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
|
/**
* Copyright (C) 2021-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::kTenantMigration
#include "mongo/platform/basic.h"
#include "mongo/db/client.h"
#include "mongo/db/repl/read_concern_args.h"
#include "mongo/db/repl/replication_coordinator.h"
#include "mongo/db/repl/storage_interface.h"
#include "mongo/db/repl/tenant_migration_access_blocker_executor.h"
#include "mongo/db/repl/tenant_migration_decoration.h"
#include "mongo/db/repl/tenant_migration_recipient_access_blocker.h"
#include "mongo/logv2/log.h"
#include "mongo/util/cancellation.h"
#include "mongo/util/fail_point.h"
#include "mongo/util/future_util.h"
namespace mongo {
namespace {
MONGO_FAIL_POINT_DEFINE(tenantMigrationRecipientNotRejectReads);
}
TenantMigrationRecipientAccessBlocker::TenantMigrationRecipientAccessBlocker(
ServiceContext* serviceContext,
UUID migrationId,
std::string tenantId,
std::string donorConnString)
: TenantMigrationAccessBlocker(BlockerType::kRecipient),
_serviceContext(serviceContext),
_migrationId(migrationId),
_tenantId(std::move(tenantId)),
_donorConnString(std::move(donorConnString)) {
_asyncBlockingOperationsExecutor = TenantMigrationAccessBlockerExecutor::get(serviceContext)
.getOrCreateBlockedOperationsExecutor();
}
Status TenantMigrationRecipientAccessBlocker::checkIfCanWrite() {
// This is guaranteed by the migration protocol. The recipient will not get any writes until the
// migration is committed on the donor.
return Status::OK();
}
Status TenantMigrationRecipientAccessBlocker::waitUntilCommittedOrAborted(OperationContext* opCtx) {
// Recipient nodes will not throw TenantMigrationConflict errors and so we should never need
// to wait for a migration to commit/abort on the recipient set.
MONGO_UNREACHABLE;
}
SharedSemiFuture<void> TenantMigrationRecipientAccessBlocker::getCanReadFuture(
OperationContext* opCtx, StringData command) {
if (MONGO_unlikely(tenantMigrationRecipientNotRejectReads.shouldFail())) {
return SharedSemiFuture<void>();
}
// Exclude internal reads decorated with 'tenantMigrationRecipientInfo' from any logic.
if (repl::tenantMigrationRecipientInfo(opCtx).has_value()) {
LOGV2_DEBUG(5492000,
1,
"Internal tenant read got excluded from the MTAB filtering",
"tenantId"_attr = _tenantId,
"opId"_attr = opCtx->getOpID());
return SharedSemiFuture<void>();
}
auto readConcernArgs = repl::ReadConcernArgs::get(opCtx);
auto atClusterTime = [opCtx, &readConcernArgs]() -> boost::optional<Timestamp> {
if (auto atClusterTime = readConcernArgs.getArgsAtClusterTime()) {
return atClusterTime->asTimestamp();
} else if (readConcernArgs.getLevel() == repl::ReadConcernLevel::kSnapshotReadConcern) {
return repl::StorageInterface::get(opCtx)->getPointInTimeReadTimestamp(opCtx);
}
return boost::none;
}();
stdx::lock_guard<Latch> lk(_mutex);
if (_state.isReject()) {
return SharedSemiFuture<void>(Status(
ErrorCodes::SnapshotTooOld, "Tenant read is not allowed before migration completes"));
}
invariant(_state.isRejectBefore());
invariant(_rejectBeforeTimestamp);
if (atClusterTime && *atClusterTime < *_rejectBeforeTimestamp) {
return SharedSemiFuture<void>(Status(
ErrorCodes::SnapshotTooOld, "Tenant read is not allowed before migration completes"));
}
if (readConcernArgs.getLevel() == repl::ReadConcernLevel::kMajorityReadConcern) {
// Speculative majority reads are only used for change streams (against the oplog
// collection) or when enableMajorityReadConcern=false. So we don't expect speculative
// majority reads in serverless.
invariant(readConcernArgs.getMajorityReadMechanism() !=
repl::ReadConcernArgs::MajorityReadMechanism::kSpeculative);
return ExecutorFuture(_asyncBlockingOperationsExecutor)
.then([timestamp = *_rejectBeforeTimestamp, deadline = opCtx->getDeadline()] {
auto uniqueOpCtx = cc().makeOperationContext();
auto opCtx = uniqueOpCtx.get();
opCtx->setDeadlineByDate(deadline, ErrorCodes::MaxTimeMSExpired);
repl::ReplicationCoordinator::get(opCtx)->waitUntilSnapshotCommitted(opCtx,
timestamp);
})
.share();
}
return SharedSemiFuture<void>();
}
Status TenantMigrationRecipientAccessBlocker::checkIfLinearizableReadWasAllowed(
OperationContext* opCtx) {
// The donor will block all writes at the blockOpTime, and will not signal the proxy to allow
// reading from the recipient until that blockOpTime is majority committed on the recipient.
// This means any writes made on the donor set are available in the majority snapshot of the
// recipient, so linearizable guarantees will hold using the existing linearizable read
// mechanism of doing a no-op write and waiting for it to be majority committed.
return Status::OK();
}
Status TenantMigrationRecipientAccessBlocker::checkIfCanBuildIndex() {
return Status::OK();
}
bool TenantMigrationRecipientAccessBlocker::checkIfShouldBlockTTL() const {
stdx::lock_guard<Latch> lg(_mutex);
return _ttlIsBlocked;
}
void TenantMigrationRecipientAccessBlocker::stopBlockingTTL() {
stdx::lock_guard<Latch> lg(_mutex);
_ttlIsBlocked = false;
}
void TenantMigrationRecipientAccessBlocker::onMajorityCommitPointUpdate(repl::OpTime opTime) {
// Nothing to do.
return;
}
void TenantMigrationRecipientAccessBlocker::appendInfoForServerStatus(
BSONObjBuilder* builder) const {
stdx::lock_guard<Latch> lg(_mutex);
BSONObjBuilder tenantBuilder;
tenantBuilder.append("state", _state.toString());
if (_rejectBeforeTimestamp) {
tenantBuilder.append("rejectBeforeTimestamp", _rejectBeforeTimestamp.get());
}
tenantBuilder.append("ttlIsBlocked", _ttlIsBlocked);
tenantBuilder.append("tenantId", _tenantId);
builder->append("recipient", tenantBuilder.obj());
}
std::string TenantMigrationRecipientAccessBlocker::BlockerState::toString() const {
switch (_state) {
case State::kReject:
return "reject";
case State::kRejectBefore:
return "rejectBefore";
default:
MONGO_UNREACHABLE;
}
}
UUID TenantMigrationRecipientAccessBlocker::getMigrationId() const {
return _migrationId;
}
BSONObj TenantMigrationRecipientAccessBlocker::getDebugInfo() const {
return BSON("tenantId" << _tenantId << "donorConnectionString" << _donorConnString);
}
void TenantMigrationRecipientAccessBlocker::startRejectingReadsBefore(const Timestamp& timestamp) {
stdx::lock_guard<Latch> lk(_mutex);
_state.transitionToRejectBefore();
if (!_rejectBeforeTimestamp || timestamp > *_rejectBeforeTimestamp) {
LOGV2(5358100,
"Tenant migration recipient starting to reject reads before timestamp",
"tenantId"_attr = _tenantId,
"timestamp"_attr = timestamp);
_rejectBeforeTimestamp = timestamp;
}
}
} // namespace mongo
|
