summaryrefslogtreecommitdiff
path: root/src/mongo/db/repl/session_update_tracker.cpp
blob: c539e60867002e16e043b3fc76ac834b1c318f53 (plain)
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
/**
 *    Copyright (C) 2018-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::kReplication

#include "mongo/platform/basic.h"

#include "mongo/db/repl/session_update_tracker.h"

#include "mongo/db/namespace_string.h"
#include "mongo/db/repl/oplog_entry.h"
#include "mongo/db/server_options.h"
#include "mongo/db/session.h"
#include "mongo/db/session_txn_record_gen.h"
#include "mongo/db/transaction_participant_gen.h"
#include "mongo/util/assert_util.h"
#include "mongo/util/log.h"

namespace mongo {
namespace repl {
namespace {

/**
 * Creates an oplog entry to perform an update on the transaction table.
 */
OplogEntry createOplogEntryForTransactionTableUpdate(repl::OpTime opTime,
                                                     const BSONObj& updateBSON,
                                                     const BSONObj& o2Field,
                                                     Date_t wallClockTime) {
    return repl::OplogEntry(opTime,
                            boost::none,  // hash
                            repl::OpTypeEnum::kUpdate,
                            NamespaceString::kSessionTransactionsTableNamespace,
                            boost::none,  // uuid
                            false,        // fromMigrate
                            repl::OplogEntry::kOplogVersion,
                            updateBSON,
                            o2Field,
                            {},    // sessionInfo
                            true,  // upsert
                            wallClockTime,
                            boost::none,   // statementId
                            boost::none,   // prevWriteOpTime
                            boost::none,   // preImangeOpTime
                            boost::none);  // postImageOpTime
}

/**
 * Constructs a new oplog entry if the given entry has transaction state embedded within in. The new
 * oplog entry will contain the operation needed to replicate the transaction table.
 *
 * Returns boost::none if the given oplog doesn't have any transaction state or does not support
 * update to the transaction table.
 */
boost::optional<repl::OplogEntry> createMatchingTransactionTableUpdate(
    const repl::OplogEntry& entry) {
    auto sessionInfo = entry.getOperationSessionInfo();
    if (!sessionInfo.getTxnNumber()) {
        return boost::none;
    }

    invariant(sessionInfo.getSessionId());

    const auto updateBSON = [&] {
        SessionTxnRecord newTxnRecord;
        newTxnRecord.setSessionId(*sessionInfo.getSessionId());
        newTxnRecord.setTxnNum(*sessionInfo.getTxnNumber());
        newTxnRecord.setLastWriteOpTime(entry.getOpTime());
        newTxnRecord.setLastWriteDate(entry.getWallClockTime());

        return newTxnRecord.toBSON();
    }();

    return createOplogEntryForTransactionTableUpdate(
        entry.getOpTime(),
        updateBSON,
        BSON(SessionTxnRecord::kSessionIdFieldName << sessionInfo.getSessionId()->toBSON()),
        entry.getWallClockTime());
}

/**
 * Returns true if the oplog entry represents an operation in a transaction and false otherwise.
 */
bool isTransactionEntry(OplogEntry entry) {
    auto sessionInfo = entry.getOperationSessionInfo();
    if (!sessionInfo.getTxnNumber()) {
        return false;
    }

    return entry.isPartialTransaction() ||
        entry.getCommandType() == repl::OplogEntry::CommandType::kAbortTransaction ||
        entry.getCommandType() == repl::OplogEntry::CommandType::kCommitTransaction ||
        entry.getCommandType() == repl::OplogEntry::CommandType::kApplyOps;
}

}  // namespace

boost::optional<std::vector<OplogEntry>> SessionUpdateTracker::_updateOrFlush(
    const OplogEntry& entry) {
    const auto& ns = entry.getNss();

    if (ns == NamespaceString::kSessionTransactionsTableNamespace ||
        (ns.isConfigDB() && ns.isCommand())) {
        return _flush(entry);
    }

    _updateSessionInfo(entry);
    return boost::none;
}

boost::optional<std::vector<OplogEntry>> SessionUpdateTracker::updateSession(
    const OplogEntry& entry) {
    if (!isTransactionEntry(entry)) {
        return _updateOrFlush(entry);
    }

    // If we generate an update from a multi-statement transaction operation, we must clear (then
    // replace) a possibly queued transaction table update for a retryable write on this session.
    // It is okay to clear the transaction table update because retryable writes only care about
    // the final state of the transaction table entry for a given session, not the full history
    // of updates for the session. By contrast, we care about each transaction table update for
    // multi-statement transactions -- we must maintain the timestamps and transaction states for
    // each entry originating from a multi-statement transaction. For this reason, we cannot defer
    // entries originating from multi-statement transactions.
    if (auto txnTableUpdate = _createTransactionTableUpdateFromTransactionOp(entry)) {
        _sessionsToUpdate.erase(*entry.getOperationSessionInfo().getSessionId());
        return boost::optional<std::vector<OplogEntry>>({*txnTableUpdate});
    }

    return boost::none;
}

void SessionUpdateTracker::_updateSessionInfo(const OplogEntry& entry) {
    const auto& sessionInfo = entry.getOperationSessionInfo();

    if (!sessionInfo.getTxnNumber()) {
        return;
    }

    const auto& lsid = sessionInfo.getSessionId();
    invariant(lsid);

    // Ignore any no-op oplog entries, except for the ones generated from session migration
    // of CRUD ops. These entries will have an o2 field that contains the original CRUD
    // oplog entry.
    if (entry.getOpType() == OpTypeEnum::kNoop) {
        if (!entry.getFromMigrate() || !*entry.getFromMigrate()) {
            return;
        }

        if (!entry.getObject2()) {
            return;
        }
    }

    auto iter = _sessionsToUpdate.find(*lsid);
    if (iter == _sessionsToUpdate.end()) {
        _sessionsToUpdate.emplace(*lsid, entry);
        return;
    }

    const auto& existingSessionInfo = iter->second.getOperationSessionInfo();
    if (*sessionInfo.getTxnNumber() >= *existingSessionInfo.getTxnNumber()) {
        iter->second = entry;
        return;
    }

    severe() << "Entry for session " << lsid->toBSON() << " has txnNumber "
             << *sessionInfo.getTxnNumber() << " < " << *existingSessionInfo.getTxnNumber();
    severe() << "New oplog entry: " << redact(entry.toString());
    severe() << "Existing oplog entry: " << redact(iter->second.toString());

    fassertFailedNoTrace(50843);
}

std::vector<OplogEntry> SessionUpdateTracker::_flush(const OplogEntry& entry) {
    switch (entry.getOpType()) {
        case OpTypeEnum::kInsert:
        case OpTypeEnum::kNoop:
            // Session table is keyed by session id, so nothing to do here because
            // it would have triggered a unique index violation in the primary if
            // it was trying to insert with the same session id with existing ones.
            return {};

        case OpTypeEnum::kUpdate:
            return _flushForQueryPredicate(*entry.getObject2());

        case OpTypeEnum::kDelete:
            return _flushForQueryPredicate(entry.getObject());

        case OpTypeEnum::kCommand:
            return flushAll();
    }

    MONGO_UNREACHABLE;
}

std::vector<OplogEntry> SessionUpdateTracker::flushAll() {
    std::vector<OplogEntry> opList;

    for (auto&& entry : _sessionsToUpdate) {
        auto newUpdate = createMatchingTransactionTableUpdate(entry.second);
        invariant(newUpdate);
        opList.push_back(std::move(*newUpdate));
    }
    _sessionsToUpdate.clear();

    return opList;
}

std::vector<OplogEntry> SessionUpdateTracker::_flushForQueryPredicate(
    const BSONObj& queryPredicate) {
    auto idField = queryPredicate["_id"].Obj();
    auto lsid = LogicalSessionId::parse(IDLParserErrorContext("lsidInOplogQuery"), idField);
    auto iter = _sessionsToUpdate.find(lsid);

    if (iter == _sessionsToUpdate.end()) {
        return {};
    }

    std::vector<OplogEntry> opList;
    auto updateOplog = createMatchingTransactionTableUpdate(iter->second);
    invariant(updateOplog);
    opList.push_back(std::move(*updateOplog));
    _sessionsToUpdate.erase(iter);

    return opList;
}

boost::optional<OplogEntry> SessionUpdateTracker::_createTransactionTableUpdateFromTransactionOp(
    const repl::OplogEntry& entry) {
    auto sessionInfo = entry.getOperationSessionInfo();

    // We only update the transaction table on the first partialTxn operation.
    if (entry.isPartialTransaction() && !entry.getPrevWriteOpTimeInTransaction()->isNull()) {
        return boost::none;
    }
    invariant(sessionInfo.getSessionId());

    const auto updateBSON = [&] {
        SessionTxnRecord newTxnRecord;
        newTxnRecord.setSessionId(*sessionInfo.getSessionId());
        newTxnRecord.setTxnNum(*sessionInfo.getTxnNumber());
        newTxnRecord.setLastWriteOpTime(entry.getOpTime());
        newTxnRecord.setLastWriteDate(entry.getWallClockTime());

        if (entry.isPartialTransaction()) {
            invariant(entry.getPrevWriteOpTimeInTransaction()->isNull());
            newTxnRecord.setState(DurableTxnStateEnum::kInProgress);
            newTxnRecord.setStartOpTime(entry.getOpTime());
            return newTxnRecord.toBSON();
        }
        switch (entry.getCommandType()) {
            case repl::OplogEntry::CommandType::kApplyOps:
                if (entry.shouldPrepare()) {
                    newTxnRecord.setState(DurableTxnStateEnum::kPrepared);
                    if (entry.getPrevWriteOpTimeInTransaction()->isNull()) {
                        // The prepare oplog entry is the first operation of the transaction.
                        newTxnRecord.setStartOpTime(entry.getOpTime());
                    } else {
                        // Update the transaction record using $set to avoid overwriting the
                        // startOpTime.
                        return BSON("$set" << newTxnRecord.toBSON());
                    }
                } else {
                    newTxnRecord.setState(DurableTxnStateEnum::kCommitted);
                }
                break;
            case repl::OplogEntry::CommandType::kCommitTransaction:
                newTxnRecord.setState(DurableTxnStateEnum::kCommitted);
                break;
            case repl::OplogEntry::CommandType::kAbortTransaction:
                newTxnRecord.setState(DurableTxnStateEnum::kAborted);
                break;
            default:
                break;
        }
        return newTxnRecord.toBSON();
    }();

    return createOplogEntryForTransactionTableUpdate(
        entry.getOpTime(),
        updateBSON,
        BSON(SessionTxnRecord::kSessionIdFieldName << sessionInfo.getSessionId()->toBSON()),
        entry.getWallClockTime());
}

}  // namespace repl
}  // namespace mongo