/
resharding_txn_cloner.cpp
230 lines (197 loc) · 9.67 KB
/
resharding_txn_cloner.cpp
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
/**
* Copyright (C) 2020-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_LOGV2_DEFAULT_COMPONENT ::mongo::logv2::LogComponent::kSharding
#include "mongo/platform/basic.h"
#include "mongo/db/s/resharding_txn_cloner.h"
#include <fmt/format.h>
#include "mongo/bson/bsonobj.h"
#include "mongo/client/dbclient_connection.h"
#include "mongo/client/fetcher.h"
#include "mongo/client/read_preference.h"
#include "mongo/client/remote_command_targeter.h"
#include "mongo/db/concurrency/write_conflict_exception.h"
#include "mongo/db/operation_context.h"
#include "mongo/db/pipeline/document_source_match.h"
#include "mongo/db/pipeline/document_source_sort.h"
#include "mongo/db/pipeline/sharded_agg_helpers.h"
#include "mongo/db/session_catalog_mongod.h"
#include "mongo/db/session_txn_record_gen.h"
#include "mongo/db/transaction_participant.h"
#include "mongo/logv2/redaction.h"
#include "mongo/s/shard_id.h"
namespace mongo {
using namespace fmt::literals;
std::unique_ptr<Pipeline, PipelineDeleter> createConfigTxnCloningPipelineForResharding(
const boost::intrusive_ptr<ExpressionContext>& expCtx,
Timestamp fetchTimestamp,
boost::optional<LogicalSessionId> startAfter) {
invariant(!fetchTimestamp.isNull());
std::list<boost::intrusive_ptr<DocumentSource>> stages;
if (startAfter) {
stages.emplace_back(DocumentSourceMatch::create(
BSON(SessionTxnRecord::kSessionIdFieldName << BSON("$gt" << startAfter->toBSON())),
expCtx));
}
stages.emplace_back(
DocumentSourceSort::create(expCtx, BSON(SessionTxnRecord::kSessionIdFieldName << 1)));
stages.emplace_back(DocumentSourceMatch::create(
BSON((SessionTxnRecord::kLastWriteOpTimeFieldName + "." + repl::OpTime::kTimestampFieldName)
<< BSON("$lt" << fetchTimestamp)),
expCtx));
return Pipeline::create(std::move(stages), expCtx);
}
std::unique_ptr<Fetcher> cloneConfigTxnsForResharding(
OperationContext* opCtx,
const ShardId& shardId,
Timestamp fetchTimestamp,
boost::optional<LogicalSessionId> startAfter,
std::function<void(OperationContext*, BSONObj)> merge,
Status* status) {
boost::intrusive_ptr<ExpressionContext> expCtx = make_intrusive<ExpressionContext>(
opCtx, nullptr, NamespaceString::kSessionTransactionsTableNamespace);
auto pipeline =
createConfigTxnCloningPipelineForResharding(expCtx, fetchTimestamp, std::move(startAfter));
AggregationRequest request(NamespaceString::kSessionTransactionsTableNamespace,
pipeline->serializeToBson());
request.setReadConcern(BSON(repl::ReadConcernArgs::kLevelFieldName
<< repl::readConcernLevels::kMajorityName
<< repl::ReadConcernArgs::kAfterClusterTimeFieldName
<< fetchTimestamp));
request.setHint(BSON("_id" << 1));
auto shard = uassertStatusOK(Grid::get(opCtx)->shardRegistry()->getShard(opCtx, shardId));
const auto targetHost = uassertStatusOK(
shard->getTargeter()->findHost(opCtx, ReadPreferenceSetting{ReadPreference::Nearest}));
auto serviceContext = opCtx->getServiceContext();
auto fetcherCallback =
[merge, status, serviceContext](const Fetcher::QueryResponseStatus& dataStatus,
Fetcher::NextAction* nextAction,
BSONObjBuilder* getMoreBob) {
if (!dataStatus.isOK()) {
*status = dataStatus.getStatus();
return;
}
ThreadClient threadClient(serviceContext);
auto uniqueOpCtx = cc().makeOperationContext();
auto fetcherOpCtx = uniqueOpCtx.get();
auto data = dataStatus.getValue();
for (BSONObj doc : data.documents) {
try {
merge(fetcherOpCtx, doc);
} catch (const DBException& ex) {
*status = ex.toStatus();
return;
}
}
if (!getMoreBob) {
return;
}
getMoreBob->append("getMore", data.cursorId);
getMoreBob->append("collection", data.nss.coll());
};
auto executor = Grid::get(opCtx)->getExecutorPool()->getFixedExecutor();
auto fetcher = std::make_unique<Fetcher>(
executor.get(),
targetHost,
"config",
request.serializeToCommandObj().toBson(),
fetcherCallback,
ReadPreferenceSetting(ReadPreference::Nearest).toContainingBSON());
uassertStatusOK(fetcher->schedule());
return fetcher;
}
void configTxnsMergerForResharding(OperationContext* opCtx, BSONObj donorBsonTransaction) {
SessionTxnRecord donorTransaction;
donorTransaction = SessionTxnRecord::parse(
IDLParserErrorContext("resharding config transactions cloning"), donorBsonTransaction);
opCtx->setLogicalSessionId(donorTransaction.getSessionId());
while (true) {
auto ocs = std::make_unique<MongoDOperationContextSession>(opCtx);
auto txnParticipant = TransactionParticipant::get(opCtx);
// Which error code should this be? what message?
uassert(4989900, "Failed to get transaction Participant", txnParticipant);
try {
txnParticipant.beginOrContinue(
opCtx, donorTransaction.getTxnNum(), boost::none, boost::none);
} catch (const DBException& ex) {
if (ex.code() == ErrorCodes::TransactionTooOld) {
// donorTransaction.getTxnNum() < recipientTxnNumber
return;
} else if (ex.code() == ErrorCodes::IncompleteTransactionHistory) {
// donorTransaction.getTxnNum() == recipientTxnNumber &&
// !txnParticipant.transactionIsInRetryableWriteMode()
return;
} else if (ex.code() == ErrorCodes::PreparedTransactionInProgress) {
// txnParticipant.transactionIsPrepared()
ocs.reset();
// TODO SERVER-51493 Change to not block here.
txnParticipant.onExitPrepare().wait();
continue;
}
throw;
}
repl::MutableOplogEntry oplogEntry;
oplogEntry.setObject(BSON("$sessionMigrateInfo" << 1));
oplogEntry.setObject2(TransactionParticipant::kDeadEndSentinel);
oplogEntry.setOpType(repl::OpTypeEnum::kNoop);
oplogEntry.setSessionId(donorTransaction.getSessionId());
oplogEntry.setTxnNumber(donorTransaction.getTxnNum());
oplogEntry.setStatementId(kIncompleteHistoryStmtId);
oplogEntry.setPrevWriteOpTimeInTransaction(repl::OpTime());
oplogEntry.setNss({});
oplogEntry.setWallClockTime(Date_t::now());
writeConflictRetry(
opCtx,
"ReshardingTxnCloner",
NamespaceString::kSessionTransactionsTableNamespace.ns(),
[&] {
// Need to take global lock here so repl::logOp will not unlock it and
// trigger the invariant that disallows unlocking global lock while
// inside a WUOW. Take the transaction table db lock to ensure the same
// lock ordering with normal replicated updates to the table.
Lock::DBLock lk(
opCtx, NamespaceString::kSessionTransactionsTableNamespace.db(), MODE_IX);
WriteUnitOfWork wunit(opCtx);
repl::OpTime opTime = repl::logOp(opCtx, &oplogEntry);
uassert(4989901,
str::stream() << "Failed to create new oplog entry for oplog with opTime: "
<< oplogEntry.getOpTime().toString() << ": "
<< redact(oplogEntry.toBSON()),
!opTime.isNull());
SessionTxnRecord sessionTxnRecord(donorTransaction.getSessionId(),
donorTransaction.getTxnNum(),
opTime,
Date_t::now());
txnParticipant.onRetryableWriteCloningCompleted(
opCtx, {kIncompleteHistoryStmtId}, sessionTxnRecord);
wunit.commit();
});
return;
}
}
} // namespace mongo