forked from dgraph-io/dgraph
-
Notifications
You must be signed in to change notification settings - Fork 0
/
oracle.go
424 lines (386 loc) · 10.2 KB
/
oracle.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
/*
* Copyright (C) 2017 Dgraph Labs, Inc. and Contributors
*
* This program 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 zero
import (
"errors"
"math/rand"
"time"
"github.com/dgraph-io/dgraph/protos/api"
"github.com/dgraph-io/dgraph/protos/intern"
"github.com/dgraph-io/dgraph/x"
"golang.org/x/net/context"
)
type syncMark struct {
index uint64
ts uint64
}
type Oracle struct {
x.SafeMutex
commits map[uint64]uint64 // startTs -> commitTs
// TODO: Check if we need LRU.
rowCommit map[string]uint64 // fp(key) -> commitTs. Used to detect conflict.
aborts map[uint64]struct{} // key is startTs
maxPending uint64 // max transaction startTs given out by us.
// timestamp at the time of start of server or when it became leader. Used to detect conflicts.
tmax uint64
// All transactions with startTs < startTxnTs return true for hasConflict.
startTxnTs uint64
subscribers map[int]chan *intern.OracleDelta
updates chan *intern.OracleDelta
doneUntil x.WaterMark
syncMarks []syncMark
}
func (o *Oracle) Init() {
o.commits = make(map[uint64]uint64)
o.rowCommit = make(map[string]uint64)
o.aborts = make(map[uint64]struct{})
o.subscribers = make(map[int]chan *intern.OracleDelta)
o.updates = make(chan *intern.OracleDelta, 100000) // Keeping 1 second worth of updates.
o.doneUntil.Init()
go o.sendDeltasToSubscribers()
}
func (o *Oracle) updateStartTxnTs(ts uint64) {
o.Lock()
defer o.Unlock()
o.startTxnTs = ts
o.rowCommit = make(map[string]uint64)
}
func (o *Oracle) hasConflict(src *api.TxnContext) bool {
// This transaction was started before I became leader.
if src.StartTs < o.startTxnTs {
return true
}
for _, k := range src.Keys {
if last := o.rowCommit[k]; last > src.StartTs {
return true
}
}
return false
}
func (o *Oracle) purgeBelow(minTs uint64) {
x.Printf("purging below ts:%d, len(o.commits):%d, len(o.aborts):%d"+
", len(o.rowCommit):%d\n",
minTs, len(o.commits), len(o.aborts), len(o.rowCommit))
o.Lock()
defer o.Unlock()
// Dropping would be cheaper if abort/commits map is sharded
for ts := range o.commits {
if ts < minTs {
delete(o.commits, ts)
}
}
for ts := range o.aborts {
if ts < minTs {
delete(o.aborts, ts)
}
}
// There is no transaction running with startTs less than minTs
// So we can delete everything from rowCommit whose commitTs < minTs
for key, ts := range o.rowCommit {
if ts < minTs {
delete(o.rowCommit, key)
}
}
o.tmax = minTs
}
func (o *Oracle) commit(src *api.TxnContext) error {
o.Lock()
defer o.Unlock()
if o.hasConflict(src) {
return errConflict
}
for _, k := range src.Keys {
o.rowCommit[k] = src.CommitTs // CommitTs is handed out before calling this func.
}
return nil
}
func (o *Oracle) currentState() *intern.OracleDelta {
o.AssertRLock()
resp := &intern.OracleDelta{
Commits: make(map[uint64]uint64, len(o.commits)),
}
for start, commit := range o.commits {
resp.Commits[start] = commit
}
for abort := range o.aborts {
resp.Aborts = append(resp.Aborts, abort)
}
resp.MaxPending = o.maxPending
return resp
}
func (o *Oracle) newSubscriber() (<-chan *intern.OracleDelta, int) {
o.Lock()
defer o.Unlock()
var id int
for {
id = rand.Int()
if _, has := o.subscribers[id]; !has {
break
}
}
ch := make(chan *intern.OracleDelta, 1000)
ch <- o.currentState() // Queue up the full state as the first entry.
o.subscribers[id] = ch
return ch, id
}
func (o *Oracle) removeSubscriber(id int) {
o.Lock()
defer o.Unlock()
delete(o.subscribers, id)
}
func (o *Oracle) sendDeltasToSubscribers() {
delta := &intern.OracleDelta{
Commits: make(map[uint64]uint64),
}
for {
update, open := <-o.updates
if !open {
return
}
slurp_loop:
for {
// Consume tctx.
if update.MaxPending > delta.MaxPending {
delta.MaxPending = update.MaxPending
}
for _, startTs := range update.Aborts {
delta.Aborts = append(delta.Aborts, startTs)
}
for startTs, commitTs := range update.Commits {
delta.Commits[startTs] = commitTs
}
select {
case update, open = <-o.updates:
if !open {
return
}
default:
break slurp_loop
}
}
o.Lock()
for id, ch := range o.subscribers {
select {
case ch <- delta:
default:
close(ch)
delete(o.subscribers, id)
}
}
o.Unlock()
delta = &intern.OracleDelta{
Commits: make(map[uint64]uint64),
}
}
}
func (o *Oracle) updateCommitStatusHelper(index uint64, src *api.TxnContext) bool {
o.Lock()
defer o.Unlock()
if _, ok := o.commits[src.StartTs]; ok {
return false
}
if _, ok := o.aborts[src.StartTs]; ok {
return false
}
if src.Aborted {
o.aborts[src.StartTs] = struct{}{}
} else {
o.commits[src.StartTs] = src.CommitTs
}
o.syncMarks = append(o.syncMarks, syncMark{index: index, ts: src.StartTs})
return true
}
func (o *Oracle) updateCommitStatus(index uint64, src *api.TxnContext) {
if o.updateCommitStatusHelper(index, src) {
delta := new(intern.OracleDelta)
if src.Aborted {
delta.Aborts = append(delta.Aborts, src.StartTs)
} else {
delta.Commits = make(map[uint64]uint64)
delta.Commits[src.StartTs] = src.CommitTs
}
o.updates <- delta
}
}
func (o *Oracle) commitTs(startTs uint64) uint64 {
o.RLock()
defer o.RUnlock()
return o.commits[startTs]
}
func (o *Oracle) storePending(ids *api.AssignedIds) {
// Wait to finish up processing everything before start id.
o.doneUntil.WaitForMark(context.Background(), ids.EndId)
// Now send it out to updates.
o.updates <- &intern.OracleDelta{MaxPending: ids.EndId}
o.Lock()
defer o.Unlock()
max := ids.EndId
if o.maxPending < max {
o.maxPending = max
}
}
var errConflict = errors.New("Transaction conflict")
func (s *Server) proposeTxn(ctx context.Context, src *api.TxnContext) error {
var zp intern.ZeroProposal
zp.Txn = &api.TxnContext{
StartTs: src.StartTs,
CommitTs: src.CommitTs,
Aborted: src.Aborted,
}
return s.Node.proposeAndWait(ctx, &zp)
}
func (s *Server) commit(ctx context.Context, src *api.TxnContext) error {
if src.Aborted {
return s.proposeTxn(ctx, src)
}
// Use the start timestamp to check if we have a conflict, before we need to assign a commit ts.
s.orc.RLock()
conflict := s.orc.hasConflict(src)
s.orc.RUnlock()
if conflict {
src.Aborted = true
return s.proposeTxn(ctx, src)
}
var num intern.Num
num.Val = 1
assigned, err := s.lease(ctx, &num, true)
if err != nil {
return err
}
src.CommitTs = assigned.StartId
if err := s.orc.commit(src); err != nil {
src.Aborted = true
}
// Propose txn should be used to set watermark as done.
err = s.proposeTxn(ctx, src)
// Mark the transaction as done, irrespective of whether the proposal succeeded or not.
s.orc.doneUntil.Done(src.CommitTs)
return err
}
func (s *Server) CommitOrAbort(ctx context.Context, src *api.TxnContext) (*api.TxnContext, error) {
if ctx.Err() != nil {
return nil, ctx.Err()
}
if !s.Node.AmLeader() {
return nil, x.Errorf("Only leader can decide to commit or abort")
}
err := s.commit(ctx, src)
return src, err
}
var errClosed = errors.New("Streaming closed by Oracle.")
var errNotLeader = errors.New("Node is no longer leader.")
func (s *Server) Oracle(unused *api.Payload, server intern.Zero_OracleServer) error {
if !s.Node.AmLeader() {
return errNotLeader
}
ch, id := s.orc.newSubscriber()
defer s.orc.removeSubscriber(id)
ctx := server.Context()
leaderChangeCh := s.leaderChangeChannel()
for {
select {
case <-leaderChangeCh:
return errNotLeader
case delta, open := <-ch:
if !open {
return errClosed
}
if err := server.Send(delta); err != nil {
return err
}
case <-ctx.Done():
return ctx.Err()
case <-s.shutDownCh:
return errServerShutDown
}
}
return nil
}
func (s *Server) SyncedUntil() uint64 {
s.orc.Lock()
defer s.orc.Unlock()
// Find max index with timestamp less than tmax
var idx int
for i, sm := range s.orc.syncMarks {
idx = i
if sm.ts >= s.orc.tmax {
break
}
}
var syncUntil uint64
if idx > 0 {
syncUntil = s.orc.syncMarks[idx-1].index
}
s.orc.syncMarks = s.orc.syncMarks[idx:]
return syncUntil
}
func (s *Server) purgeOracle() {
ticker := time.NewTicker(time.Second * 10)
defer ticker.Stop()
var lastPurgeTs uint64
OUTER:
for {
<-ticker.C
groups := s.KnownGroups()
var minTs uint64
for _, group := range groups {
pl := s.Leader(group)
if pl == nil {
x.Printf("No healthy connection found to leader of group %d\n", group)
goto OUTER
}
c := intern.NewWorkerClient(pl.Get())
num, err := c.MinTxnTs(context.Background(), &api.Payload{})
if err != nil {
x.Printf("Error while fetching minTs from group %d, err: %v\n", group, err)
goto OUTER
}
if minTs == 0 || num.Val < minTs {
minTs = num.Val
}
}
if minTs > 0 && minTs != lastPurgeTs {
s.orc.purgeBelow(minTs)
lastPurgeTs = minTs
}
}
}
func (s *Server) TryAbort(ctx context.Context, txns *intern.TxnTimestamps) (*intern.TxnTimestamps, error) {
commitTimestamps := new(intern.TxnTimestamps)
for _, startTs := range txns.Ts {
// Do via proposals to avoid race
tctx := &api.TxnContext{StartTs: startTs, Aborted: true}
if err := s.proposeTxn(ctx, tctx); err != nil {
return commitTimestamps, err
}
// Txn should be aborted if not already committed.
commitTimestamps.Ts = append(commitTimestamps.Ts, s.orc.commitTs(startTs))
}
return commitTimestamps, nil
}
// Timestamps is used to assign startTs for a new transaction
func (s *Server) Timestamps(ctx context.Context, num *intern.Num) (*api.AssignedIds, error) {
if ctx.Err() != nil {
return &emptyAssignedIds, ctx.Err()
}
reply, err := s.lease(ctx, num, true)
if err == nil {
s.orc.doneUntil.Done(reply.EndId)
go s.orc.storePending(reply)
}
return reply, err
}