-
Notifications
You must be signed in to change notification settings - Fork 262
/
service_v1.go
1453 lines (1270 loc) · 48.5 KB
/
service_v1.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
/*
* Copyright 2020 The Dragonfly Authors
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package service
import (
"context"
"errors"
"fmt"
"io"
"math"
"strings"
"time"
"google.golang.org/grpc/status"
commonv1 "d7y.io/api/pkg/apis/common/v1"
commonv2 "d7y.io/api/pkg/apis/common/v2"
errordetailsv1 "d7y.io/api/pkg/apis/errordetails/v1"
schedulerv1 "d7y.io/api/pkg/apis/scheduler/v1"
"d7y.io/dragonfly/v2/internal/dferrors"
logger "d7y.io/dragonfly/v2/internal/dflog"
"d7y.io/dragonfly/v2/pkg/container/set"
"d7y.io/dragonfly/v2/pkg/digest"
"d7y.io/dragonfly/v2/pkg/idgen"
"d7y.io/dragonfly/v2/pkg/net/http"
"d7y.io/dragonfly/v2/pkg/rpc/common"
"d7y.io/dragonfly/v2/pkg/types"
"d7y.io/dragonfly/v2/scheduler/config"
"d7y.io/dragonfly/v2/scheduler/metrics"
"d7y.io/dragonfly/v2/scheduler/resource"
"d7y.io/dragonfly/v2/scheduler/scheduling"
"d7y.io/dragonfly/v2/scheduler/storage"
)
// V1 is the interface for v1 version of the service.
type V1 struct {
// Resource interface.
resource resource.Resource
// Scheduling interface.
scheduling scheduling.Scheduling
// Scheduelr service config.
config *config.Config
// Dynamic config.
dynconfig config.DynconfigInterface
// Storage interface.
storage storage.Storage
}
// New v1 version of service instance.
func NewV1(
cfg *config.Config,
resource resource.Resource,
scheduling scheduling.Scheduling,
dynconfig config.DynconfigInterface,
storage storage.Storage,
) *V1 {
return &V1{
resource: resource,
scheduling: scheduling,
config: cfg,
dynconfig: dynconfig,
storage: storage,
}
}
// RegisterPeerTask registers peer and triggers seed peer download task.
func (v *V1) RegisterPeerTask(ctx context.Context, req *schedulerv1.PeerTaskRequest) (*schedulerv1.RegisterResult, error) {
logger.WithPeer(req.PeerHost.Id, req.TaskId, req.PeerId).Infof("register peer task request: %#v %#v",
req, req.UrlMeta)
// Store resource.
task := v.storeTask(ctx, req, commonv2.TaskType_DFDAEMON)
host := v.storeHost(ctx, req.PeerHost)
peer := v.storePeer(ctx, req.PeerId, req.UrlMeta.Priority, req.UrlMeta.Range, task, host)
// Trigger the first download of the task.
if err := v.triggerTask(ctx, req, task, host, peer, v.dynconfig); err != nil {
peer.Log.Error(err)
v.handleRegisterFailure(ctx, peer)
return nil, dferrors.New(commonv1.Code_SchedForbidden, err.Error())
}
// If the task does not succeed, it is scheduled as a normal task.
if !task.FSM.Is(resource.TaskStateSucceeded) {
peer.Log.Infof("register as normal task, because of task state is %s",
task.FSM.Current())
result, err := v.registerNormalTask(ctx, peer)
if err != nil {
peer.Log.Error(err)
v.handleRegisterFailure(ctx, peer)
return nil, dferrors.New(commonv1.Code_SchedError, err.Error())
}
return result, nil
}
// If SizeScope is SizeScope_UNKNOW, then register as SizeScope_NORMAL.
sizeScope := types.SizeScopeV2ToV1(task.SizeScope())
peer.Log.Infof("task size scope is %s", sizeScope)
// The task state is TaskStateSucceeded and SizeScope is not invalid.
switch sizeScope {
case commonv1.SizeScope_EMPTY:
result, err := v.registerEmptyTask(ctx, peer)
if err != nil {
peer.Log.Error(err)
v.handleRegisterFailure(ctx, peer)
return nil, dferrors.New(commonv1.Code_SchedError, err.Error())
}
return result, nil
case commonv1.SizeScope_TINY:
// Validate data of direct piece.
if !peer.Task.CanReuseDirectPiece() {
peer.Log.Warnf("register as normal task, because of length of direct piece is %d, content length is %d",
len(task.DirectPiece), task.ContentLength.Load())
break
}
result, err := v.registerTinyTask(ctx, peer)
if err != nil {
peer.Log.Warnf("register as normal task, because of %s", err.Error())
break
}
return result, nil
case commonv1.SizeScope_SMALL:
result, err := v.registerSmallTask(ctx, peer)
if err != nil {
peer.Log.Warnf("register as normal task, because of %s", err.Error())
break
}
return result, nil
}
result, err := v.registerNormalTask(ctx, peer)
if err != nil {
peer.Log.Error(err)
v.handleRegisterFailure(ctx, peer)
return nil, dferrors.New(commonv1.Code_SchedError, err.Error())
}
peer.Log.Info("register as normal task, because of invalid size scope")
return result, nil
}
// ReportPieceResult handles the piece information reported by dfdaemon.
func (v *V1) ReportPieceResult(stream schedulerv1.Scheduler_ReportPieceResultServer) error {
ctx, cancel := context.WithCancel(stream.Context())
defer cancel()
var (
peer *resource.Peer
initialized bool
loaded bool
)
for {
select {
case <-ctx.Done():
logger.Infof("context was done")
return ctx.Err()
default:
}
piece, err := stream.Recv()
if err != nil {
if err == io.EOF {
return nil
}
logger.Errorf("receive piece failed: %s", err.Error())
return err
}
if !initialized {
initialized = true
// Get peer from peer manager.
peer, loaded = v.resource.PeerManager().Load(piece.SrcPid)
if !loaded {
// If the scheduler cannot find the peer,
// then the peer has not been registered in this scheduler.
msg := fmt.Sprintf("peer %s not found", piece.SrcPid)
logger.Error(msg)
return dferrors.New(commonv1.Code_SchedReregister, msg)
}
// Peer setting stream.
peer.StoreReportPieceResultStream(stream)
defer peer.DeleteReportPieceResultStream()
}
if piece.PieceInfo != nil {
// Handle begin of piece.
if piece.PieceInfo.PieceNum == common.BeginOfPiece {
peer.Log.Infof("receive begin of piece: %#v %#v", piece, piece.PieceInfo)
v.handleBeginOfPiece(ctx, peer)
continue
}
// Handle end of piece.
if piece.PieceInfo.PieceNum == common.EndOfPiece {
peer.Log.Infof("receive end of piece: %#v %#v", piece, piece.PieceInfo)
v.handleEndOfPiece(ctx, peer)
continue
}
}
// Handle piece download successfully.
if piece.Success {
peer.Log.Infof("receive success piece: %#v %#v", piece, piece.PieceInfo)
v.handlePieceSuccess(ctx, peer, piece)
// Collect host traffic metrics.
if v.config.Metrics.Enable && v.config.Metrics.EnableHost {
metrics.HostTraffic.WithLabelValues(metrics.HostTrafficDownloadType, peer.Task.Type.String(), peer.Task.Tag, peer.Task.Application,
peer.Host.Type.Name(), peer.Host.ID, peer.Host.IP, peer.Host.Hostname).Add(float64(piece.PieceInfo.RangeSize))
if parent, loaded := v.resource.PeerManager().Load(piece.DstPid); loaded {
metrics.HostTraffic.WithLabelValues(metrics.HostTrafficUploadType, peer.Task.Type.String(), peer.Task.Tag, peer.Task.Application,
parent.Host.Type.Name(), parent.Host.ID, parent.Host.IP, parent.Host.Hostname).Add(float64(piece.PieceInfo.RangeSize))
} else if !resource.IsPieceBackToSource(piece.DstPid) {
peer.Log.Warnf("dst peer %s not found", piece.DstPid)
}
}
// Collect traffic metrics.
if !resource.IsPieceBackToSource(piece.DstPid) {
metrics.Traffic.WithLabelValues(commonv2.TrafficType_REMOTE_PEER.String(), peer.Task.Type.String(),
peer.Task.Tag, peer.Task.Application, peer.Host.Type.Name()).Add(float64(piece.PieceInfo.RangeSize))
} else {
metrics.Traffic.WithLabelValues(commonv2.TrafficType_BACK_TO_SOURCE.String(), peer.Task.Type.String(),
peer.Task.Tag, peer.Task.Application, peer.Host.Type.Name()).Add(float64(piece.PieceInfo.RangeSize))
}
continue
}
// Handle piece download code.
if piece.Code != commonv1.Code_Success {
if piece.Code == commonv1.Code_ClientWaitPieceReady {
peer.Log.Debug("receive wait piece")
continue
}
// Handle piece download failed.
peer.Log.Errorf("receive failed piece: %#v", piece)
v.handlePieceFailure(ctx, peer, piece)
continue
}
peer.Log.Warnf("receive unknow piece: %#v %#v", piece, piece.PieceInfo)
}
}
// ReportPeerResult handles peer result reported by dfdaemon.
func (v *V1) ReportPeerResult(ctx context.Context, req *schedulerv1.PeerResult) error {
logger.WithTaskAndPeerID(req.TaskId, req.PeerId).Infof("report peer result request: %#v", req)
peer, loaded := v.resource.PeerManager().Load(req.PeerId)
if !loaded {
msg := fmt.Sprintf("peer %s not found", req.PeerId)
logger.Error(msg)
return dferrors.New(commonv1.Code_SchedPeerNotFound, msg)
}
// Collect DownloadPeerCount metrics.
priority := peer.CalculatePriority(v.dynconfig)
metrics.DownloadPeerCount.WithLabelValues(priority.String(), peer.Task.Type.String(),
peer.Task.Tag, peer.Task.Application, peer.Host.Type.Name()).Inc()
parents := peer.Parents()
if !req.Success {
peer.Log.Error("report failed peer")
if peer.FSM.Is(resource.PeerStateBackToSource) {
// Collect DownloadPeerBackToSourceFailureCount metrics.
metrics.DownloadPeerBackToSourceFailureCount.WithLabelValues(priority.String(), peer.Task.Type.String(),
peer.Task.Tag, peer.Task.Application, peer.Host.Type.Name()).Inc()
go v.createRecord(peer, parents, req)
v.handleTaskFailure(ctx, peer.Task, req.GetSourceError(), nil)
v.handlePeerFailure(ctx, peer)
return nil
}
// Collect DownloadPeerFailureCount metrics.
metrics.DownloadPeerFailureCount.WithLabelValues(priority.String(), peer.Task.Type.String(),
peer.Task.Tag, peer.Task.Application, peer.Host.Type.Name()).Inc()
go v.createRecord(peer, parents, req)
v.handlePeerFailure(ctx, peer)
return nil
}
// Collect DownloadPeerDuration metrics.
metrics.DownloadPeerDuration.WithLabelValues(priority.String(), peer.Task.Type.String(),
peer.Task.Tag, peer.Task.Application, peer.Host.Type.Name()).Observe(float64(req.Cost))
peer.Log.Info("report success peer")
if peer.FSM.Is(resource.PeerStateBackToSource) {
go v.createRecord(peer, parents, req)
v.handleTaskSuccess(ctx, peer.Task, req)
v.handlePeerSuccess(ctx, peer)
return nil
}
go v.createRecord(peer, parents, req)
v.handlePeerSuccess(ctx, peer)
return nil
}
// AnnounceTask informs scheduler a peer has completed task.
func (v *V1) AnnounceTask(ctx context.Context, req *schedulerv1.AnnounceTaskRequest) error {
logger.WithPeer(req.PeerHost.Id, req.TaskId, req.PiecePacket.DstPid).Infof("announce task request: %#v %#v %#v %#v",
req, req.UrlMeta, req.PeerHost, req.PiecePacket,
)
taskID := req.TaskId
peerID := req.PiecePacket.DstPid
options := []resource.TaskOption{}
if d, err := digest.Parse(req.UrlMeta.Digest); err == nil {
options = append(options, resource.WithDigest(d))
}
task := resource.NewTask(taskID, req.Url, req.UrlMeta.Tag, req.UrlMeta.Application, types.TaskTypeV1ToV2(req.TaskType),
strings.Split(req.UrlMeta.Filter, idgen.URLFilterSeparator), req.UrlMeta.Header, int32(v.config.Scheduler.BackToSourceCount), options...)
task, _ = v.resource.TaskManager().LoadOrStore(task)
host := v.storeHost(ctx, req.PeerHost)
peer := v.storePeer(ctx, peerID, req.UrlMeta.Priority, req.UrlMeta.Range, task, host)
// If the task state is not TaskStateSucceeded,
// advance the task state to TaskStateSucceeded.
if !task.FSM.Is(resource.TaskStateSucceeded) {
if task.FSM.Can(resource.TaskEventDownload) {
if err := task.FSM.Event(ctx, resource.TaskEventDownload); err != nil {
msg := fmt.Sprintf("task fsm event failed: %s", err.Error())
peer.Log.Error(msg)
return dferrors.New(commonv1.Code_SchedError, msg)
}
}
// Construct piece.
for _, pieceInfo := range req.PiecePacket.PieceInfos {
piece := &resource.Piece{
Number: pieceInfo.PieceNum,
ParentID: req.PiecePacket.DstPid,
Offset: pieceInfo.RangeStart,
Length: uint64(pieceInfo.RangeSize),
TrafficType: commonv2.TrafficType_LOCAL_PEER,
Cost: 0,
CreatedAt: time.Now(),
}
if len(pieceInfo.PieceMd5) > 0 {
piece.Digest = digest.New(digest.AlgorithmMD5, pieceInfo.PieceMd5)
}
peer.StorePiece(piece)
peer.FinishedPieces.Set(uint(pieceInfo.PieceNum))
peer.AppendPieceCost(piece.Cost)
task.StorePiece(piece)
}
v.handleTaskSuccess(ctx, task, &schedulerv1.PeerResult{
TotalPieceCount: req.PiecePacket.TotalPiece,
ContentLength: req.PiecePacket.ContentLength,
})
}
// If the peer state is not PeerStateSucceeded,
// advance the peer state to PeerStateSucceeded.
if !peer.FSM.Is(resource.PeerStateSucceeded) {
if peer.FSM.Is(resource.PeerStatePending) {
if err := peer.FSM.Event(ctx, resource.PeerEventRegisterNormal); err != nil {
msg := fmt.Sprintf("peer fsm event failed: %s", err.Error())
peer.Log.Error(msg)
return dferrors.New(commonv1.Code_SchedError, msg)
}
}
if peer.FSM.Is(resource.PeerStateReceivedTiny) ||
peer.FSM.Is(resource.PeerStateReceivedSmall) ||
peer.FSM.Is(resource.PeerStateReceivedNormal) {
if err := peer.FSM.Event(ctx, resource.PeerEventDownload); err != nil {
msg := fmt.Sprintf("peer fsm event failed: %s", err.Error())
peer.Log.Error(msg)
return dferrors.New(commonv1.Code_SchedError, msg)
}
v.handlePeerSuccess(ctx, peer)
}
}
return nil
}
// StatTask checks the current state of the task.
func (v *V1) StatTask(ctx context.Context, req *schedulerv1.StatTaskRequest) (*schedulerv1.Task, error) {
logger.WithTaskID(req.TaskId).Infof("stat task request: %#v", req)
task, loaded := v.resource.TaskManager().Load(req.TaskId)
if !loaded {
msg := fmt.Sprintf("task %s not found", req.TaskId)
logger.Info(msg)
return nil, dferrors.New(commonv1.Code_PeerTaskNotFound, msg)
}
return &schedulerv1.Task{
Id: task.ID,
Type: types.TaskTypeV2ToV1(task.Type),
ContentLength: task.ContentLength.Load(),
TotalPieceCount: task.TotalPieceCount.Load(),
State: task.FSM.Current(),
PeerCount: int32(task.PeerCount()),
HasAvailablePeer: task.HasAvailablePeer(set.NewSafeSet[string]()),
}, nil
}
// LeaveTask releases peer in scheduler.
func (v *V1) LeaveTask(ctx context.Context, req *schedulerv1.PeerTarget) error {
logger.WithTaskAndPeerID(req.TaskId, req.PeerId).Infof("leave task request: %#v", req)
peer, loaded := v.resource.PeerManager().Load(req.PeerId)
if !loaded {
msg := fmt.Sprintf("peer %s not found", req.PeerId)
logger.Error(msg)
return dferrors.New(commonv1.Code_SchedPeerNotFound, msg)
}
if err := peer.FSM.Event(ctx, resource.PeerEventLeave); err != nil {
msg := fmt.Sprintf("peer fsm event failed: %s", err.Error())
peer.Log.Error(msg)
return dferrors.New(commonv1.Code_SchedTaskStatusError, msg)
}
return nil
}
// AnnounceHost announces host to scheduler.
func (v *V1) AnnounceHost(ctx context.Context, req *schedulerv1.AnnounceHostRequest) error {
// Get scheduler cluster client config by manager.
var concurrentUploadLimit int32
if clientConfig, err := v.dynconfig.GetSchedulerClusterClientConfig(); err == nil {
concurrentUploadLimit = int32(clientConfig.LoadLimit)
}
host, loaded := v.resource.HostManager().Load(req.Id)
if !loaded {
options := []resource.HostOption{
resource.WithOS(req.Os),
resource.WithPlatform(req.Platform),
resource.WithPlatformFamily(req.PlatformFamily),
resource.WithPlatformVersion(req.PlatformVersion),
resource.WithKernelVersion(req.KernelVersion),
}
if concurrentUploadLimit > 0 {
options = append(options, resource.WithConcurrentUploadLimit(concurrentUploadLimit))
}
if req.Cpu != nil {
options = append(options, resource.WithCPU(resource.CPU{
LogicalCount: req.Cpu.LogicalCount,
PhysicalCount: req.Cpu.PhysicalCount,
Percent: req.Cpu.Percent,
ProcessPercent: req.Cpu.ProcessPercent,
Times: resource.CPUTimes{
User: req.Cpu.Times.User,
System: req.Cpu.Times.System,
Idle: req.Cpu.Times.Idle,
Nice: req.Cpu.Times.Nice,
Iowait: req.Cpu.Times.Iowait,
Irq: req.Cpu.Times.Irq,
Softirq: req.Cpu.Times.Softirq,
Steal: req.Cpu.Times.Steal,
Guest: req.Cpu.Times.Guest,
GuestNice: req.Cpu.Times.GuestNice,
},
}))
}
if req.Memory != nil {
options = append(options, resource.WithMemory(resource.Memory{
Total: req.Memory.Total,
Available: req.Memory.Available,
Used: req.Memory.Used,
UsedPercent: req.Memory.UsedPercent,
ProcessUsedPercent: req.Memory.ProcessUsedPercent,
Free: req.Memory.Free,
}))
}
if req.Network != nil {
options = append(options, resource.WithNetwork(resource.Network{
TCPConnectionCount: req.Network.TcpConnectionCount,
UploadTCPConnectionCount: req.Network.UploadTcpConnectionCount,
SecurityDomain: req.Network.SecurityDomain,
Location: req.Network.Location,
IDC: req.Network.Idc,
}))
}
if req.Disk != nil {
options = append(options, resource.WithDisk(resource.Disk{
Total: req.Disk.Total,
Free: req.Disk.Free,
Used: req.Disk.Used,
UsedPercent: req.Disk.UsedPercent,
InodesTotal: req.Disk.InodesTotal,
InodesUsed: req.Disk.InodesUsed,
InodesFree: req.Disk.InodesFree,
InodesUsedPercent: req.Disk.InodesUsedPercent,
}))
}
if req.Build != nil {
options = append(options, resource.WithBuild(resource.Build{
GitVersion: req.Build.GitVersion,
GitCommit: req.Build.GitCommit,
GoVersion: req.Build.GoVersion,
Platform: req.Build.Platform,
}))
}
host = resource.NewHost(
req.Id, req.Ip, req.Hostname,
req.Port, req.DownloadPort, types.ParseHostType(req.Type),
options...,
)
v.resource.HostManager().Store(host)
host.Log.Infof("announce new host: %#v", req)
return nil
}
// Host already exists and updates properties.
host.Port = req.Port
host.DownloadPort = req.DownloadPort
host.Type = types.ParseHostType(req.Type)
host.OS = req.Os
host.Platform = req.Platform
host.PlatformFamily = req.PlatformFamily
host.PlatformVersion = req.PlatformVersion
host.KernelVersion = req.KernelVersion
host.UpdatedAt.Store(time.Now())
if concurrentUploadLimit > 0 {
host.ConcurrentUploadLimit.Store(concurrentUploadLimit)
}
if req.Cpu != nil {
host.CPU = resource.CPU{
LogicalCount: req.Cpu.LogicalCount,
PhysicalCount: req.Cpu.PhysicalCount,
Percent: req.Cpu.Percent,
ProcessPercent: req.Cpu.ProcessPercent,
Times: resource.CPUTimes{
User: req.Cpu.Times.User,
System: req.Cpu.Times.System,
Idle: req.Cpu.Times.Idle,
Nice: req.Cpu.Times.Nice,
Iowait: req.Cpu.Times.Iowait,
Irq: req.Cpu.Times.Irq,
Softirq: req.Cpu.Times.Softirq,
Steal: req.Cpu.Times.Steal,
Guest: req.Cpu.Times.Guest,
GuestNice: req.Cpu.Times.GuestNice,
},
}
}
if req.Memory != nil {
host.Memory = resource.Memory{
Total: req.Memory.Total,
Available: req.Memory.Available,
Used: req.Memory.Used,
UsedPercent: req.Memory.UsedPercent,
ProcessUsedPercent: req.Memory.ProcessUsedPercent,
Free: req.Memory.Free,
}
}
if req.Network != nil {
host.Network = resource.Network{
TCPConnectionCount: req.Network.TcpConnectionCount,
UploadTCPConnectionCount: req.Network.UploadTcpConnectionCount,
SecurityDomain: req.Network.SecurityDomain,
Location: req.Network.Location,
IDC: req.Network.Idc,
}
}
if req.Disk != nil {
host.Disk = resource.Disk{
Total: req.Disk.Total,
Free: req.Disk.Free,
Used: req.Disk.Used,
UsedPercent: req.Disk.UsedPercent,
InodesTotal: req.Disk.InodesTotal,
InodesUsed: req.Disk.InodesUsed,
InodesFree: req.Disk.InodesFree,
InodesUsedPercent: req.Disk.InodesUsedPercent,
}
}
if req.Build != nil {
host.Build = resource.Build{
GitVersion: req.Build.GitVersion,
GitCommit: req.Build.GitCommit,
GoVersion: req.Build.GoVersion,
Platform: req.Build.Platform,
}
}
return nil
}
// LeaveHost releases host in scheduler.
func (v *V1) LeaveHost(ctx context.Context, req *schedulerv1.LeaveHostRequest) error {
logger.WithHostID(req.Id).Infof("leave host request: %#v", req)
host, loaded := v.resource.HostManager().Load(req.Id)
if !loaded {
msg := fmt.Sprintf("host %s not found", req.Id)
logger.Error(msg)
return dferrors.New(commonv1.Code_BadRequest, msg)
}
host.LeavePeers()
return nil
}
// triggerTask triggers the first download of the task.
func (v *V1) triggerTask(ctx context.Context, req *schedulerv1.PeerTaskRequest, task *resource.Task, host *resource.Host, peer *resource.Peer, dynconfig config.DynconfigInterface) error {
// If task has available peer, peer does not need to be triggered.
blocklist := set.NewSafeSet[string]()
blocklist.Add(peer.ID)
if (task.FSM.Is(resource.TaskStateRunning) ||
task.FSM.Is(resource.TaskStateSucceeded)) &&
task.HasAvailablePeer(blocklist) {
peer.Log.Info("peer does not need to trigger")
return nil
}
// If the task triggers the TaskEventDownload failed and it has no available peer,
// let the peer do the scheduling.
if task.FSM.Can(resource.TaskEventDownload) {
if err := task.FSM.Event(ctx, resource.TaskEventDownload); err != nil {
peer.Log.Errorf("task fsm event failed: %s", err.Error())
return err
}
}
// If host type is not HostTypeNormal, then it needs to back-to-source.
if host.Type != types.HostTypeNormal {
peer.Log.Infof("peer back-to-source, because of host type is %d", host.Type)
peer.NeedBackToSource.Store(true)
return nil
}
// The first download is triggered according to
// the different priorities of the peer and
// priority of the RegisterPeerTask parameter is
// higher than parameter of the application.
var priority commonv1.Priority
if req.UrlMeta.Priority != commonv1.Priority_LEVEL0 {
priority = req.UrlMeta.Priority
} else {
// Compatible with v1 version of priority enum.
priority = types.PriorityV2ToV1(peer.CalculatePriority(dynconfig))
}
peer.Log.Infof("peer priority is %d", priority)
switch priority {
case commonv1.Priority_LEVEL6, commonv1.Priority_LEVEL0:
if v.config.SeedPeer.Enable && !task.IsSeedPeerFailed() {
if len(req.UrlMeta.Range) > 0 {
if rg, err := http.ParseURLMetaRange(req.UrlMeta.Range, math.MaxInt64); err == nil {
go v.triggerSeedPeerTask(ctx, &rg, task)
return nil
}
peer.Log.Errorf("range %s is invalid", req.UrlMeta.Range)
} else {
go v.triggerSeedPeerTask(ctx, nil, task)
return nil
}
}
fallthrough
case commonv1.Priority_LEVEL5:
fallthrough
case commonv1.Priority_LEVEL4:
fallthrough
case commonv1.Priority_LEVEL3:
peer.Log.Infof("peer back-to-source, because of hitting priority %d", commonv1.Priority_LEVEL3)
peer.NeedBackToSource.Store(true)
return nil
case commonv1.Priority_LEVEL2:
return fmt.Errorf("priority is %d and no available peers", commonv1.Priority_LEVEL2)
case commonv1.Priority_LEVEL1:
return fmt.Errorf("priority is %d", commonv1.Priority_LEVEL1)
}
peer.Log.Infof("peer back-to-source, because of peer has invalid priority %d", priority)
peer.NeedBackToSource.Store(true)
return nil
}
// triggerSeedPeerTask starts to trigger seed peer task.
func (v *V1) triggerSeedPeerTask(ctx context.Context, rg *http.Range, task *resource.Task) {
task.Log.Info("trigger seed peer")
peer, endOfPiece, err := v.resource.SeedPeer().TriggerTask(ctx, rg, task)
if err != nil {
task.Log.Errorf("trigger seed peer failed: %s", err.Error())
v.handleTaskFailure(ctx, task, nil, err)
return
}
// Update the task status first to help peer scheduling evaluation and scoring.
peer.Log.Info("trigger seed peer successfully")
v.handleTaskSuccess(ctx, task, endOfPiece)
v.handlePeerSuccess(ctx, peer)
}
// storeTask stores a new task or reuses a previous task.
func (v *V1) storeTask(ctx context.Context, req *schedulerv1.PeerTaskRequest, typ commonv2.TaskType) *resource.Task {
filters := strings.Split(req.UrlMeta.Filter, idgen.URLFilterSeparator)
task, loaded := v.resource.TaskManager().Load(req.TaskId)
if !loaded {
options := []resource.TaskOption{}
if d, err := digest.Parse(req.UrlMeta.Digest); err == nil {
options = append(options, resource.WithDigest(d))
}
task := resource.NewTask(req.TaskId, req.Url, req.UrlMeta.Tag, req.UrlMeta.Application,
typ, filters, req.UrlMeta.Header, int32(v.config.Scheduler.BackToSourceCount), options...)
v.resource.TaskManager().Store(task)
task.Log.Info("create new task")
return task
}
// Task is the pointer, if the task already exists, the next request will
// update the task's Url and UrlMeta in task manager.
task.URL = req.Url
task.Filters = filters
task.Header = req.UrlMeta.Header
task.Log.Info("task already exists")
return task
}
// storeHost stores a new host or reuses a previous host.
func (v *V1) storeHost(ctx context.Context, peerHost *schedulerv1.PeerHost) *resource.Host {
host, loaded := v.resource.HostManager().Load(peerHost.Id)
if !loaded {
options := []resource.HostOption{resource.WithNetwork(resource.Network{
SecurityDomain: peerHost.SecurityDomain,
Location: peerHost.Location,
IDC: peerHost.Idc,
})}
if clientConfig, err := v.dynconfig.GetSchedulerClusterClientConfig(); err == nil && clientConfig.LoadLimit > 0 {
options = append(options, resource.WithConcurrentUploadLimit(int32(clientConfig.LoadLimit)))
}
host := resource.NewHost(
peerHost.Id, peerHost.Ip, peerHost.Hostname,
peerHost.RpcPort, peerHost.DownPort, types.HostTypeNormal,
options...,
)
v.resource.HostManager().Store(host)
host.Log.Info("create new host")
return host
}
host.Port = peerHost.RpcPort
host.DownloadPort = peerHost.DownPort
host.Network.SecurityDomain = peerHost.SecurityDomain
host.Network.Location = peerHost.Location
host.Network.IDC = peerHost.Idc
host.UpdatedAt.Store(time.Now())
host.Log.Info("host already exists")
return host
}
// storePeer stores a new peer or reuses a previous peer.
func (v *V1) storePeer(ctx context.Context, id string, priority commonv1.Priority, rg string, task *resource.Task, host *resource.Host) *resource.Peer {
peer, loaded := v.resource.PeerManager().Load(id)
if !loaded {
options := []resource.PeerOption{}
if priority != commonv1.Priority_LEVEL0 {
options = append(options, resource.WithPriority(types.PriorityV1ToV2(priority)))
}
if len(rg) > 0 {
if r, err := http.ParseURLMetaRange(rg, math.MaxInt64); err == nil {
options = append(options, resource.WithRange(r))
} else {
logger.WithPeer(host.ID, task.ID, id).Error(err)
}
}
peer := resource.NewPeer(id, task, host, options...)
v.resource.PeerManager().Store(peer)
peer.Log.Info("create new peer")
return peer
}
peer.Log.Info("peer already exists")
return peer
}
// registerEmptyTask registers the empty task.
func (v *V1) registerEmptyTask(ctx context.Context, peer *resource.Peer) (*schedulerv1.RegisterResult, error) {
if err := peer.FSM.Event(ctx, resource.PeerEventRegisterEmpty); err != nil {
return nil, err
}
return &schedulerv1.RegisterResult{
TaskId: peer.Task.ID,
TaskType: types.TaskTypeV2ToV1(peer.Task.Type),
SizeScope: commonv1.SizeScope_EMPTY,
DirectPiece: &schedulerv1.RegisterResult_PieceContent{
PieceContent: []byte{},
},
}, nil
}
// registerEmptyTask registers the tiny task.
func (v *V1) registerTinyTask(ctx context.Context, peer *resource.Peer) (*schedulerv1.RegisterResult, error) {
if err := peer.FSM.Event(ctx, resource.PeerEventRegisterTiny); err != nil {
return nil, err
}
return &schedulerv1.RegisterResult{
TaskId: peer.Task.ID,
TaskType: types.TaskTypeV2ToV1(peer.Task.Type),
SizeScope: commonv1.SizeScope_TINY,
DirectPiece: &schedulerv1.RegisterResult_PieceContent{
PieceContent: peer.Task.DirectPiece,
},
}, nil
}
// registerSmallTask registers the small task.
func (v *V1) registerSmallTask(ctx context.Context, peer *resource.Peer) (*schedulerv1.RegisterResult, error) {
candidateParents, found := v.scheduling.FindCandidateParents(ctx, peer, set.NewSafeSet[string]())
if !found {
return nil, errors.New("candidate parent not found")
}
candidateParent := candidateParents[0]
// When task size scope is small, parent must be downloaded successfully
// before returning to the parent directly.
if !candidateParent.FSM.Is(resource.PeerStateSucceeded) {
return nil, fmt.Errorf("candidate parent state is %s", candidateParent.FSM.Current())
}
piece, loaded := peer.Task.LoadPiece(0)
if !loaded {
return nil, fmt.Errorf("first piece not found")
}
// Delete inedges of peer.
if err := peer.Task.DeletePeerInEdges(peer.ID); err != nil {
return nil, err
}
// Add edges between parent and peer.
if err := peer.Task.AddPeerEdge(candidateParent, peer); err != nil {
return nil, err
}
if err := peer.FSM.Event(ctx, resource.PeerEventRegisterSmall); err != nil {
return nil, err
}
// Construct piece information.
pieceInfo := &commonv1.PieceInfo{
PieceNum: piece.Number,
RangeStart: piece.Offset,
RangeSize: uint32(piece.Length),
PieceOffset: piece.Offset,
// FIXME Use value of piece.PieceStyle.
PieceStyle: commonv1.PieceStyle_PLAIN,
DownloadCost: uint64(piece.Cost.Milliseconds()),
}
if piece.Digest != nil {
pieceInfo.PieceMd5 = piece.Digest.Encoded
}
return &schedulerv1.RegisterResult{
TaskId: peer.Task.ID,
TaskType: types.TaskTypeV2ToV1(peer.Task.Type),
SizeScope: commonv1.SizeScope_SMALL,
DirectPiece: &schedulerv1.RegisterResult_SinglePiece{
SinglePiece: &schedulerv1.SinglePiece{
DstPid: candidateParent.ID,
DstAddr: fmt.Sprintf("%s:%d", candidateParent.Host.IP, candidateParent.Host.DownloadPort),
PieceInfo: pieceInfo,
},
},
}, nil
}
// registerNormalTask registers the tiny task.
func (v *V1) registerNormalTask(ctx context.Context, peer *resource.Peer) (*schedulerv1.RegisterResult, error) {
if err := peer.FSM.Event(ctx, resource.PeerEventRegisterNormal); err != nil {
return nil, err
}
return &schedulerv1.RegisterResult{
TaskId: peer.Task.ID,
TaskType: types.TaskTypeV2ToV1(peer.Task.Type),
SizeScope: commonv1.SizeScope_NORMAL,
}, nil
}
// handleRegisterFailure handles failure of register.
func (v *V1) handleRegisterFailure(ctx context.Context, peer *resource.Peer) {
if err := peer.FSM.Event(ctx, resource.PeerEventLeave); err != nil {
peer.Log.Error(err)
}
v.resource.PeerManager().Delete(peer.ID)
return
}
// handleBeginOfPiece handles begin of piece.
func (v *V1) handleBeginOfPiece(ctx context.Context, peer *resource.Peer) {
state := peer.FSM.Current()
peer.Log.Infof("peer state is %s", state)
switch state {
case resource.PeerStateBackToSource:
// Back to the source download process, peer directly returns.
return
case resource.PeerStateReceivedTiny:
// When the task is tiny,
// the peer has already returned to piece data when registering.
if err := peer.FSM.Event(ctx, resource.PeerEventDownload); err != nil {
peer.Log.Errorf("peer fsm event failed: %s", err.Error())
return
}
case resource.PeerStateReceivedSmall:
// When the task is small,
// the peer has already returned to the parent when registering.
if err := peer.FSM.Event(ctx, resource.PeerEventDownload); err != nil {
peer.Log.Errorf("peer fsm event failed: %s", err.Error())
return
}
case resource.PeerStateReceivedNormal:
if err := peer.FSM.Event(ctx, resource.PeerEventDownload); err != nil {
peer.Log.Errorf("peer fsm event failed: %s", err.Error())
return
}
v.scheduling.ScheduleParentAndCandidateParents(ctx, peer, set.NewSafeSet[string]())
default:
}
}
// handleEndOfPiece handles end of piece.
func (v *V1) handleEndOfPiece(ctx context.Context, peer *resource.Peer) {}
// handlePieceSuccess handles successful piece.
func (v *V1) handlePieceSuccess(ctx context.Context, peer *resource.Peer, pieceResult *schedulerv1.PieceResult) {
// Distinguish traffic type.
trafficType := commonv2.TrafficType_REMOTE_PEER
if resource.IsPieceBackToSource(pieceResult.DstPid) {
trafficType = commonv2.TrafficType_BACK_TO_SOURCE
}
// Construct piece.
cost := time.Duration(int64(pieceResult.PieceInfo.DownloadCost) * int64(time.Millisecond))
piece := &resource.Piece{
Number: pieceResult.PieceInfo.PieceNum,
ParentID: pieceResult.DstPid,
Offset: pieceResult.PieceInfo.RangeStart,
Length: uint64(pieceResult.PieceInfo.RangeSize),
TrafficType: trafficType,