-
Notifications
You must be signed in to change notification settings - Fork 2.3k
/
RaftMachine.java
381 lines (321 loc) · 11.7 KB
/
RaftMachine.java
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
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
/*
* Copyright (c) 2002-2016 "Neo Technology,"
* Network Engine for Objects in Lund AB [http://neotechnology.com]
*
* This file is part of Neo4j.
*
* Neo4j is free software: you can redistribute it and/or modify
* it under the terms of the GNU Affero General Public License as
* published by the Free Software Foundation, either version 3 of the
* License, or (at your option) any later version.
*
* 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
* GNU Affero General Public License for more details.
*
* You should have received a copy of the GNU Affero General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
package org.neo4j.causalclustering.core.consensus;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Collection;
import java.util.Set;
import java.util.concurrent.TimeoutException;
import java.util.function.Predicate;
import org.neo4j.causalclustering.core.consensus.log.RaftLog;
import org.neo4j.causalclustering.core.consensus.log.RaftLogEntry;
import org.neo4j.causalclustering.core.consensus.log.segmented.InFlightMap;
import org.neo4j.causalclustering.core.consensus.membership.RaftMembershipManager;
import org.neo4j.causalclustering.core.consensus.outcome.ConsensusOutcome;
import org.neo4j.causalclustering.core.consensus.outcome.Outcome;
import org.neo4j.causalclustering.core.consensus.roles.Role;
import org.neo4j.causalclustering.core.consensus.schedule.RenewableTimeoutService;
import org.neo4j.causalclustering.core.consensus.shipping.RaftLogShippingManager;
import org.neo4j.causalclustering.core.consensus.state.ExposedRaftState;
import org.neo4j.causalclustering.core.consensus.state.RaftState;
import org.neo4j.causalclustering.core.consensus.term.TermState;
import org.neo4j.causalclustering.core.consensus.vote.VoteState;
import org.neo4j.causalclustering.core.state.snapshot.RaftCoreState;
import org.neo4j.causalclustering.core.state.storage.StateStorage;
import org.neo4j.causalclustering.helper.VolatileFuture;
import org.neo4j.causalclustering.identity.MemberId;
import org.neo4j.causalclustering.messaging.Outbound;
import org.neo4j.kernel.impl.util.Listener;
import org.neo4j.kernel.monitoring.Monitors;
import org.neo4j.logging.Log;
import org.neo4j.logging.LogProvider;
import static java.lang.String.format;
import static org.neo4j.causalclustering.core.consensus.roles.Role.LEADER;
/**
* Implements the Raft Consensus Algorithm.
*
* The algorithm is driven by incoming messages provided to {@link #handle}.
*/
public class RaftMachine implements LeaderLocator, CoreMetaData
{
private final LeaderNotFoundMonitor leaderNotFoundMonitor;
private RenewableTimeoutService.RenewableTimeout heartbeatTimer;
public enum Timeouts implements RenewableTimeoutService.TimeoutName
{
ELECTION, HEARTBEAT
}
private final RaftState state;
private final MemberId myself;
private final RenewableTimeoutService renewableTimeoutService;
private final long heartbeatInterval;
private RenewableTimeoutService.RenewableTimeout electionTimer;
private RaftMembershipManager membershipManager;
private final long electionTimeout;
private final VolatileFuture<MemberId> volatileLeader = new VolatileFuture<>( null );
private final Outbound<MemberId, RaftMessages.RaftMessage> outbound;
private final Log log;
private Role currentRole = Role.FOLLOWER;
private RaftLogShippingManager logShipping;
public RaftMachine( MemberId myself, StateStorage<TermState> termStorage,
StateStorage<VoteState> voteStorage, RaftLog entryLog,
long electionTimeout, long heartbeatInterval,
RenewableTimeoutService renewableTimeoutService,
Outbound<MemberId, RaftMessages.RaftMessage> outbound,
LogProvider logProvider, RaftMembershipManager membershipManager,
RaftLogShippingManager logShipping,
InFlightMap<RaftLogEntry> inFlightMap,
Monitors monitors )
{
this.myself = myself;
this.electionTimeout = electionTimeout;
this.heartbeatInterval = heartbeatInterval;
this.renewableTimeoutService = renewableTimeoutService;
this.outbound = outbound;
this.logShipping = logShipping;
this.log = logProvider.getLog( getClass() );
this.membershipManager = membershipManager;
this.state = new RaftState( myself, termStorage, membershipManager, entryLog, voteStorage, inFlightMap, logProvider );
leaderNotFoundMonitor = monitors.newMonitor( LeaderNotFoundMonitor.class );
initTimers();
}
private void initTimers()
{
electionTimer = renewableTimeoutService.create( Timeouts.ELECTION, electionTimeout, randomTimeoutRange(),
timeout -> {
try
{
handle( new RaftMessages.Timeout.Election( myself ) );
}
catch ( IOException e )
{
log.error( "Failed to process election timeout.", e );
}
timeout.renew();
} );
heartbeatTimer = renewableTimeoutService.create( Timeouts.HEARTBEAT, heartbeatInterval, 0,
timeout -> {
try
{
handle( new RaftMessages.Timeout.Heartbeat( myself ) );
}
catch ( IOException e )
{
log.error( "Failed to process heartbeat timeout.", e );
}
timeout.renew();
} );
}
public void triggerElection() throws IOException
{
handle( new RaftMessages.Timeout.Election( myself ) );
}
public void panic()
{
heartbeatTimer.cancel();
electionTimer.cancel();
}
public synchronized RaftCoreState coreState()
{
return new RaftCoreState( membershipManager.getCommitted() );
}
public synchronized void installCoreState( RaftCoreState coreState ) throws IOException
{
membershipManager.install( coreState.committed() );
}
public synchronized void setTargetMembershipSet( Set<MemberId> targetMembers )
{
membershipManager.setTargetMembershipSet( targetMembers );
if ( currentRole == LEADER )
{
membershipManager.onFollowerStateChange( state.followerStates() );
}
}
@Override
public MemberId getLeader() throws NoLeaderFoundException
{
return waitForLeader( 0, member -> member != null );
}
private MemberId waitForLeader( long timeoutMillis, Predicate<MemberId> predicate ) throws NoLeaderFoundException
{
try
{
return volatileLeader.get( timeoutMillis, predicate );
}
catch ( InterruptedException e )
{
Thread.currentThread().interrupt();
leaderNotFoundMonitor.increment();
throw new NoLeaderFoundException( e );
}
catch ( TimeoutException e )
{
leaderNotFoundMonitor.increment();
throw new NoLeaderFoundException( e );
}
}
private Collection<Listener<MemberId>> leaderListeners = new ArrayList<>();
@Override
public synchronized void registerListener( Listener<MemberId> listener )
{
leaderListeners.add( listener );
listener.receive( state.leader() );
}
@Override
public synchronized void unregisterListener( Listener listener )
{
leaderListeners.remove( listener );
}
/**
* Every call to state() gives you an immutable copy of the current state.
*
* @return A fresh view of the state.
*/
public synchronized ExposedRaftState state()
{
return state.copy();
}
private void notifyLeaderChanges( Outcome outcome )
{
for ( Listener<MemberId> listener : leaderListeners )
{
listener.receive( outcome.getLeader() );
}
}
private void handleLogShipping( Outcome outcome ) throws IOException
{
LeaderContext leaderContext = new LeaderContext( outcome.getTerm(), outcome.getLeaderCommit() );
if ( outcome.isElectedLeader() )
{
logShipping.resume( leaderContext );
}
else if ( outcome.isSteppingDown() )
{
logShipping.pause();
}
if ( outcome.getRole() == LEADER )
{
logShipping.handleCommands( outcome.getShipCommands(), leaderContext );
}
}
private boolean leaderChanged( Outcome outcome, MemberId oldLeader )
{
if ( oldLeader == null && outcome.getLeader() != null )
{
return true;
}
else if ( oldLeader != null && !oldLeader.equals( outcome.getLeader() ) )
{
return true;
}
return false;
}
public synchronized ConsensusOutcome handle( RaftMessages.RaftMessage incomingMessage ) throws IOException
{
Outcome outcome = currentRole.handler.handle( incomingMessage, state, log );
boolean newLeaderWasElected = leaderChanged( outcome, state.leader() );
state.update( outcome ); // updates to raft log happen within
sendMessages( outcome );
handleTimers( outcome );
handleLogShipping( outcome );
driveMembership( outcome );
volatileLeader.set( outcome.getLeader() );
if ( newLeaderWasElected )
{
notifyLeaderChanges( outcome );
}
return outcome;
}
private void driveMembership( Outcome outcome ) throws IOException
{
membershipManager.processLog( outcome.getCommitIndex(), outcome.getLogCommands() );
currentRole = outcome.getRole();
membershipManager.onRole( currentRole );
if ( currentRole == LEADER )
{
membershipManager.onFollowerStateChange( state.followerStates() );
}
}
private void handleTimers( Outcome outcome )
{
if ( outcome.electionTimeoutRenewed() )
{
electionTimer.renew();
}
}
private void sendMessages( Outcome outcome )
{
for ( RaftMessages.Directed outgoingMessage : outcome.getOutgoingMessages() )
{
try
{
outbound.send( outgoingMessage.to(), outgoingMessage.message() );
}
catch ( Exception e )
{
log.warn( format( "Failed to send message %s.", outgoingMessage ), e );
}
}
}
@Override
public boolean isLeader()
{
return currentRole == LEADER;
}
public Role currentRole()
{
return currentRole;
}
public MemberId identity()
{
return myself;
}
public RaftLogShippingManager logShippingManager()
{
return logShipping;
}
@Override
public String toString()
{
return format( "RaftInstance{role=%s, term=%d, currentMembers=%s}", currentRole, term(), votingMembers() );
}
public static class BootstrapException extends Exception
{
BootstrapException( Throwable cause )
{
super( cause );
}
}
public long term()
{
return state.term();
}
private long randomTimeoutRange()
{
return electionTimeout;
}
public Set<MemberId> votingMembers()
{
return membershipManager.votingMembers();
}
public Set<MemberId> replicationMembers()
{
return membershipManager.replicationMembers();
}
}