harness-drone/operator/manager/teardown.go

355 lines
8.3 KiB
Go
Raw Normal View History

2019-03-13 21:47:47 +00:00
// Copyright 2019 Drone IO, Inc.
//
// 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.
2019-02-19 23:56:41 +00:00
package manager
import (
"context"
"encoding/json"
"time"
"github.com/drone/drone/core"
2019-03-13 21:47:47 +00:00
"github.com/drone/drone/store/shared/db"
2019-02-19 23:56:41 +00:00
"github.com/drone/go-scm/scm"
"github.com/hashicorp/go-multierror"
"github.com/sirupsen/logrus"
)
type teardown struct {
Builds core.BuildStore
Events core.Pubsub
Logs core.LogStream
Scheduler core.Scheduler
Repos core.RepositoryStore
Steps core.StepStore
Status core.StatusService
Stages core.StageStore
Users core.UserStore
Webhook core.WebhookSender
2019-02-19 23:56:41 +00:00
}
func (t *teardown) do(ctx context.Context, stage *core.Stage) error {
logger := logrus.WithField("stage.id", stage.ID)
logger.Debugln("manager: stage is complete. teardown")
build, err := t.Builds.Find(noContext, stage.BuildID)
if err != nil {
logger.WithError(err).Warnln("manager: cannot find the build")
return err
}
logger = logger.WithFields(
logrus.Fields{
"build.number": build.Number,
"build.id": build.ID,
"repo.id": build.RepoID,
},
)
repo, err := t.Repos.Find(noContext, build.RepoID)
if err != nil {
logger.WithError(err).Warnln("manager: cannot find the repository")
return err
}
for _, step := range stage.Steps {
if len(step.Error) > 500 {
step.Error = step.Error[:500]
}
2019-02-19 23:56:41 +00:00
err := t.Steps.Update(noContext, step)
if err != nil {
logger.WithError(err).
WithField("stage.status", stage.Status).
WithField("step.name", step.Name).
WithField("step.id", step.ID).
Warnln("manager: cannot persist the step")
return err
}
}
if len(stage.Error) > 500 {
stage.Error = stage.Error[:500]
}
2019-02-19 23:56:41 +00:00
stage.Updated = time.Now().Unix()
err = t.Stages.Update(noContext, stage)
if err != nil {
logger.WithError(err).
Warnln("manager: cannot update the stage")
return err
}
for _, step := range stage.Steps {
t.Logs.Delete(noContext, step.ID)
}
stages, err := t.Stages.ListSteps(noContext, build.ID)
if err != nil {
logger.WithError(err).Warnln("manager: cannot get stages")
return err
}
2019-04-22 22:24:00 +00:00
//
//
//
2019-02-19 23:56:41 +00:00
err = t.cancelDownstream(ctx, stages)
if err != nil {
2019-04-22 22:24:00 +00:00
logger.WithError(err).
Errorln("manager: cannot cancel downstream builds")
2019-02-19 23:56:41 +00:00
return err
}
2019-04-22 22:24:00 +00:00
2019-02-19 23:56:41 +00:00
err = t.scheduleDownstream(ctx, stage, stages)
if err != nil {
2019-04-22 22:24:00 +00:00
logger.WithError(err).
Errorln("manager: cannot schedule downstream builds")
2019-02-19 23:56:41 +00:00
return err
}
2019-04-22 22:24:00 +00:00
//
//
//
2019-02-19 23:56:41 +00:00
if isBuildComplete(stages) == false {
logger.Debugln("manager: build pending completion of additional stages")
return nil
}
logger.Debugln("manager: build is finished, teardown")
build.Status = core.StatusPassing
build.Finished = time.Now().Unix()
for _, sibling := range stages {
if sibling.Status == core.StatusKilled {
build.Status = core.StatusKilled
break
}
if sibling.Status == core.StatusFailing {
build.Status = core.StatusFailing
break
}
if sibling.Status == core.StatusError {
build.Status = core.StatusError
break
}
}
2019-06-23 02:00:12 +00:00
if build.Started == 0 {
build.Started = build.Finished
}
2019-02-19 23:56:41 +00:00
err = t.Builds.Update(noContext, build)
if err == db.ErrOptimisticLock {
logger.WithError(err).
Warnln("manager: build updated by another goroutine")
return nil
}
if err != nil {
logger.WithError(err).
Warnln("manager: cannot update the build")
return err
}
repo.Build = build
repo.Build.Stages = stages
data, _ := json.Marshal(repo)
err = t.Events.Publish(noContext, &core.Message{
Repository: repo.Slug,
Visibility: repo.Visibility,
Data: data,
})
if err != nil {
logger.WithError(err).
Warnln("manager: cannot publish build event")
}
payload := &core.WebhookData{
Event: core.WebhookEventBuild,
Action: core.WebhookActionUpdated,
Repo: repo,
Build: build,
}
err = t.Webhook.Send(noContext, payload)
if err != nil {
logger.WithError(err).Warnln("manager: cannot send global webhook")
}
2019-02-19 23:56:41 +00:00
user, err := t.Users.Find(noContext, repo.UserID)
if err != nil {
logger.WithError(err).
Warnln("manager: cannot find repository owner")
2019-04-22 22:24:00 +00:00
// this error is insufficient to fail the function,
// however, execution of the function should be halted
// to prevent a nil pointer in subsequent operations.
return nil
2019-02-19 23:56:41 +00:00
}
req := &core.StatusInput{
Repo: repo,
Build: build,
}
err = t.Status.Send(noContext, user, req)
if err != nil && err != scm.ErrNotSupported {
logger.WithError(err).
Warnln("manager: cannot publish status")
}
return nil
}
// cancelDownstream is a helper function that tests for
// downstream stages and cancels them based on the overall
// pipeline state.
func (t *teardown) cancelDownstream(
ctx context.Context,
stages []*core.Stage,
) error {
failed := false
for _, s := range stages {
if s.IsFailed() {
failed = true
}
}
var errs error
for _, s := range stages {
if s.Status != core.StatusWaiting {
continue
}
2019-04-22 22:24:00 +00:00
var skip bool
if failed == true && s.OnFailure == false {
skip = true
}
if failed == false && s.OnSuccess == false {
skip = true
}
if skip == false {
2019-02-19 23:56:41 +00:00
continue
}
2019-04-22 22:24:00 +00:00
if areDepsComplete(s, stages) == false {
2019-02-19 23:56:41 +00:00
continue
}
logger := logrus.WithFields(
logrus.Fields{
"stage.id": s.ID,
"stage.on_success": s.OnSuccess,
"stage.on_failure": s.OnFailure,
"stage.is_failure": failed,
"stage.depends_on": s.DependsOn,
},
)
logger.Debugln("manager: skipping step")
s.Status = core.StatusSkipped
s.Started = time.Now().Unix()
s.Stopped = time.Now().Unix()
err := t.Stages.Update(noContext, s)
2019-04-22 22:24:00 +00:00
if err == db.ErrOptimisticLock {
t.resync(ctx, s)
continue
}
if err != nil {
2019-02-19 23:56:41 +00:00
logger.WithError(err).
Warnln("manager: cannot update stage status")
errs = multierror.Append(errs, err)
}
}
return errs
}
// scheduleDownstream is a helper function that tests for
// downstream stages and schedules stages if all dependencies
// and execution requirements are met.
func (t *teardown) scheduleDownstream(
ctx context.Context,
stage *core.Stage,
stages []*core.Stage,
) error {
var errs error
for _, sibling := range stages {
if sibling.Status == core.StatusWaiting {
if len(sibling.DependsOn) == 0 {
continue
}
2019-04-22 22:24:00 +00:00
// PROBLEM: isDep only checks the direct parent
// i think ....
// if isDep(stage, sibling) == false {
// continue
// }
2019-02-19 23:56:41 +00:00
if areDepsComplete(sibling, stages) == false {
continue
}
2019-04-22 22:24:00 +00:00
// if isLastDep(stage, sibling, stages) == false {
// continue
// }
2019-02-19 23:56:41 +00:00
logger := logrus.WithFields(
logrus.Fields{
"stage.id": sibling.ID,
"stage.name": sibling.Name,
"stage.depends_on": sibling.DependsOn,
},
)
logger.Debugln("manager: schedule next stage")
sibling.Status = core.StatusPending
sibling.Updated = time.Now().Unix()
err := t.Stages.Update(noContext, sibling)
2019-04-22 22:24:00 +00:00
if err == db.ErrOptimisticLock {
t.resync(ctx, sibling)
continue
}
2019-02-19 23:56:41 +00:00
if err != nil {
logger.WithError(err).
Warnln("manager: cannot update stage status")
errs = multierror.Append(errs, err)
}
err = t.Scheduler.Schedule(noContext, sibling)
if err != nil {
logger.WithError(err).
Warnln("manager: cannot schedule stage")
errs = multierror.Append(errs, err)
}
}
}
return errs
}
2019-04-22 22:24:00 +00:00
// resync updates the stage from the database. Note that it does
// not update the Version field. This is by design. It prevents
// the current go routine from updating a stage that has been
// updated by another go routine.
func (t *teardown) resync(ctx context.Context, stage *core.Stage) error {
updated, err := t.Stages.Find(ctx, stage.ID)
if err != nil {
return err
}
stage.Status = updated.Status
stage.Error = updated.Error
stage.ExitCode = updated.ExitCode
stage.Machine = updated.Machine
stage.Started = updated.Started
stage.Stopped = updated.Stopped
stage.Created = updated.Created
stage.Updated = updated.Updated
return nil
}