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
|
/**
* 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::kSharding
#include "mongo/platform/basic.h"
#include "mongo/db/auth/action_set.h"
#include "mongo/db/auth/action_type.h"
#include "mongo/db/auth/authorization_session.h"
#include "mongo/db/auth/privilege.h"
#include "mongo/db/catalog_raii.h"
#include "mongo/db/commands.h"
#include "mongo/db/concurrency/write_conflict_exception.h"
#include "mongo/db/repl/replication_process.h"
#include "mongo/db/s/active_migrations_registry.h"
#include "mongo/db/s/collection_sharding_runtime.h"
#include "mongo/db/s/migration_chunk_cloner_source_legacy.h"
#include "mongo/db/s/migration_source_manager.h"
#include "mongo/db/write_concern.h"
/**
* This file contains commands, which are specific to the legacy chunk cloner source.
*/
namespace mongo {
namespace {
/**
* Shortcut class to perform the appropriate checks and acquire the cloner associated with the
* currently active migration. Uses the currently registered migration for this shard and ensures
* the session ids match.
*/
class AutoGetActiveCloner {
MONGO_DISALLOW_COPYING(AutoGetActiveCloner);
public:
AutoGetActiveCloner(OperationContext* opCtx, const MigrationSessionId& migrationSessionId) {
const auto nss = ActiveMigrationsRegistry::get(opCtx).getActiveDonateChunkNss();
uassert(ErrorCodes::NotYetInitialized, "No active migrations were found", nss);
// Once the collection is locked, the migration status cannot change
_autoColl.emplace(opCtx, *nss, MODE_IS);
uassert(ErrorCodes::NamespaceNotFound,
str::stream() << "Collection " << nss->ns() << " does not exist",
_autoColl->getCollection());
if (auto msm = MigrationSourceManager::get(CollectionShardingRuntime::get(opCtx, *nss))) {
// It is now safe to access the cloner
_chunkCloner = dynamic_cast<MigrationChunkClonerSourceLegacy*>(msm->getCloner());
invariant(_chunkCloner);
} else {
uasserted(ErrorCodes::IllegalOperation,
str::stream() << "No active migrations were found for collection "
<< nss->ns());
}
// Ensure the session ids are correct
uassert(ErrorCodes::IllegalOperation,
str::stream() << "Requested migration session id " << migrationSessionId.toString()
<< " does not match active session id "
<< _chunkCloner->getSessionId().toString(),
migrationSessionId.matches(_chunkCloner->getSessionId()));
}
Database* getDb() const {
invariant(_autoColl);
return _autoColl->getDb();
}
Collection* getColl() const {
invariant(_autoColl);
return _autoColl->getCollection();
}
MigrationChunkClonerSourceLegacy* getCloner() const {
invariant(_chunkCloner);
return _chunkCloner;
}
private:
// Scoped database + collection lock
boost::optional<AutoGetCollection> _autoColl;
// Contains the active cloner for the namespace
MigrationChunkClonerSourceLegacy* _chunkCloner;
};
class InitialCloneCommand : public BasicCommand {
public:
InitialCloneCommand() : BasicCommand("_migrateClone") {}
std::string help() const override {
return "internal";
}
virtual bool supportsWriteConcern(const BSONObj& cmd) const override {
return false;
}
AllowedOnSecondary secondaryAllowed(ServiceContext*) const override {
return AllowedOnSecondary::kNever;
}
virtual bool adminOnly() const {
return true;
}
virtual void addRequiredPrivileges(const std::string& dbname,
const BSONObj& cmdObj,
std::vector<Privilege>* out) const {
ActionSet actions;
actions.addAction(ActionType::internal);
out->push_back(Privilege(ResourcePattern::forClusterResource(), actions));
}
bool run(OperationContext* opCtx,
const std::string&,
const BSONObj& cmdObj,
BSONObjBuilder& result) {
const MigrationSessionId migrationSessionId(
uassertStatusOK(MigrationSessionId::extractFromBSON(cmdObj)));
boost::optional<BSONArrayBuilder> arrBuilder;
// Try to maximize on the size of the buffer, which we are returning in order to have less
// round-trips
int arrSizeAtPrevIteration = -1;
while (!arrBuilder || arrBuilder->arrSize() > arrSizeAtPrevIteration) {
AutoGetActiveCloner autoCloner(opCtx, migrationSessionId);
if (!arrBuilder) {
arrBuilder.emplace(autoCloner.getCloner()->getCloneBatchBufferAllocationSize());
}
arrSizeAtPrevIteration = arrBuilder->arrSize();
uassertStatusOK(autoCloner.getCloner()->nextCloneBatch(
opCtx, autoCloner.getColl(), arrBuilder.get_ptr()));
}
invariant(arrBuilder);
result.appendArray("objects", arrBuilder->arr());
return true;
}
} initialCloneCommand;
class TransferModsCommand : public BasicCommand {
public:
TransferModsCommand() : BasicCommand("_transferMods") {}
std::string help() const override {
return "internal";
}
virtual bool supportsWriteConcern(const BSONObj& cmd) const override {
return false;
}
AllowedOnSecondary secondaryAllowed(ServiceContext*) const override {
return AllowedOnSecondary::kNever;
}
virtual bool adminOnly() const {
return true;
}
virtual void addRequiredPrivileges(const std::string& dbname,
const BSONObj& cmdObj,
std::vector<Privilege>* out) const {
ActionSet actions;
actions.addAction(ActionType::internal);
out->push_back(Privilege(ResourcePattern::forClusterResource(), actions));
}
bool run(OperationContext* opCtx,
const std::string&,
const BSONObj& cmdObj,
BSONObjBuilder& result) {
const MigrationSessionId migrationSessionId(
uassertStatusOK(MigrationSessionId::extractFromBSON(cmdObj)));
AutoGetActiveCloner autoCloner(opCtx, migrationSessionId);
uassertStatusOK(autoCloner.getCloner()->nextModsBatch(opCtx, autoCloner.getDb(), &result));
return true;
}
} transferModsCommand;
/**
* Command for extracting the oplog entries that needs to be migrated for the given migration
* session id.
* Note: this command is not stateless. Calling this command has a side-effect of gradually
* depleting the buffer that contains the oplog entries to be transfered.
*/
class MigrateSessionCommand : public BasicCommand {
public:
MigrateSessionCommand() : BasicCommand("_getNextSessionMods") {}
std::string help() const override {
return "internal";
}
virtual bool supportsWriteConcern(const BSONObj& cmd) const override {
return false;
}
AllowedOnSecondary secondaryAllowed(ServiceContext*) const override {
return AllowedOnSecondary::kNever;
}
virtual bool adminOnly() const {
return true;
}
virtual void addRequiredPrivileges(const std::string& dbname,
const BSONObj& cmdObj,
std::vector<Privilege>* out) const {
ActionSet actions;
actions.addAction(ActionType::internal);
out->push_back(Privilege(ResourcePattern::forClusterResource(), actions));
}
bool run(OperationContext* opCtx,
const std::string&,
const BSONObj& cmdObj,
BSONObjBuilder& result) {
const MigrationSessionId migrationSessionId(
uassertStatusOK(MigrationSessionId::extractFromBSON(cmdObj)));
BSONArrayBuilder arrBuilder;
boost::optional<repl::OpTime> opTime;
writeConflictRetry(opCtx,
"Fetching session related oplogs for migration",
NamespaceString::kRsOplogNamespace.ns(),
[&]() {
AutoGetActiveCloner autoCloner(opCtx, migrationSessionId);
opTime = autoCloner.getCloner()->nextSessionMigrationBatch(
opCtx, &arrBuilder);
});
// If the batch returns something, we wait for write concern to ensure that all the entries
// in the batch have been majority committed. We then need to check that the rollback id
// hasn't changed since we started migration, because a change would indicate that some data
// in this batch may have been rolled back. In this case, we abort the migration.
if (opTime) {
WriteConcernResult wcResult;
WriteConcernOptions majorityWC(
WriteConcernOptions::kMajority, WriteConcernOptions::SyncMode::UNSET, 0);
uassertStatusOK(waitForWriteConcern(opCtx, opTime.get(), majorityWC, &wcResult));
auto rollbackIdAtMigrationInit = [&]() {
AutoGetActiveCloner autoCloner(opCtx, migrationSessionId);
return autoCloner.getCloner()->getRollbackIdAtInit();
}();
// The check for rollback id must be done after having waited for majority in order to
// ensure that whatever was waited on didn't get rolled back.
auto rollbackId = repl::ReplicationProcess::get(opCtx)->getRollbackID();
uassert(50881,
str::stream() << "rollback detected, rollbackId was "
<< rollbackIdAtMigrationInit
<< " but is now "
<< rollbackId,
rollbackId == rollbackIdAtMigrationInit);
}
result.appendArray("oplog", arrBuilder.arr());
return true;
}
} migrateSessionCommand;
} // namespace
} // namespace mongo
|