recordBson = record.toBson();
} catch (...) {
return exceptionToStatus();
}
const Status status = validateBSON(
recordBson.objdata(), recordBson.objsize(), Validator<BSONObj>::enabledBSONVersion());
if (status.isOK()) {
*dataSize = recordBson.objsize();
} else {
return status;
}
const IndexCatalog* indexCatalog = _validateState->getCollection()->getIndexCatalog();
if (!indexCatalog->haveAnyIndexes()) {
return status;
}
for (const auto& index : _validateState->getIndexes()) {
const IndexDescriptor* descriptor = index->descriptor();
const IndexAccessMethod* iam = index->accessMethod();
if (descriptor->isPartial()) {
const IndexCatalogEntry* ice = indexCatalog->getEntry(descriptor);
if (!ice->getFilterExpression()->matchesBSON(recordBson)) {
continue;
}
}
KeyStringSet documentKeySet;
KeyStringSet multikeyMetadataKeys;
MultikeyPaths multikeyPaths;
iam->getKeys(recordBson,
IndexAccessMethod::GetKeysMode::kEnforceConstraints,
IndexAccessMethod::GetKeysContext::kReadOrAddKeys,
&documentKeySet,
&multikeyMetadataKeys,
&multikeyPaths,
recordId,
IndexAccessMethod::kNoopOnSuppressedErrorFn);
if (!descriptor->isMultikey() &&
iam->shouldMarkIndexAsMultikey(
documentKeySet.size(),
{multikeyMetadataKeys.begin(), multikeyMetadataKeys.end()},
multikeyPaths)) {
std::string msg = str::stream()
<< "Index " << descriptor->indexName() << " is not multi-key but has more than one"
<< " key in document " << recordId;
ValidateResults& curRecordResults = (*_indexNsResultsMap)[descriptor->indexName()];
curRecordResults.errors.push_back(msg);
curRecordResults.valid = false;
}
IndexInfo& indexInfo = _indexConsistency->getIndexInfo(descriptor->indexName());
for (const auto& keyString : multikeyMetadataKeys) {
try {
_indexConsistency->addMultikeyMetadataPath(keyString, &indexInfo);
} catch (...) {
return exceptionToStatus();
}
}
for (const auto& keyString : documentKeySet) {
try {
_totalIndexKeys++;
_indexConsistency->addDocKey(opCtx, keyString, &indexInfo, recordId);
} catch (...) {
return exceptionToStatus();
}
}
}
return status;
}
void ValidateAdaptor::traverseIndex(OperationContext* opCtx,
const IndexCatalogEntry* index,
int64_t* numTraversedKeys,
ValidateResults* results) {
const IndexDescriptor* descriptor = index->descriptor();
auto indexName = descriptor->indexName();
IndexInfo& indexInfo = _indexConsistency->getIndexInfo(indexName);
int64_t numKeys = 0;
bool isFirstEntry = true;
// The progress meter will be inactive after traversing the record store to allow the message
// and the total to be set to different values.
if (!_progress->isActive()) {
const char* curopMessage = "Validate: scanning index entries";
stdx::unique_lock<Client> lk(*opCtx->getClient());
_progress.set(CurOp::get(opCtx)->setProgress_inlock(curopMessage, _totalIndexKeys));
}
const KeyString::Version version =
index->accessMethod()->getSortedDataInterface()->getKeyStringVersion();
KeyString::Builder firstKeyString(
version, BSONObj(), indexInfo.ord, KeyString::Discriminator::kExclusiveBefore);
KeyString::Value prevIndexKeyStringValue;
// Ensure that this index has an open index cursor.
const auto indexCursorIt = _validateState->getIndexCursors().find(indexName);
invariant(indexCursorIt != _validateState->getIndexCursors().end());
const std::unique_ptr<SortedDataInterfaceThrottleCursor>& indexCursor = indexCursorIt->second;
for (auto indexEntry = indexCursor->seekForKeyString(opCtx, firstKeyString.getValueCopy());
indexEntry;
indexEntry = indexCursor->nextKeyString(opCtx)) {
// Ensure that the index entries are in increasing or decreasing order.
if (!isFirstEntry && indexEntry->keyString < prevIndexKeyStringValue) {
if (results && results->valid) {
results->errors.push_back(
"one or more indexes are not in strictly ascending or descending order");
}
if (results) {
results->valid = false;
}
}
const RecordId kWildcardMultikeyMetadataRecordId{
RecordId::ReservedId::kWildcardMultikeyMetadataId};
if (descriptor->getIndexType() == IndexType::INDEX_WILDCARD &&
indexEntry->loc == kWildcardMultikeyMetadataRecordId) {
_indexConsistency->removeMultikeyMetadataPath(indexEntry->keyString, &indexInfo);
_progress->hit();
numKeys++;
continue;
}
_indexConsistency->addIndexKey(indexEntry->keyString, &indexInfo, indexEntry->loc);
_progress->hit();
numKeys++;
isFirstEntry = false;
prevIndexKeyStringValue = indexEntry->keyString;
if (numKeys % kInterruptIntervalNumRecords == 0) {
// Periodically checks for interrupts and yields.
opCtx->checkForInterrupt();
_validateState->yield(opCtx);
}
}
if (results && _indexConsistency->getMultikeyMetadataPathCount(&indexInfo) > 0) {
results->errors.push_back(str::stream()
<< "Index '" << descriptor->indexName()
<< "' has one or more missing multikey metadata index keys");
results->valid = false;
}
if (numTraversedKeys) {
*numTraversedKeys = numKeys;
}
}
void ValidateAdaptor::traverseRecordStore(OperationContext* opCtx,
ValidateResults* results,
BSONObjBuilder* output) {
_numRecords = 0; // need to reset it because this function can be called more than once.
long long dataSizeTotal = 0;
long long interruptIntervalNumBytes = 0;
long long nInvalid = 0;
results->valid = true;
RecordId prevRecordId;
// In case validation occurs twice and the progress meter persists after index traversal
if (_progress.get() && _progress->isActive()) {
_progress->finished();
}
// Because the progress meter is intended as an approximation, it's sufficient to get the number
// of records when we begin traversing, even if this number may deviate from the final number.
const char* curopMessage = "Validate: scanning documents";
const auto totalRecords = _validateState->getCollection()->getRecordStore()->numRecords(opCtx);
{
stdx::unique_lock<Client> lk(*opCtx->getClient());
_progress.set(CurOp::get(opCtx)->setProgress_inlock(curopMessage, totalRecords));
}
const std::unique_ptr<SeekableRecordThrottleCursor>& traverseRecordStoreCursor =
_validateState->getTraverseRecordStoreCursor();
for (auto record =
traverseRecordStoreCursor->seekExact(opCtx, _validateState->getFirstRecordId());
record;
record = traverseRecordStoreCursor->next(opCtx)) {
_progress->hit();
++_numRecords;
auto dataSize = record->data.size();
interruptIntervalNumBytes += dataSize;
dataSizeTotal += dataSize;
size_t validatedSize = 0;
Status status = validateRecord(opCtx, record->id, record->data, &validatedSize);
// Checks to ensure isInRecordIdOrder() is being used properly.
if (prevRecordId.isValid()) {
invariant(prevRecordId < record->id);
}
// validatedSize = dataSize is not a general requirement as some storage engines may use
// padding, but we still require that they return the unpadded record data.
if (!status.isOK() || validatedSize != static_cast<size_t>(dataSize)) {
str::stream ss;
ss << "Document with RecordId " << record->id << " is corrupted. ";
if (!status.isOK() && validatedSize != static_cast<size_t>(dataSize)) {
ss << "Reasons: (1) " << status << "; (2) Validated size of " << validatedSize
<< " bytes does not equal the record size of " << dataSize << " bytes";
} else if (!status.isOK()) {
ss << "Reason: " << status;
} else {
ss << "Reason: Validated size of " << validatedSize
<< " bytes does not equal the record size of " << dataSize << " bytes";
}
LOGV2(20404, "{std_string_ss}", "std_string_ss"_attr = std::string(ss));
// Only log once
if (results->valid) {
results->errors.push_back("Detected one or more invalid documents. See logs.");
results->valid = false;
}
nInvalid++;
}
prevRecordId = record->id;
if (_numRecords % kInterruptIntervalNumRecords == 0 ||
interruptIntervalNumBytes >= kInterruptIntervalNumBytes) {
// Periodically checks for interrupts and yields.
opCtx->checkForInterrupt();
_validateState->yield(opCtx);
if (interruptIntervalNumBytes >= kInterruptIntervalNumBytes) {
interruptIntervalNumBytes = 0;
}
}
}
// Do not update the record store stats if we're in the background as we've validated a
// checkpoint and it may not have the most up-to-date changes.
if (results->valid && !_validateState->isBackground()) {
_validateState->getCollection()->getRecordStore()->updateStatsAfterRepair(
opCtx, _numRecords, dataSizeTotal);
}
_progress->finished();
output->appendNumber("nInvalidDocuments", nInvalid);
output->appendNumber("nrecords", _numRecords);
}
void ValidateAdaptor::validateIndexKeyCount(const IndexDescriptor* idx, ValidateResults& results) {
// Fetch the total number of index entries we previously found traversing the index.
const std::string indexName = idx->indexName();
IndexInfo* indexInfo = &_indexConsistency->getIndexInfo(indexName);
auto numTotalKeys = indexInfo->numKeys;
// Do not fail on finding too few index entries compared to collection entries when full:false.
bool hasTooFewKeys = false;
bool noErrorOnTooFewKeys = !_validateState->isFullIndexValidation();
if (idx->isIdIndex() && numTotalKeys != _numRecords) {
hasTooFewKeys = (numTotalKeys < _numRecords);
std::string msg = str::stream()
<< "number of _id index entries (" << numTotalKeys
<< ") does not match the number of documents in the index (" << _numRecords << ")";
if (noErrorOnTooFewKeys && (numTotalKeys < _numRecords)) {
results.warnings.push_back(msg);
} else {
results.errors.push_back(msg);
results.valid = false;
}
}
// Confirm that the number of index entries is not greater than the number of documents in the
// collection. This check is only valid for indexes that are not multikey (indexed arrays
// produce an index key per array entry) and not $** indexes which can produce index keys for
// multiple paths within a single document.
if (results.valid && !idx->isMultikey() && idx->getIndexType() != IndexType::INDEX_WILDCARD &&
numTotalKeys > _numRecords) {
std::string err = str::stream()
<< "index " << idx->indexName() << " is not multi-key, but has more entries ("
<< numTotalKeys << ") than documents in the index (" << _numRecords << ")";
results.errors.push_back(err);
results.valid = false;
}
// Ignore any indexes with a special access method. If an access method name is given, the
// index may be a full text, geo or special index plugin with different semantics.
if (results.valid && !idx->isSparse() && !idx->isPartial() && !idx->isIdIndex() &&
idx->getAccessMethodName() == "" && numTotalKeys < _numRecords) {
hasTooFewKeys = true;
std::string msg = str::stream()
<< "index " << idx->indexName() << " is not sparse or partial, but has fewer entries ("
<< numTotalKeys << ") than documents in the index (" << _numRecords << ")";
if (noErrorOnTooFewKeys) {
results.warnings.push_back(msg);
} else {
results.errors.push_back(msg);
results.valid = false;
}
}
if (!_validateState->isFullIndexValidation() && hasTooFewKeys) {
std::string warning = str::stream()
<< "index " << idx->indexName() << " has fewer keys than records."
<< " Please re-run the validate command with {full: true}";
results.warnings.push_back(warning);
}
}
} // namespace mongo
|