-
Notifications
You must be signed in to change notification settings - Fork 1.3k
Expand file tree
/
Copy pathlifecycle_executor.go
More file actions
716 lines (643 loc) · 26.7 KB
/
Copy pathlifecycle_executor.go
File metadata and controls
716 lines (643 loc) · 26.7 KB
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
package autobuild
import (
"context"
"database/sql"
"fmt"
"net/http"
"slices"
"strings"
"sync"
"sync/atomic"
"time"
"github.com/dustin/go-humanize"
"github.com/google/uuid"
"github.com/prometheus/client_golang/prometheus"
"github.com/prometheus/client_golang/prometheus/promauto"
"golang.org/x/sync/errgroup"
"golang.org/x/xerrors"
"cdr.dev/slog/v3"
"github.com/coder/coder/v2/coderd/audit"
"github.com/coder/coder/v2/coderd/database"
"github.com/coder/coder/v2/coderd/database/dbauthz"
"github.com/coder/coder/v2/coderd/database/dbtime"
"github.com/coder/coder/v2/coderd/database/provisionerjobs"
"github.com/coder/coder/v2/coderd/database/pubsub"
"github.com/coder/coder/v2/coderd/files"
"github.com/coder/coder/v2/coderd/notifications"
"github.com/coder/coder/v2/coderd/pproflabel"
"github.com/coder/coder/v2/coderd/provisionerdserver"
"github.com/coder/coder/v2/coderd/schedule"
"github.com/coder/coder/v2/coderd/wsbuilder"
"github.com/coder/coder/v2/codersdk"
)
// Executor automatically starts or stops workspaces.
type Executor struct {
ctx context.Context
db database.Store
ps pubsub.Pubsub
fileCache *files.Cache
templateScheduleStore *atomic.Pointer[schedule.TemplateScheduleStore]
accessControlStore *atomic.Pointer[dbauthz.AccessControlStore]
auditor *atomic.Pointer[audit.Auditor]
buildUsageChecker *atomic.Pointer[wsbuilder.UsageChecker]
log slog.Logger
tick <-chan time.Time
statsCh chan<- Stats
// NotificationsEnqueuer handles enqueueing notifications for delivery by SMTP, webhook, etc.
notificationsEnqueuer notifications.Enqueuer
reg prometheus.Registerer
experiments codersdk.Experiments
workspaceBuilderMetrics *wsbuilder.Metrics
metrics executorMetrics
}
type executorMetrics struct {
autobuildExecutionDuration prometheus.Histogram
}
// Stats contains information about one run of Executor.
type Stats struct {
Transitions map[uuid.UUID]database.WorkspaceTransition
Elapsed time.Duration
Errors map[uuid.UUID]error
}
// New returns a new wsactions executor.
func NewExecutor(ctx context.Context, db database.Store, ps pubsub.Pubsub, fc *files.Cache, reg prometheus.Registerer, tss *atomic.Pointer[schedule.TemplateScheduleStore], auditor *atomic.Pointer[audit.Auditor], acs *atomic.Pointer[dbauthz.AccessControlStore], buildUsageChecker *atomic.Pointer[wsbuilder.UsageChecker], log slog.Logger, tick <-chan time.Time, enqueuer notifications.Enqueuer, exp codersdk.Experiments, workspaceBuilderMetrics *wsbuilder.Metrics) *Executor {
factory := promauto.With(reg)
le := &Executor{
//nolint:gocritic // Autostart has a limited set of permissions.
ctx: dbauthz.AsAutostart(ctx),
db: db,
ps: ps,
fileCache: fc,
templateScheduleStore: tss,
tick: tick,
log: log.Named("autobuild"),
auditor: auditor,
accessControlStore: acs,
buildUsageChecker: buildUsageChecker,
notificationsEnqueuer: enqueuer,
reg: reg,
experiments: exp,
workspaceBuilderMetrics: workspaceBuilderMetrics,
metrics: executorMetrics{
autobuildExecutionDuration: factory.NewHistogram(prometheus.HistogramOpts{
Namespace: "coderd",
Subsystem: "lifecycle",
Name: "autobuild_execution_duration_seconds",
Help: "Duration of each autobuild execution.",
Buckets: prometheus.DefBuckets,
}),
},
}
return le
}
// WithStatsChannel will cause Executor to push a RunStats to ch after
// every tick.
func (e *Executor) WithStatsChannel(ch chan<- Stats) *Executor {
e.statsCh = ch
return e
}
// Run will cause executor to start or stop workspaces on every
// tick from its channel. It will stop when its context is Done, or when
// its channel is closed.
func (e *Executor) Run() {
pproflabel.Go(e.ctx, pproflabel.Service(pproflabel.ServiceLifecycles), func(ctx context.Context) {
for {
select {
case <-ctx.Done():
return
case t, ok := <-e.tick:
if !ok {
return
}
stats := e.runOnce(t)
e.metrics.autobuildExecutionDuration.Observe(stats.Elapsed.Seconds())
if e.statsCh != nil {
select {
case <-ctx.Done():
return
case e.statsCh <- stats:
}
}
e.log.Debug(ctx, "run stats", slog.F("elapsed", stats.Elapsed), slog.F("transitions", stats.Transitions))
}
}
})
}
// hasValidProvisioner checks whether there is at least one valid (non-stale, correct tags) provisioner
// based on time t and the tags maps (such as from a templateVersionJob).
func (e *Executor) hasValidProvisioner(ctx context.Context, tx database.Store, t time.Time, ws database.Workspace, tags map[string]string) (bool, error) {
queryParams := database.GetProvisionerDaemonsByOrganizationParams{
OrganizationID: ws.OrganizationID,
WantTags: tags,
}
// nolint: gocritic // The user (in this case, the user/context for autostart builds) may not have the full
// permissions to read provisioner daemons, but we need to check if there's any for the job prior to the
// execution of the job via autostart to fix: https://github.com/coder/coder/issues/17941
provisionerDaemons, err := tx.GetProvisionerDaemonsByOrganization(dbauthz.AsSystemReadProvisionerDaemons(ctx), queryParams)
if err != nil {
return false, xerrors.Errorf("get provisioner daemons: %w", err)
}
logger := e.log.With(slog.F("tags", tags))
// Check if any provisioners are active (not stale)
for _, pd := range provisionerDaemons {
if pd.LastSeenAt.Valid {
age := t.Sub(pd.LastSeenAt.Time)
if age <= provisionerdserver.StaleInterval {
logger.Debug(ctx, "hasValidProvisioner: found active provisioner",
slog.F("daemon_id", pd.ID),
)
return true, nil
}
}
}
logger.Debug(ctx, "hasValidProvisioner: no active provisioners found")
return false, nil
}
func (e *Executor) runOnce(t time.Time) Stats {
stats := Stats{
Transitions: make(map[uuid.UUID]database.WorkspaceTransition),
Errors: make(map[uuid.UUID]error),
}
// we build the map of transitions concurrently, so need a mutex to serialize writes to the map
statsMu := sync.Mutex{}
defer func() {
stats.Elapsed = time.Since(t)
}()
currentTick := t.Truncate(time.Minute)
// TTL is set at the workspace level, and deadline at the workspace build level.
// When a workspace build is created, its deadline initially starts at zero.
// When provisionerd successfully completes a provision job, the deadline is
// set to now + TTL if the associated workspace has a TTL set. This deadline
// is what we compare against when performing autostop operations, rounded down
// to the minute.
//
// NOTE: If a workspace build is created with a given TTL and then the user either
// changes or unsets the TTL, the deadline for the workspace build will not
// have changed. This behavior is as expected per #2229.
workspaces, err := e.db.GetWorkspacesEligibleForTransition(e.ctx, currentTick)
if err != nil {
e.log.Error(e.ctx, "get workspaces for autostart or autostop", slog.Error(err))
return stats
}
// Sort the workspaces by build template version ID so that we can group
// identical template versions together. This is a slight (and imperfect)
// optimization.
//
// `wsbuilder` needs to load the terraform files for a given template version
// into memory. If 2 workspaces are using the same template version, they will
// share the same files in the FileCache. This only happens if the builds happen
// in parallel.
// TODO: Actually make sure the cache has the files in the cache for the full
// set of identical template versions. Then unload the files when the builds
// are done. Right now, this relies on luck for the 10 goroutine workers to
// overlap and keep the file reference in the cache alive.
slices.SortFunc(workspaces, func(a, b database.GetWorkspacesEligibleForTransitionRow) int {
return strings.Compare(a.BuildTemplateVersionID.UUID.String(), b.BuildTemplateVersionID.UUID.String())
})
// We only use errgroup here for convenience of API, not for early
// cancellation. This means we only return nil errors in th eg.Go.
eg := errgroup.Group{}
// Limit the concurrency to avoid overloading the database.
eg.SetLimit(10)
for _, ws := range workspaces {
wsID := ws.ID
wsName := ws.Name
log := e.log.With(
slog.F("workspace_id", wsID),
slog.F("workspace_name", wsName),
)
eg.Go(func() error {
err := func() error {
var (
job *database.ProvisionerJob
auditLog *auditParams
shouldNotifyDormancy bool
shouldNotifyTaskPause bool
nextBuild *database.WorkspaceBuild
activeTemplateVersion database.TemplateVersion
ws database.Workspace
tmpl database.Template
didAutoUpdate bool
)
err := e.db.InTx(func(tx database.Store) error {
var err error
ok, err := tx.TryAcquireLock(e.ctx, database.GenLockID(fmt.Sprintf("lifecycle-executor:%s", wsID)))
if err != nil {
return xerrors.Errorf("try acquire lifecycle executor lock: %w", err)
}
if !ok {
log.Debug(e.ctx, "unable to acquire lock for workspace, skipping")
return nil
}
// Re-check eligibility since the first check was outside the
// transaction and the workspace settings may have changed.
ws, err = tx.GetWorkspaceByID(e.ctx, wsID)
if err != nil {
return xerrors.Errorf("get workspace by id: %w", err)
}
user, err := tx.GetUserByID(e.ctx, ws.OwnerID)
if err != nil {
return xerrors.Errorf("get user by id: %w", err)
}
// Determine the workspace state based on its latest build.
latestBuild, err := tx.GetLatestWorkspaceBuildByWorkspaceID(e.ctx, ws.ID)
if err != nil {
return xerrors.Errorf("get latest workspace build: %w", err)
}
latestJob, err := tx.GetProvisionerJobByID(e.ctx, latestBuild.JobID)
if err != nil {
return xerrors.Errorf("get latest provisioner job: %w", err)
}
templateSchedule, err := (*(e.templateScheduleStore.Load())).Get(e.ctx, tx, ws.TemplateID)
if err != nil {
return xerrors.Errorf("get template scheduling options: %w", err)
}
// If next start at is not valid we need to re-compute it
if !ws.NextStartAt.Valid && ws.AutostartSchedule.Valid {
next, err := schedule.NextAllowedAutostart(currentTick, ws.AutostartSchedule.String, templateSchedule)
if err == nil {
nextStartAt := sql.NullTime{Valid: true, Time: dbtime.Time(next.UTC())}
if err = tx.UpdateWorkspaceNextStartAt(e.ctx, database.UpdateWorkspaceNextStartAtParams{
ID: wsID,
NextStartAt: nextStartAt,
}); err != nil {
return xerrors.Errorf("update workspace next start at: %w", err)
}
// Save re-fetching the workspace
ws.NextStartAt = nextStartAt
}
}
tmpl, err = tx.GetTemplateByID(e.ctx, ws.TemplateID)
if err != nil {
return xerrors.Errorf("get template by ID: %w", err)
}
activeTemplateVersion, err = tx.GetTemplateVersionByID(e.ctx, tmpl.ActiveVersionID)
if err != nil {
return xerrors.Errorf("get active template version by ID: %w", err)
}
accessControl := (*(e.accessControlStore.Load())).GetTemplateAccessControl(tmpl)
nextTransition, reason, err := getNextTransition(user, ws, latestBuild, latestJob, templateSchedule, currentTick)
if err != nil {
log.Debug(e.ctx, "skipping workspace", slog.Error(err))
// err is used to indicate that a workspace is not eligible
// so returning nil here is ok although ultimately the distinction
// doesn't matter since the transaction is read-only up to
// this point.
return nil
}
if reason == database.BuildReasonTaskAutoPause {
shouldNotifyTaskPause = true
}
// Get the template version job to access tags
templateVersionJob, err := tx.GetProvisionerJobByID(e.ctx, activeTemplateVersion.JobID)
if err != nil {
return xerrors.Errorf("get template version job: %w", err)
}
// Before creating the workspace build, check for available provisioners
hasProvisioners, err := e.hasValidProvisioner(e.ctx, tx, t, ws, templateVersionJob.Tags)
if err != nil {
return xerrors.Errorf("check provisioner availability: %w", err)
}
if !hasProvisioners {
log.Warn(e.ctx, "skipping autostart - no available provisioners")
return nil // Skip this workspace
}
if nextTransition != "" {
builder := wsbuilder.New(ws, nextTransition, *e.buildUsageChecker.Load()).
SetLastWorkspaceBuildInTx(&latestBuild).
SetLastWorkspaceBuildJobInTx(&latestJob).
Experiments(e.experiments).
Reason(reason).
BuildMetrics(e.workspaceBuilderMetrics)
log.Debug(e.ctx, "auto building workspace", slog.F("transition", nextTransition))
if nextTransition == database.WorkspaceTransitionStart &&
useActiveVersion(accessControl, ws) {
log.Debug(e.ctx, "autostarting with active version")
builder = builder.ActiveVersion()
if latestBuild.TemplateVersionID != tmpl.ActiveVersionID {
// control flag to know if the workspace was auto-updated,
// so the lifecycle executor can notify the user
didAutoUpdate = true
}
}
nextBuild, job, _, err = builder.Build(e.ctx, tx, e.fileCache, nil, audit.WorkspaceBuildBaggage{IP: "127.0.0.1"})
if err != nil {
return xerrors.Errorf("build workspace with transition %q: %w", nextTransition, err)
}
}
// Transition the workspace to dormant if it has breached the template's
// threshold for inactivity.
if reason == database.BuildReasonDormancy {
wsOld := ws
wsNew, err := tx.UpdateWorkspaceDormantDeletingAt(e.ctx, database.UpdateWorkspaceDormantDeletingAtParams{
ID: ws.ID,
DormantAt: sql.NullTime{
Time: dbtime.Now(),
Valid: true,
},
})
if err != nil {
return xerrors.Errorf("update workspace dormant deleting at: %w", err)
}
auditLog = &auditParams{
Old: wsOld.WorkspaceTable(),
New: wsNew,
}
// To keep the `ws` accurate without doing a sql fetch
ws.DormantAt = wsNew.DormantAt
shouldNotifyDormancy = true
log.Info(e.ctx, "dormant workspace",
slog.F("last_used_at", ws.LastUsedAt),
slog.F("time_til_dormant", templateSchedule.TimeTilDormant),
slog.F("since_last_used_at", time.Since(ws.LastUsedAt)),
)
}
if reason == database.BuildReasonAutodelete {
log.Info(e.ctx, "deleted workspace",
slog.F("dormant_at", ws.DormantAt.Time),
slog.F("time_til_dormant_autodelete", templateSchedule.TimeTilDormantAutoDelete),
)
}
if nextTransition == "" {
return nil
}
statsMu.Lock()
stats.Transitions[ws.ID] = nextTransition
statsMu.Unlock()
log.Info(e.ctx, "scheduling workspace transition",
slog.F("transition", nextTransition),
slog.F("reason", reason),
)
return nil
// Run with RepeatableRead isolation so that the build process sees the same data
// as our calculation that determines whether an autobuild is necessary.
}, &database.TxOptions{
Isolation: sql.LevelRepeatableRead,
TxIdentifier: "lifecycle",
})
if auditLog != nil {
// If the transition didn't succeed then updating the workspace
// to indicate dormant didn't either.
auditLog.Success = err == nil
auditBuild(e.ctx, log, *e.auditor.Load(), *auditLog)
}
if didAutoUpdate && err == nil {
nextBuildReason := ""
if nextBuild != nil {
nextBuildReason = string(nextBuild.Reason)
}
templateVersionMessage := activeTemplateVersion.Message
if templateVersionMessage == "" {
templateVersionMessage = "None provided"
}
if _, err := e.notificationsEnqueuer.Enqueue(e.ctx, ws.OwnerID, notifications.TemplateWorkspaceAutoUpdated,
map[string]string{
"name": ws.Name,
"initiator": "autobuild",
"reason": nextBuildReason,
"template_version_name": activeTemplateVersion.Name,
"template_version_message": templateVersionMessage,
}, "autobuild",
// Associate this notification with all the related entities.
ws.ID, ws.OwnerID, ws.TemplateID, ws.OrganizationID,
); err != nil {
log.Warn(e.ctx, "failed to notify of autoupdated workspace", slog.Error(err))
}
}
if err != nil {
return xerrors.Errorf("transition workspace: %w", err)
}
if job != nil {
// Note that we can't refactor such that posting the job happens inside wsbuilder because it's called
// with an outer transaction like this, and we need to make sure the outer transaction commits before
// posting the job. If we post before the transaction commits, provisionerd might try to acquire the
// job, fail, and then sit idle instead of picking up the job.
err = provisionerjobs.PostJob(e.ps, *job)
if err != nil {
return xerrors.Errorf("post provisioner job to pubsub: %w", err)
}
}
if shouldNotifyDormancy {
dormantTime := dbtime.Now().Add(time.Duration(tmpl.TimeTilDormant))
_, err = e.notificationsEnqueuer.Enqueue(
e.ctx,
ws.OwnerID,
notifications.TemplateWorkspaceDormant,
map[string]string{
"name": ws.Name,
"reason": "inactivity exceeded the dormancy threshold",
"timeTilDormant": humanize.Time(dormantTime),
},
"lifecycle_executor",
ws.ID,
ws.OwnerID,
ws.TemplateID,
ws.OrganizationID,
)
if err != nil {
log.Warn(e.ctx, "failed to notify of workspace marked as dormant", slog.Error(err), slog.F("workspace_id", ws.ID))
}
}
if shouldNotifyTaskPause {
task, err := e.db.GetTaskByID(e.ctx, ws.TaskID.UUID)
if err != nil {
log.Warn(e.ctx, "failed to get task for pause notification", slog.Error(err), slog.F("task_id", ws.TaskID.UUID), slog.F("workspace_id", ws.ID))
} else {
if _, err := e.notificationsEnqueuer.Enqueue(
e.ctx,
ws.OwnerID,
notifications.TemplateTaskPaused,
map[string]string{
"task": task.Name,
"task_id": task.ID.String(),
"workspace": ws.Name,
"pause_reason": "idle timeout",
},
"lifecycle_executor",
ws.ID, ws.OwnerID, ws.OrganizationID,
); err != nil {
log.Warn(e.ctx, "failed to notify of task paused", slog.Error(err), slog.F("task_id", ws.TaskID.UUID), slog.F("workspace_id", ws.ID))
}
}
}
return nil
}()
if err != nil && !xerrors.Is(err, context.Canceled) {
log.Error(e.ctx, "failed to transition workspace", slog.Error(err))
statsMu.Lock()
stats.Errors[wsID] = err
statsMu.Unlock()
}
// Even though we got an error we still return nil to avoid
// short-circuiting the evaluation loop.
return nil
})
}
// This should not happen since we don't want early cancellation.
err = eg.Wait()
if err != nil {
e.log.Error(e.ctx, "workspace scheduling errgroup failed", slog.Error(err))
}
return stats
}
// getNextTransition returns the next eligible transition for the workspace
// as well as the reason for why it is transitioning. It is possible
// for this function to return a nil error as well as an empty transition.
// In such cases it means no provisioning should occur but the workspace
// may be "transitioning" to a new state (such as an inactive, stopped
// workspace transitioning to the dormant state).
func getNextTransition(
user database.User,
ws database.Workspace,
latestBuild database.WorkspaceBuild,
latestJob database.ProvisionerJob,
templateSchedule schedule.TemplateScheduleOptions,
currentTick time.Time,
) (
database.WorkspaceTransition,
database.BuildReason,
error,
) {
switch {
case isEligibleForAutostop(user, ws, latestBuild, latestJob, currentTick):
// Use task-specific reason for AI task workspaces.
if ws.TaskID.Valid {
return database.WorkspaceTransitionStop, database.BuildReasonTaskAutoPause, nil
}
return database.WorkspaceTransitionStop, database.BuildReasonAutostop, nil
case isEligibleForAutostart(user, ws, latestBuild, latestJob, templateSchedule, currentTick):
return database.WorkspaceTransitionStart, database.BuildReasonAutostart, nil
case isEligibleForFailedStop(latestBuild, latestJob, templateSchedule, currentTick):
// Use task-specific reason for AI task workspaces.
if ws.TaskID.Valid {
return database.WorkspaceTransitionStop, database.BuildReasonTaskAutoPause, nil
}
return database.WorkspaceTransitionStop, database.BuildReasonAutostop, nil
case isEligibleForDormantStop(ws, templateSchedule, currentTick):
// Only stop started workspaces.
if latestBuild.Transition == database.WorkspaceTransitionStart {
return database.WorkspaceTransitionStop, database.BuildReasonDormancy, nil
}
// We shouldn't transition the workspace but we should still
// make it dormant.
return "", database.BuildReasonDormancy, nil
case isEligibleForDelete(ws, templateSchedule, latestBuild, latestJob, currentTick):
return database.WorkspaceTransitionDelete, database.BuildReasonAutodelete, nil
default:
return "", "", xerrors.Errorf("last transition not valid for autostart or autostop")
}
}
// isEligibleForAutostart returns true if the workspace should be autostarted.
func isEligibleForAutostart(user database.User, ws database.Workspace, build database.WorkspaceBuild, job database.ProvisionerJob, templateSchedule schedule.TemplateScheduleOptions, currentTick time.Time) bool {
// Don't attempt to autostart workspaces for suspended users.
if user.Status != database.UserStatusActive {
return false
}
// Don't attempt to autostart failed workspaces.
if job.JobStatus == database.ProvisionerJobStatusFailed {
return false
}
// If the workspace is dormant we should not autostart it.
if ws.DormantAt.Valid {
return false
}
// If the last transition for the workspace was not 'stop' then the workspace
// cannot be started.
if build.Transition != database.WorkspaceTransitionStop {
return false
}
// If autostart isn't enabled, or the schedule isn't valid/populated we can't
// autostart the workspace.
if !templateSchedule.UserAutostartEnabled || !ws.AutostartSchedule.Valid || ws.AutostartSchedule.String == "" {
return false
}
// Get the next allowed autostart time after the build's creation time,
// based on the workspace's schedule and the template's allowed days.
nextTransition, err := schedule.NextAllowedAutostart(build.CreatedAt, ws.AutostartSchedule.String, templateSchedule)
if err != nil {
return false
}
// Must use '.Before' vs '.After' so equal times are considered "valid for autostart".
return !currentTick.Before(nextTransition)
}
// isEligibleForAutostop returns true if the workspace should be autostopped.
func isEligibleForAutostop(user database.User, ws database.Workspace, build database.WorkspaceBuild, job database.ProvisionerJob, currentTick time.Time) bool {
if job.JobStatus == database.ProvisionerJobStatusFailed {
return false
}
// If the workspace is dormant we should not autostop it.
if ws.DormantAt.Valid {
return false
}
if build.Transition == database.WorkspaceTransitionStart && user.Status == database.UserStatusSuspended {
return true
}
// A workspace must be started in order for it to be auto-stopped.
return build.Transition == database.WorkspaceTransitionStart &&
!build.Deadline.IsZero() &&
// We do not want to stop a workspace prior to it breaching its deadline.
!currentTick.Before(build.Deadline)
}
// isEligibleForDormantStop returns true if the workspace should be dormant
// for breaching the inactivity threshold of the template.
func isEligibleForDormantStop(ws database.Workspace, templateSchedule schedule.TemplateScheduleOptions, currentTick time.Time) bool {
// Only attempt against workspaces not already dormant.
return !ws.DormantAt.Valid &&
// The template must specify an time_til_dormant value.
templateSchedule.TimeTilDormant > 0 &&
// The workspace must breach the time_til_dormant value.
currentTick.Sub(ws.LastUsedAt) > templateSchedule.TimeTilDormant
}
func isEligibleForDelete(ws database.Workspace, templateSchedule schedule.TemplateScheduleOptions, lastBuild database.WorkspaceBuild, lastJob database.ProvisionerJob, currentTick time.Time) bool {
eligible := ws.DormantAt.Valid && ws.DeletingAt.Valid &&
// Dormant workspaces should only be deleted if a time_til_dormant_autodelete value is specified.
templateSchedule.TimeTilDormantAutoDelete > 0 &&
// The workspace must breach the time_til_dormant_autodelete value.
currentTick.After(ws.DeletingAt.Time)
// If the last delete job failed we should wait 24 hours before trying again.
// Builds are resource-intensive so retrying every minute is not productive
// and will hold compute hostage.
if lastBuild.Transition == database.WorkspaceTransitionDelete && lastJob.JobStatus == database.ProvisionerJobStatusFailed {
return eligible && lastJob.Finished() && currentTick.Sub(lastJob.FinishedAt()) > time.Hour*24
}
return eligible
}
// isEligibleForFailedStop returns true if the workspace is eligible to be stopped
// due to a failed build.
func isEligibleForFailedStop(build database.WorkspaceBuild, job database.ProvisionerJob, templateSchedule schedule.TemplateScheduleOptions, currentTick time.Time) bool {
// If the template has specified a failure TLL.
return templateSchedule.FailureTTL > 0 &&
// And the job resulted in failure.
job.JobStatus == database.ProvisionerJobStatusFailed &&
build.Transition == database.WorkspaceTransitionStart &&
// And sufficient time has elapsed since the job has completed.
job.CompletedAt.Valid &&
currentTick.Sub(job.CompletedAt.Time) > templateSchedule.FailureTTL
}
type auditParams struct {
Old database.WorkspaceTable
New database.WorkspaceTable
Success bool
}
func auditBuild(ctx context.Context, log slog.Logger, auditor audit.Auditor, params auditParams) {
status := http.StatusInternalServerError
if params.Success {
status = http.StatusOK
}
audit.BackgroundAudit(ctx, &audit.BackgroundAuditParams[database.WorkspaceTable]{
Audit: auditor,
Log: log,
UserID: params.New.OwnerID,
OrganizationID: params.New.OrganizationID,
// Right now there's no request associated with an autobuild
// operation.
RequestID: uuid.Nil,
Action: database.AuditActionWrite,
Old: params.Old,
New: params.New,
Status: status,
})
}
func useActiveVersion(opts dbauthz.TemplateAccessControl, ws database.Workspace) bool {
return opts.RequireActiveVersion || ws.AutomaticUpdates == database.AutomaticUpdatesAlways
}