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
|
/**
* 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::kDefault
#include "mongo/platform/basic.h"
#include "mongo/db/logical_clock.h"
#include "mongo/base/status.h"
#include "mongo/db/global_settings.h"
#include "mongo/db/operation_context.h"
#include "mongo/db/server_parameters.h"
#include "mongo/db/service_context.h"
#include "mongo/db/time_proof_service.h"
#include "mongo/util/log.h"
namespace mongo {
constexpr Seconds LogicalClock::kMaxAcceptableLogicalClockDriftSecs;
MONGO_EXPORT_STARTUP_SERVER_PARAMETER(maxAcceptableLogicalClockDriftSecs,
long long,
LogicalClock::kMaxAcceptableLogicalClockDriftSecs.count())
->withValidator([](const long long& potentialNewValue) {
if (potentialNewValue <= 0) {
return Status(ErrorCodes::BadValue,
str::stream() << "maxAcceptableLogicalClockDriftSecs must be positive, "
"but attempted to set to: "
<< potentialNewValue);
}
return Status::OK();
});
namespace {
const auto getLogicalClock = ServiceContext::declareDecoration<std::unique_ptr<LogicalClock>>();
bool lessThanOrEqualToMaxPossibleTime(LogicalTime time, uint64_t nTicks) {
return time.asTimestamp().getSecs() <= LogicalClock::kMaxSignedInt &&
time.asTimestamp().getInc() <= (LogicalClock::kMaxSignedInt - nTicks);
}
}
LogicalTime LogicalClock::getClusterTimeForReplicaSet(OperationContext* opCtx) {
if (getGlobalReplSettings().usingReplSets()) {
return get(opCtx)->getClusterTime();
}
return {};
}
LogicalClock* LogicalClock::get(ServiceContext* service) {
return getLogicalClock(service).get();
}
LogicalClock* LogicalClock::get(OperationContext* ctx) {
return get(ctx->getClient()->getServiceContext());
}
void LogicalClock::set(ServiceContext* service, std::unique_ptr<LogicalClock> clockArg) {
auto& clock = getLogicalClock(service);
clock = std::move(clockArg);
}
LogicalClock::LogicalClock(ServiceContext* service) : _service(service) {}
LogicalTime LogicalClock::getClusterTime() {
stdx::lock_guard<stdx::mutex> lock(_mutex);
return _clusterTime;
}
Status LogicalClock::advanceClusterTime(const LogicalTime newTime) {
stdx::lock_guard<stdx::mutex> lock(_mutex);
auto rateLimitStatus = _passesRateLimiter_inlock(newTime);
if (!rateLimitStatus.isOK()) {
return rateLimitStatus;
}
if (newTime > _clusterTime) {
_clusterTime = newTime;
}
return Status::OK();
}
LogicalTime LogicalClock::reserveTicks(uint64_t nTicks) {
invariant(nTicks > 0 && nTicks <= kMaxSignedInt);
stdx::lock_guard<stdx::mutex> lock(_mutex);
LogicalTime clusterTime = _clusterTime;
const unsigned wallClockSecs =
durationCount<Seconds>(_service->getFastClockSource()->now().toDurationSinceEpoch());
unsigned clusterTimeSecs = clusterTime.asTimestamp().getSecs();
// Synchronize clusterTime with wall clock time, if clusterTime was behind in seconds.
if (clusterTimeSecs < wallClockSecs) {
clusterTime = LogicalTime(Timestamp(wallClockSecs, 0));
}
// If reserving 'nTicks' would force the cluster timestamp's increment field to exceed (2^31-1),
// overflow by moving to the next second. We use the signed integer maximum as an overflow point
// in order to preserve compatibility with potentially signed or unsigned integral Timestamp
// increment types. It is also unlikely to apply more than 2^31 oplog entries in the span of one
// second.
else if (clusterTime.asTimestamp().getInc() > (kMaxSignedInt - nTicks)) {
log() << "Exceeded maximum allowable increment value within one second. Moving clusterTime "
"forward to the next second.";
// Move time forward to the next second
clusterTime = LogicalTime(Timestamp(clusterTime.asTimestamp().getSecs() + 1, 0));
}
uassert(40482,
"cluster time cannot be advanced beyond its maximum value",
lessThanOrEqualToMaxPossibleTime(clusterTime, nTicks));
// Save the next cluster time.
clusterTime.addTicks(1);
_clusterTime = clusterTime;
// Add the rest of the requested ticks if needed.
if (nTicks > 1) {
_clusterTime.addTicks(nTicks - 1);
}
return clusterTime;
}
void LogicalClock::setClusterTimeFromTrustedSource(LogicalTime newTime) {
stdx::lock_guard<stdx::mutex> lock(_mutex);
// Rate limit checks are skipped here so a server with no activity for longer than
// maxAcceptableLogicalClockDriftSecs seconds can still have its cluster time initialized.
uassert(40483,
"cluster time cannot be advanced beyond its maximum value",
lessThanOrEqualToMaxPossibleTime(newTime, 0));
if (newTime > _clusterTime) {
_clusterTime = newTime;
}
}
Status LogicalClock::_passesRateLimiter_inlock(LogicalTime newTime) {
const unsigned wallClockSecs =
durationCount<Seconds>(_service->getFastClockSource()->now().toDurationSinceEpoch());
auto maxAcceptableDriftSecs = static_cast<const unsigned>(maxAcceptableLogicalClockDriftSecs);
auto newTimeSecs = newTime.asTimestamp().getSecs();
// Both values are unsigned, so compare them first to avoid wrap-around.
if ((newTimeSecs > wallClockSecs) && (newTimeSecs - wallClockSecs) > maxAcceptableDriftSecs) {
return Status(ErrorCodes::ClusterTimeFailsRateLimiter,
str::stream() << "New cluster time, " << newTimeSecs
<< ", is too far from this node's wall clock time, "
<< wallClockSecs
<< ".");
}
uassert(40484,
"cluster time cannot be advanced beyond its maximum value",
lessThanOrEqualToMaxPossibleTime(newTime, 0));
return Status::OK();
}
bool LogicalClock::isEnabled() const {
stdx::lock_guard<stdx::mutex> lock(_mutex);
return _isEnabled;
}
void LogicalClock::disable() {
stdx::lock_guard<stdx::mutex> lock(_mutex);
_isEnabled = false;
}
} // namespace mongo
|
