forked from tendermint/tendermint
-
Notifications
You must be signed in to change notification settings - Fork 0
/
state.go
1749 lines (1511 loc) · 60.4 KB
/
state.go
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
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
package consensus
import (
"bytes"
"errors"
"fmt"
"reflect"
"runtime/debug"
"sync"
"time"
cmn "github.com/tendermint/tendermint/libs/common"
"github.com/tendermint/tendermint/libs/fail"
"github.com/tendermint/tendermint/libs/log"
tmtime "github.com/tendermint/tendermint/types/time"
cfg "github.com/tendermint/tendermint/config"
cstypes "github.com/tendermint/tendermint/consensus/types"
tmevents "github.com/tendermint/tendermint/libs/events"
"github.com/tendermint/tendermint/p2p"
sm "github.com/tendermint/tendermint/state"
"github.com/tendermint/tendermint/types"
)
//-----------------------------------------------------------------------------
// Config
const (
proposalHeartbeatIntervalSeconds = 2
)
//-----------------------------------------------------------------------------
// Errors
var (
ErrInvalidProposalSignature = errors.New("Error invalid proposal signature")
ErrInvalidProposalPOLRound = errors.New("Error invalid proposal POL round")
ErrAddingVote = errors.New("Error adding vote")
ErrVoteHeightMismatch = errors.New("Error vote height mismatch")
)
//-----------------------------------------------------------------------------
var (
msgQueueSize = 1000
)
// msgs from the reactor which may update the state
type msgInfo struct {
Msg ConsensusMessage `json:"msg"`
PeerID p2p.ID `json:"peer_key"`
}
// internally generated messages which may update the state
type timeoutInfo struct {
Duration time.Duration `json:"duration"`
Height int64 `json:"height"`
Round int `json:"round"`
Step cstypes.RoundStepType `json:"step"`
}
func (ti *timeoutInfo) String() string {
return fmt.Sprintf("%v ; %d/%d %v", ti.Duration, ti.Height, ti.Round, ti.Step)
}
// ConsensusState handles execution of the consensus algorithm.
// It processes votes and proposals, and upon reaching agreement,
// commits blocks to the chain and executes them against the application.
// The internal state machine receives input from peers, the internal validator, and from a timer.
type ConsensusState struct {
cmn.BaseService
// config details
config *cfg.ConsensusConfig
privValidator types.PrivValidator // for signing votes
// services for creating and executing blocks
blockExec *sm.BlockExecutor
blockStore sm.BlockStore
mempool sm.Mempool
evpool sm.EvidencePool
// internal state
mtx sync.RWMutex
cstypes.RoundState
triggeredTimeoutPrecommit bool
state sm.State // State until height-1.
// state changes may be triggered by: msgs from peers,
// msgs from ourself, or by timeouts
peerMsgQueue chan msgInfo
internalMsgQueue chan msgInfo
timeoutTicker TimeoutTicker
// information about about added votes and block parts are written on this channel
// so statistics can be computed by reactor
statsMsgQueue chan msgInfo
// we use eventBus to trigger msg broadcasts in the reactor,
// and to notify external subscribers, eg. through a websocket
eventBus *types.EventBus
// a Write-Ahead Log ensures we can recover from any kind of crash
// and helps us avoid signing conflicting votes
wal WAL
replayMode bool // so we don't log signing errors during replay
doWALCatchup bool // determines if we even try to do the catchup
// for tests where we want to limit the number of transitions the state makes
nSteps int
// some functions can be overwritten for testing
decideProposal func(height int64, round int)
doPrevote func(height int64, round int)
setProposal func(proposal *types.Proposal) error
// closed when we finish shutting down
done chan struct{}
// synchronous pubsub between consensus state and reactor.
// state only emits EventNewRoundStep, EventVote and EventProposalHeartbeat
evsw tmevents.EventSwitch
// for reporting metrics
metrics *Metrics
}
// StateOption sets an optional parameter on the ConsensusState.
type StateOption func(*ConsensusState)
// NewConsensusState returns a new ConsensusState.
func NewConsensusState(
config *cfg.ConsensusConfig,
state sm.State,
blockExec *sm.BlockExecutor,
blockStore sm.BlockStore,
mempool sm.Mempool,
evpool sm.EvidencePool,
options ...StateOption,
) *ConsensusState {
cs := &ConsensusState{
config: config,
blockExec: blockExec,
blockStore: blockStore,
mempool: mempool,
peerMsgQueue: make(chan msgInfo, msgQueueSize),
internalMsgQueue: make(chan msgInfo, msgQueueSize),
timeoutTicker: NewTimeoutTicker(),
statsMsgQueue: make(chan msgInfo, msgQueueSize),
done: make(chan struct{}),
doWALCatchup: true,
wal: nilWAL{},
evpool: evpool,
evsw: tmevents.NewEventSwitch(),
metrics: NopMetrics(),
}
// set function defaults (may be overwritten before calling Start)
cs.decideProposal = cs.defaultDecideProposal
cs.doPrevote = cs.defaultDoPrevote
cs.setProposal = cs.defaultSetProposal
cs.updateToState(state)
// Don't call scheduleRound0 yet.
// We do that upon Start().
cs.reconstructLastCommit(state)
cs.BaseService = *cmn.NewBaseService(nil, "ConsensusState", cs)
for _, option := range options {
option(cs)
}
return cs
}
//----------------------------------------
// Public interface
// SetLogger implements Service.
func (cs *ConsensusState) SetLogger(l log.Logger) {
cs.BaseService.Logger = l
cs.timeoutTicker.SetLogger(l)
}
// SetEventBus sets event bus.
func (cs *ConsensusState) SetEventBus(b *types.EventBus) {
cs.eventBus = b
cs.blockExec.SetEventBus(b)
}
// StateMetrics sets the metrics.
func StateMetrics(metrics *Metrics) StateOption {
return func(cs *ConsensusState) { cs.metrics = metrics }
}
// String returns a string.
func (cs *ConsensusState) String() string {
// better not to access shared variables
return fmt.Sprintf("ConsensusState") //(H:%v R:%v S:%v", cs.Height, cs.Round, cs.Step)
}
// GetState returns a copy of the chain state.
func (cs *ConsensusState) GetState() sm.State {
cs.mtx.RLock()
defer cs.mtx.RUnlock()
return cs.state.Copy()
}
// GetLastHeight returns the last height committed.
// If there were no blocks, returns 0.
func (cs *ConsensusState) GetLastHeight() int64 {
cs.mtx.Lock()
defer cs.mtx.Unlock()
return cs.RoundState.Height - 1
}
// GetRoundState returns a shallow copy of the internal consensus state.
func (cs *ConsensusState) GetRoundState() *cstypes.RoundState {
cs.mtx.RLock()
defer cs.mtx.RUnlock()
rs := cs.RoundState // copy
return &rs
}
// GetRoundStateJSON returns a json of RoundState, marshalled using go-amino.
func (cs *ConsensusState) GetRoundStateJSON() ([]byte, error) {
cs.mtx.RLock()
defer cs.mtx.RUnlock()
return cdc.MarshalJSON(cs.RoundState)
}
// GetRoundStateSimpleJSON returns a json of RoundStateSimple, marshalled using go-amino.
func (cs *ConsensusState) GetRoundStateSimpleJSON() ([]byte, error) {
cs.mtx.RLock()
defer cs.mtx.RUnlock()
return cdc.MarshalJSON(cs.RoundState.RoundStateSimple())
}
// GetValidators returns a copy of the current validators.
func (cs *ConsensusState) GetValidators() (int64, []*types.Validator) {
cs.mtx.RLock()
defer cs.mtx.RUnlock()
return cs.state.LastBlockHeight, cs.state.Validators.Copy().Validators
}
// SetPrivValidator sets the private validator account for signing votes.
func (cs *ConsensusState) SetPrivValidator(priv types.PrivValidator) {
cs.mtx.Lock()
defer cs.mtx.Unlock()
cs.privValidator = priv
}
// SetTimeoutTicker sets the local timer. It may be useful to overwrite for testing.
func (cs *ConsensusState) SetTimeoutTicker(timeoutTicker TimeoutTicker) {
cs.mtx.Lock()
defer cs.mtx.Unlock()
cs.timeoutTicker = timeoutTicker
}
// LoadCommit loads the commit for a given height.
func (cs *ConsensusState) LoadCommit(height int64) *types.Commit {
cs.mtx.RLock()
defer cs.mtx.RUnlock()
if height == cs.blockStore.Height() {
return cs.blockStore.LoadSeenCommit(height)
}
return cs.blockStore.LoadBlockCommit(height)
}
// OnStart implements cmn.Service.
// It loads the latest state via the WAL, and starts the timeout and receive routines.
func (cs *ConsensusState) OnStart() error {
if err := cs.evsw.Start(); err != nil {
return err
}
// we may set the WAL in testing before calling Start,
// so only OpenWAL if its still the nilWAL
if _, ok := cs.wal.(nilWAL); ok {
walFile := cs.config.WalFile()
wal, err := cs.OpenWAL(walFile)
if err != nil {
cs.Logger.Error("Error loading ConsensusState wal", "err", err.Error())
return err
}
cs.wal = wal
}
// we need the timeoutRoutine for replay so
// we don't block on the tick chan.
// NOTE: we will get a build up of garbage go routines
// firing on the tockChan until the receiveRoutine is started
// to deal with them (by that point, at most one will be valid)
if err := cs.timeoutTicker.Start(); err != nil {
return err
}
// we may have lost some votes if the process crashed
// reload from consensus log to catchup
if cs.doWALCatchup {
if err := cs.catchupReplay(cs.Height); err != nil {
cs.Logger.Error("Error on catchup replay. Proceeding to start ConsensusState anyway", "err", err.Error())
// NOTE: if we ever do return an error here,
// make sure to stop the timeoutTicker
}
}
// now start the receiveRoutine
go cs.receiveRoutine(0)
// schedule the first round!
// use GetRoundState so we don't race the receiveRoutine for access
cs.scheduleRound0(cs.GetRoundState())
return nil
}
// timeoutRoutine: receive requests for timeouts on tickChan and fire timeouts on tockChan
// receiveRoutine: serializes processing of proposoals, block parts, votes; coordinates state transitions
func (cs *ConsensusState) startRoutines(maxSteps int) {
err := cs.timeoutTicker.Start()
if err != nil {
cs.Logger.Error("Error starting timeout ticker", "err", err)
return
}
go cs.receiveRoutine(maxSteps)
}
// OnStop implements cmn.Service. It stops all routines and waits for the WAL to finish.
func (cs *ConsensusState) OnStop() {
cs.evsw.Stop()
cs.timeoutTicker.Stop()
}
// Wait waits for the the main routine to return.
// NOTE: be sure to Stop() the event switch and drain
// any event channels or this may deadlock
func (cs *ConsensusState) Wait() {
<-cs.done
}
// OpenWAL opens a file to log all consensus messages and timeouts for deterministic accountability
func (cs *ConsensusState) OpenWAL(walFile string) (WAL, error) {
wal, err := NewWAL(walFile)
if err != nil {
cs.Logger.Error("Failed to open WAL for consensus state", "wal", walFile, "err", err)
return nil, err
}
wal.SetLogger(cs.Logger.With("wal", walFile))
if err := wal.Start(); err != nil {
return nil, err
}
return wal, nil
}
//------------------------------------------------------------
// Public interface for passing messages into the consensus state, possibly causing a state transition.
// If peerID == "", the msg is considered internal.
// Messages are added to the appropriate queue (peer or internal).
// If the queue is full, the function may block.
// TODO: should these return anything or let callers just use events?
// AddVote inputs a vote.
func (cs *ConsensusState) AddVote(vote *types.Vote, peerID p2p.ID) (added bool, err error) {
if peerID == "" {
cs.internalMsgQueue <- msgInfo{&VoteMessage{vote}, ""}
} else {
cs.peerMsgQueue <- msgInfo{&VoteMessage{vote}, peerID}
}
// TODO: wait for event?!
return false, nil
}
// SetProposal inputs a proposal.
func (cs *ConsensusState) SetProposal(proposal *types.Proposal, peerID p2p.ID) error {
if peerID == "" {
cs.internalMsgQueue <- msgInfo{&ProposalMessage{proposal}, ""}
} else {
cs.peerMsgQueue <- msgInfo{&ProposalMessage{proposal}, peerID}
}
// TODO: wait for event?!
return nil
}
// AddProposalBlockPart inputs a part of the proposal block.
func (cs *ConsensusState) AddProposalBlockPart(height int64, round int, part *types.Part, peerID p2p.ID) error {
if peerID == "" {
cs.internalMsgQueue <- msgInfo{&BlockPartMessage{height, round, part}, ""}
} else {
cs.peerMsgQueue <- msgInfo{&BlockPartMessage{height, round, part}, peerID}
}
// TODO: wait for event?!
return nil
}
// SetProposalAndBlock inputs the proposal and all block parts.
func (cs *ConsensusState) SetProposalAndBlock(proposal *types.Proposal, block *types.Block, parts *types.PartSet, peerID p2p.ID) error {
if err := cs.SetProposal(proposal, peerID); err != nil {
return err
}
for i := 0; i < parts.Total(); i++ {
part := parts.GetPart(i)
if err := cs.AddProposalBlockPart(proposal.Height, proposal.Round, part, peerID); err != nil {
return err
}
}
return nil
}
//------------------------------------------------------------
// internal functions for managing the state
func (cs *ConsensusState) updateHeight(height int64) {
cs.metrics.Height.Set(float64(height))
cs.Height = height
}
func (cs *ConsensusState) updateRoundStep(round int, step cstypes.RoundStepType) {
cs.Round = round
cs.Step = step
}
// enterNewRound(height, 0) at cs.StartTime.
func (cs *ConsensusState) scheduleRound0(rs *cstypes.RoundState) {
//cs.Logger.Info("scheduleRound0", "now", tmtime.Now(), "startTime", cs.StartTime)
sleepDuration := rs.StartTime.Sub(tmtime.Now()) // nolint: gotype, gosimple
cs.scheduleTimeout(sleepDuration, rs.Height, 0, cstypes.RoundStepNewHeight)
}
// Attempt to schedule a timeout (by sending timeoutInfo on the tickChan)
func (cs *ConsensusState) scheduleTimeout(duration time.Duration, height int64, round int, step cstypes.RoundStepType) {
cs.timeoutTicker.ScheduleTimeout(timeoutInfo{duration, height, round, step})
}
// send a msg into the receiveRoutine regarding our own proposal, block part, or vote
func (cs *ConsensusState) sendInternalMessage(mi msgInfo) {
select {
case cs.internalMsgQueue <- mi:
default:
// NOTE: using the go-routine means our votes can
// be processed out of order.
// TODO: use CList here for strict determinism and
// attempt push to internalMsgQueue in receiveRoutine
cs.Logger.Info("Internal msg queue is full. Using a go-routine")
go func() { cs.internalMsgQueue <- mi }()
}
}
// Reconstruct LastCommit from SeenCommit, which we saved along with the block,
// (which happens even before saving the state)
func (cs *ConsensusState) reconstructLastCommit(state sm.State) {
if state.LastBlockHeight == 0 {
return
}
seenCommit := cs.blockStore.LoadSeenCommit(state.LastBlockHeight)
lastPrecommits := types.NewVoteSet(state.ChainID, state.LastBlockHeight, seenCommit.Round(), types.PrecommitType, state.LastValidators)
for _, precommit := range seenCommit.Precommits {
if precommit == nil {
continue
}
added, err := lastPrecommits.AddVote(precommit)
if !added || err != nil {
cmn.PanicCrisis(fmt.Sprintf("Failed to reconstruct LastCommit: %v", err))
}
}
if !lastPrecommits.HasTwoThirdsMajority() {
cmn.PanicSanity("Failed to reconstruct LastCommit: Does not have +2/3 maj")
}
cs.LastCommit = lastPrecommits
}
// Updates ConsensusState and increments height to match that of state.
// The round becomes 0 and cs.Step becomes cstypes.RoundStepNewHeight.
func (cs *ConsensusState) updateToState(state sm.State) {
if cs.CommitRound > -1 && 0 < cs.Height && cs.Height != state.LastBlockHeight {
cmn.PanicSanity(fmt.Sprintf("updateToState() expected state height of %v but found %v",
cs.Height, state.LastBlockHeight))
}
if !cs.state.IsEmpty() && cs.state.LastBlockHeight+1 != cs.Height {
// This might happen when someone else is mutating cs.state.
// Someone forgot to pass in state.Copy() somewhere?!
cmn.PanicSanity(fmt.Sprintf("Inconsistent cs.state.LastBlockHeight+1 %v vs cs.Height %v",
cs.state.LastBlockHeight+1, cs.Height))
}
// If state isn't further out than cs.state, just ignore.
// This happens when SwitchToConsensus() is called in the reactor.
// We don't want to reset e.g. the Votes, but we still want to
// signal the new round step, because other services (eg. mempool)
// depend on having an up-to-date peer state!
if !cs.state.IsEmpty() && (state.LastBlockHeight <= cs.state.LastBlockHeight) {
cs.Logger.Info("Ignoring updateToState()", "newHeight", state.LastBlockHeight+1, "oldHeight", cs.state.LastBlockHeight+1)
cs.newStep()
return
}
// Reset fields based on state.
validators := state.Validators
lastPrecommits := (*types.VoteSet)(nil)
if cs.CommitRound > -1 && cs.Votes != nil {
if !cs.Votes.Precommits(cs.CommitRound).HasTwoThirdsMajority() {
cmn.PanicSanity("updateToState(state) called but last Precommit round didn't have +2/3")
}
lastPrecommits = cs.Votes.Precommits(cs.CommitRound)
}
// Next desired block height
height := state.LastBlockHeight + 1
// RoundState fields
cs.updateHeight(height)
cs.updateRoundStep(0, cstypes.RoundStepNewHeight)
if cs.CommitTime.IsZero() {
// "Now" makes it easier to sync up dev nodes.
// We add timeoutCommit to allow transactions
// to be gathered for the first block.
// And alternative solution that relies on clocks:
// cs.StartTime = state.LastBlockTime.Add(timeoutCommit)
cs.StartTime = cs.config.Commit(tmtime.Now())
} else {
cs.StartTime = cs.config.Commit(cs.CommitTime)
}
cs.Validators = validators
cs.Proposal = nil
cs.ProposalBlock = nil
cs.ProposalBlockParts = nil
cs.LockedRound = -1
cs.LockedBlock = nil
cs.LockedBlockParts = nil
cs.ValidRound = -1
cs.ValidBlock = nil
cs.ValidBlockParts = nil
cs.Votes = cstypes.NewHeightVoteSet(state.ChainID, height, validators)
cs.CommitRound = -1
cs.LastCommit = lastPrecommits
cs.LastValidators = state.LastValidators
cs.state = state
// Finally, broadcast RoundState
cs.newStep()
}
func (cs *ConsensusState) newStep() {
rs := cs.RoundStateEvent()
cs.wal.Write(rs)
cs.nSteps++
// newStep is called by updateToState in NewConsensusState before the eventBus is set!
if cs.eventBus != nil {
cs.eventBus.PublishEventNewRoundStep(rs)
cs.evsw.FireEvent(types.EventNewRoundStep, &cs.RoundState)
}
}
//-----------------------------------------
// the main go routines
// receiveRoutine handles messages which may cause state transitions.
// it's argument (n) is the number of messages to process before exiting - use 0 to run forever
// It keeps the RoundState and is the only thing that updates it.
// Updates (state transitions) happen on timeouts, complete proposals, and 2/3 majorities.
// ConsensusState must be locked before any internal state is updated.
func (cs *ConsensusState) receiveRoutine(maxSteps int) {
onExit := func(cs *ConsensusState) {
// NOTE: the internalMsgQueue may have signed messages from our
// priv_val that haven't hit the WAL, but its ok because
// priv_val tracks LastSig
// close wal now that we're done writing to it
cs.wal.Stop()
cs.wal.Wait()
close(cs.done)
}
defer func() {
if r := recover(); r != nil {
cs.Logger.Error("CONSENSUS FAILURE!!!", "err", r, "stack", string(debug.Stack()))
// stop gracefully
//
// NOTE: We most probably shouldn't be running any further when there is
// some unexpected panic. Some unknown error happened, and so we don't
// know if that will result in the validator signing an invalid thing. It
// might be worthwhile to explore a mechanism for manual resuming via
// some console or secure RPC system, but for now, halting the chain upon
// unexpected consensus bugs sounds like the better option.
onExit(cs)
}
}()
for {
if maxSteps > 0 {
if cs.nSteps >= maxSteps {
cs.Logger.Info("reached max steps. exiting receive routine")
cs.nSteps = 0
return
}
}
rs := cs.RoundState
var mi msgInfo
select {
case <-cs.mempool.TxsAvailable():
cs.handleTxsAvailable()
case mi = <-cs.peerMsgQueue:
cs.wal.Write(mi)
// handles proposals, block parts, votes
// may generate internal events (votes, complete proposals, 2/3 majorities)
cs.handleMsg(mi)
case mi = <-cs.internalMsgQueue:
cs.wal.WriteSync(mi) // NOTE: fsync
// handles proposals, block parts, votes
cs.handleMsg(mi)
case ti := <-cs.timeoutTicker.Chan(): // tockChan:
cs.wal.Write(ti)
// if the timeout is relevant to the rs
// go to the next step
cs.handleTimeout(ti, rs)
case <-cs.Quit():
onExit(cs)
return
}
}
}
// state transitions on complete-proposal, 2/3-any, 2/3-one
func (cs *ConsensusState) handleMsg(mi msgInfo) {
cs.mtx.Lock()
defer cs.mtx.Unlock()
var err error
msg, peerID := mi.Msg, mi.PeerID
switch msg := msg.(type) {
case *ProposalMessage:
// will not cause transition.
// once proposal is set, we can receive block parts
err = cs.setProposal(msg.Proposal)
case *BlockPartMessage:
// if the proposal is complete, we'll enterPrevote or tryFinalizeCommit
added, err := cs.addProposalBlockPart(msg, peerID)
if added {
cs.statsMsgQueue <- mi
}
if err != nil && msg.Round != cs.Round {
cs.Logger.Debug("Received block part from wrong round", "height", cs.Height, "csRound", cs.Round, "blockRound", msg.Round)
err = nil
}
case *VoteMessage:
// attempt to add the vote and dupeout the validator if its a duplicate signature
// if the vote gives us a 2/3-any or 2/3-one, we transition
added, err := cs.tryAddVote(msg.Vote, peerID)
if added {
cs.statsMsgQueue <- mi
}
if err == ErrAddingVote {
// TODO: punish peer
// We probably don't want to stop the peer here. The vote does not
// necessarily comes from a malicious peer but can be just broadcasted by
// a typical peer.
// https://github.com/tendermint/tendermint/issues/1281
}
// NOTE: the vote is broadcast to peers by the reactor listening
// for vote events
// TODO: If rs.Height == vote.Height && rs.Round < vote.Round,
// the peer is sending us CatchupCommit precommits.
// We could make note of this and help filter in broadcastHasVoteMessage().
default:
cs.Logger.Error("Unknown msg type", reflect.TypeOf(msg))
}
if err != nil {
cs.Logger.Error("Error with msg", "height", cs.Height, "round", cs.Round, "type", reflect.TypeOf(msg), "peer", peerID, "err", err, "msg", msg)
}
}
func (cs *ConsensusState) handleTimeout(ti timeoutInfo, rs cstypes.RoundState) {
cs.Logger.Debug("Received tock", "timeout", ti.Duration, "height", ti.Height, "round", ti.Round, "step", ti.Step)
// timeouts must be for current height, round, step
if ti.Height != rs.Height || ti.Round < rs.Round || (ti.Round == rs.Round && ti.Step < rs.Step) {
cs.Logger.Debug("Ignoring tock because we're ahead", "height", rs.Height, "round", rs.Round, "step", rs.Step)
return
}
// the timeout will now cause a state transition
cs.mtx.Lock()
defer cs.mtx.Unlock()
switch ti.Step {
case cstypes.RoundStepNewHeight:
// NewRound event fired from enterNewRound.
// XXX: should we fire timeout here (for timeout commit)?
cs.enterNewRound(ti.Height, 0)
case cstypes.RoundStepNewRound:
cs.enterPropose(ti.Height, 0)
case cstypes.RoundStepPropose:
cs.eventBus.PublishEventTimeoutPropose(cs.RoundStateEvent())
cs.enterPrevote(ti.Height, ti.Round)
case cstypes.RoundStepPrevoteWait:
cs.eventBus.PublishEventTimeoutWait(cs.RoundStateEvent())
cs.enterPrecommit(ti.Height, ti.Round)
case cstypes.RoundStepPrecommitWait:
cs.eventBus.PublishEventTimeoutWait(cs.RoundStateEvent())
cs.enterPrecommit(ti.Height, ti.Round)
cs.enterNewRound(ti.Height, ti.Round+1)
default:
panic(fmt.Sprintf("Invalid timeout step: %v", ti.Step))
}
}
func (cs *ConsensusState) handleTxsAvailable() {
cs.mtx.Lock()
defer cs.mtx.Unlock()
// we only need to do this for round 0
cs.enterPropose(cs.Height, 0)
}
//-----------------------------------------------------------------------------
// State functions
// Used internally by handleTimeout and handleMsg to make state transitions
// Enter: `timeoutNewHeight` by startTime (commitTime+timeoutCommit),
// or, if SkipTimeout==true, after receiving all precommits from (height,round-1)
// Enter: `timeoutPrecommits` after any +2/3 precommits from (height,round-1)
// Enter: +2/3 precommits for nil at (height,round-1)
// Enter: +2/3 prevotes any or +2/3 precommits for block or any from (height, round)
// NOTE: cs.StartTime was already set for height.
func (cs *ConsensusState) enterNewRound(height int64, round int) {
logger := cs.Logger.With("height", height, "round", round)
if cs.Height != height || round < cs.Round || (cs.Round == round && cs.Step != cstypes.RoundStepNewHeight) {
logger.Debug(fmt.Sprintf("enterNewRound(%v/%v): Invalid args. Current step: %v/%v/%v", height, round, cs.Height, cs.Round, cs.Step))
return
}
if now := tmtime.Now(); cs.StartTime.After(now) {
logger.Info("Need to set a buffer and log message here for sanity.", "startTime", cs.StartTime, "now", now)
}
logger.Info(fmt.Sprintf("enterNewRound(%v/%v). Current: %v/%v/%v", height, round, cs.Height, cs.Round, cs.Step))
// Increment validators if necessary
validators := cs.Validators
if cs.Round < round {
validators = validators.Copy()
validators.IncrementAccum(round - cs.Round)
}
// Setup new round
// we don't fire newStep for this step,
// but we fire an event, so update the round step first
cs.updateRoundStep(round, cstypes.RoundStepNewRound)
cs.Validators = validators
if round == 0 {
// We've already reset these upon new height,
// and meanwhile we might have received a proposal
// for round 0.
} else {
logger.Info("Resetting Proposal info")
cs.Proposal = nil
cs.ProposalBlock = nil
cs.ProposalBlockParts = nil
}
cs.Votes.SetRound(round + 1) // also track next round (round+1) to allow round-skipping
cs.triggeredTimeoutPrecommit = false
cs.eventBus.PublishEventNewRound(cs.RoundStateEvent())
cs.metrics.Rounds.Set(float64(round))
// Wait for txs to be available in the mempool
// before we enterPropose in round 0. If the last block changed the app hash,
// we may need an empty "proof" block, and enterPropose immediately.
waitForTxs := cs.config.WaitForTxs() && round == 0 && !cs.needProofBlock(height)
if waitForTxs {
if cs.config.CreateEmptyBlocksInterval > 0 {
cs.scheduleTimeout(cs.config.CreateEmptyBlocksInterval, height, round,
cstypes.RoundStepNewRound)
}
go cs.proposalHeartbeat(height, round)
} else {
cs.enterPropose(height, round)
}
}
// needProofBlock returns true on the first height (so the genesis app hash is signed right away)
// and where the last block (height-1) caused the app hash to change
func (cs *ConsensusState) needProofBlock(height int64) bool {
if height == 1 {
return true
}
lastBlockMeta := cs.blockStore.LoadBlockMeta(height - 1)
return !bytes.Equal(cs.state.AppHash, lastBlockMeta.Header.AppHash)
}
func (cs *ConsensusState) proposalHeartbeat(height int64, round int) {
counter := 0
addr := cs.privValidator.GetAddress()
valIndex, _ := cs.Validators.GetByAddress(addr)
chainID := cs.state.ChainID
for {
rs := cs.GetRoundState()
// if we've already moved on, no need to send more heartbeats
if rs.Step > cstypes.RoundStepNewRound || rs.Round > round || rs.Height > height {
return
}
heartbeat := &types.Heartbeat{
Height: rs.Height,
Round: rs.Round,
Sequence: counter,
ValidatorAddress: addr,
ValidatorIndex: valIndex,
}
cs.privValidator.SignHeartbeat(chainID, heartbeat)
cs.eventBus.PublishEventProposalHeartbeat(types.EventDataProposalHeartbeat{heartbeat})
cs.evsw.FireEvent(types.EventProposalHeartbeat, heartbeat)
counter++
time.Sleep(proposalHeartbeatIntervalSeconds * time.Second)
}
}
// Enter (CreateEmptyBlocks): from enterNewRound(height,round)
// Enter (CreateEmptyBlocks, CreateEmptyBlocksInterval > 0 ): after enterNewRound(height,round), after timeout of CreateEmptyBlocksInterval
// Enter (!CreateEmptyBlocks) : after enterNewRound(height,round), once txs are in the mempool
func (cs *ConsensusState) enterPropose(height int64, round int) {
logger := cs.Logger.With("height", height, "round", round)
if cs.Height != height || round < cs.Round || (cs.Round == round && cstypes.RoundStepPropose <= cs.Step) {
logger.Debug(fmt.Sprintf("enterPropose(%v/%v): Invalid args. Current step: %v/%v/%v", height, round, cs.Height, cs.Round, cs.Step))
return
}
logger.Info(fmt.Sprintf("enterPropose(%v/%v). Current: %v/%v/%v", height, round, cs.Height, cs.Round, cs.Step))
defer func() {
// Done enterPropose:
cs.updateRoundStep(round, cstypes.RoundStepPropose)
cs.newStep()
// If we have the whole proposal + POL, then goto Prevote now.
// else, we'll enterPrevote when the rest of the proposal is received (in AddProposalBlockPart),
// or else after timeoutPropose
if cs.isProposalComplete() {
cs.enterPrevote(height, cs.Round)
}
}()
// If we don't get the proposal and all block parts quick enough, enterPrevote
cs.scheduleTimeout(cs.config.Propose(round), height, round, cstypes.RoundStepPropose)
// Nothing more to do if we're not a validator
if cs.privValidator == nil {
logger.Debug("This node is not a validator")
return
}
// if not a validator, we're done
if !cs.Validators.HasAddress(cs.privValidator.GetAddress()) {
logger.Debug("This node is not a validator", "addr", cs.privValidator.GetAddress(), "vals", cs.Validators)
return
}
logger.Debug("This node is a validator")
if cs.isProposer() {
logger.Info("enterPropose: Our turn to propose", "proposer", cs.Validators.GetProposer().Address, "privValidator", cs.privValidator)
cs.decideProposal(height, round)
} else {
logger.Info("enterPropose: Not our turn to propose", "proposer", cs.Validators.GetProposer().Address, "privValidator", cs.privValidator)
}
}
func (cs *ConsensusState) isProposer() bool {
return bytes.Equal(cs.Validators.GetProposer().Address, cs.privValidator.GetAddress())
}
func (cs *ConsensusState) defaultDecideProposal(height int64, round int) {
var block *types.Block
var blockParts *types.PartSet
// Decide on block
if cs.ValidBlock != nil {
// If there is valid block, choose that.
block, blockParts = cs.ValidBlock, cs.ValidBlockParts
} else {
// Create a new proposal block from state/txs from the mempool.
block, blockParts = cs.createProposalBlock()
if block == nil { // on error
return
}
}
// Make proposal
propBlockId := types.BlockID{block.Hash(), blockParts.Header()}
proposal := types.NewProposal(height, round, cs.ValidRound, propBlockId)
if err := cs.privValidator.SignProposal(cs.state.ChainID, proposal); err == nil {
// send proposal and block parts on internal msg queue
cs.sendInternalMessage(msgInfo{&ProposalMessage{proposal}, ""})
for i := 0; i < blockParts.Total(); i++ {
part := blockParts.GetPart(i)
cs.sendInternalMessage(msgInfo{&BlockPartMessage{cs.Height, cs.Round, part}, ""})
}
cs.Logger.Info("Signed proposal", "height", height, "round", round, "proposal", proposal)
cs.Logger.Debug(fmt.Sprintf("Signed proposal block: %v", block))
} else {
if !cs.replayMode {
cs.Logger.Error("enterPropose: Error signing proposal", "height", height, "round", round, "err", err)
}
}
}
// Returns true if the proposal block is complete &&
// (if POLRound was proposed, we have +2/3 prevotes from there).
func (cs *ConsensusState) isProposalComplete() bool {
if cs.Proposal == nil || cs.ProposalBlock == nil {
return false
}
// we have the proposal. if there's a POLRound,
// make sure we have the prevotes from it too
if cs.Proposal.POLRound < 0 {
return true
}
// if this is false the proposer is lying or we haven't received the POL yet
return cs.Votes.Prevotes(cs.Proposal.POLRound).HasTwoThirdsMajority()
}
// Create the next block to propose and return it.
// We really only need to return the parts, but the block
// is returned for convenience so we can log the proposal block.
// Returns nil block upon error.
// NOTE: keep it side-effect free for clarity.
func (cs *ConsensusState) createProposalBlock() (block *types.Block, blockParts *types.PartSet) {
var commit *types.Commit
if cs.Height == 1 {
// We're creating a proposal for the first block.
// The commit is empty, but not nil.
commit = &types.Commit{}
} else if cs.LastCommit.HasTwoThirdsMajority() {
// Make the commit from LastCommit
commit = cs.LastCommit.MakeCommit()
} else {
// This shouldn't happen.
cs.Logger.Error("enterPropose: Cannot propose anything: No commit for the previous block.")
return
}
maxBytes := cs.state.ConsensusParams.BlockSize.MaxBytes
maxGas := cs.state.ConsensusParams.BlockSize.MaxGas
// bound evidence to 1/10th of the block
evidence := cs.evpool.PendingEvidence(types.MaxEvidenceBytesPerBlock(maxBytes))
// Mempool validated transactions
txs := cs.mempool.ReapMaxBytesMaxGas(types.MaxDataBytes(
maxBytes,
cs.state.Validators.Size(),
len(evidence),
), maxGas)
proposerAddr := cs.privValidator.GetAddress()
block, parts := cs.state.MakeBlock(cs.Height, txs, commit, evidence, proposerAddr)
return block, parts
}
// Enter: `timeoutPropose` after entering Propose.
// Enter: proposal block and POL is ready.
// Prevote for LockedBlock if we're locked, or ProposalBlock if valid.
// Otherwise vote nil.
func (cs *ConsensusState) enterPrevote(height int64, round int) {
if cs.Height != height || round < cs.Round || (cs.Round == round && cstypes.RoundStepPrevote <= cs.Step) {
cs.Logger.Debug(fmt.Sprintf("enterPrevote(%v/%v): Invalid args. Current step: %v/%v/%v", height, round, cs.Height, cs.Round, cs.Step))
return
}
defer func() {
// Done enterPrevote:
cs.updateRoundStep(round, cstypes.RoundStepPrevote)
cs.newStep()
}()
cs.Logger.Info(fmt.Sprintf("enterPrevote(%v/%v). Current: %v/%v/%v", height, round, cs.Height, cs.Round, cs.Step))
// Sign and broadcast vote as necessary
cs.doPrevote(height, round)
// Once `addVote` hits any +2/3 prevotes, we will go to PrevoteWait
// (so we have more time to try and collect +2/3 prevotes for a single block)
}
func (cs *ConsensusState) defaultDoPrevote(height int64, round int) {