Skip to content

delete all pods and services when job is suspended #7

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
KunWuLuan wants to merge 1 commit into
base: aliyun_v0.1.0
Choose a base branch
from
Open

delete all pods and services when job is suspended #7

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
16 changes: 8 additions & 8 deletions pkg/controller.v1/pytorch/controller.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -264,13 +264,11 @@ func (pc *PyTorchController) processNextWorkItem() bool {
}

// Wait until queuing annotation is removed
if pytorchJob.Annotations != nil {
if suspend, exist := pytorchJob.Annotations[suspendInQueue]; exist && suspend == "true" {
infoMsg := fmt.Sprintf("Annotation %s is found, operator will not process until removed", suspendInQueue)
logger.Info(infoMsg)
return true
}
}
// if isSuspend(pytorchJob) {
// infoMsg := fmt.Sprintf("Annotation %s is found, operator will not process until removed", suspendInQueue)
// logger.Info(infoMsg)
// return true
// }

// Sync PyTorchJob to mapch the actual state to this desired state.
forget, err := pc.syncHandler(key)
Expand Down Expand Up @@ -373,7 +371,9 @@ func (pc *PyTorchController) reconcilePyTorchJobs(job *pyv1.PyTorchJob) error {
}

// If the PyTorchJob is terminated, delete all pods and services.
if isSucceeded(job.Status) || isFailed(job.Status) {
if isSucceeded(job.Status) || isFailed(job.Status) || isSuspend(job) {
logger.Infof("job %v is inactive(maybe in succeeed, failed or suspend status)", job.Name)

if err := pc.deletePodsAndServices(job, pods, services); err != nil {
return err
}
Expand Down
6 changes: 4 additions & 2 deletions pkg/controller.v1/pytorch/job.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,11 +2,12 @@ package pytorch

import (
"fmt"
"github.com/kubeflow/pytorch-operator/pkg/util"
"reflect"
"strings"
"time"

"github.com/kubeflow/pytorch-operator/pkg/util"

log "github.com/sirupsen/logrus"
v1 "k8s.io/api/core/v1"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
Expand Down Expand Up @@ -159,8 +160,9 @@ func (pc *PyTorchController) deletePodsAndServices(job *pyv1.PyTorchJob, pods []
return nil
}

isSuspend := isSuspend(job)
// Delete nothing when the cleanPodPolicy is None.
if *job.Spec.CleanPodPolicy == common.CleanPodPolicyNone {
if *job.Spec.CleanPodPolicy == common.CleanPodPolicyNone && !isSuspend {
return nil
}

Expand Down
13 changes: 11 additions & 2 deletions pkg/controller.v1/pytorch/status.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,12 +20,12 @@ import (
"fmt"
"time"

"k8s.io/api/core/v1"
v1 "k8s.io/api/core/v1"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
utilruntime "k8s.io/apimachinery/pkg/util/runtime"

pyv1 "github.com/kubeflow/pytorch-operator/pkg/apis/pytorch/v1"
common "github.com/kubeflow/common/job_controller/api/v1"
pyv1 "github.com/kubeflow/pytorch-operator/pkg/apis/pytorch/v1"
pylogger "github.com/kubeflow/tf-operator/pkg/logger"
"github.com/prometheus/client_golang/prometheus"
"github.com/prometheus/client_golang/prometheus/promauto"
Expand Down Expand Up @@ -212,6 +212,15 @@ func hasCondition(status common.JobStatus, condType common.JobConditionType) boo
return false
}

func isSuspend(pytorchJob *pyv1.PyTorchJob) bool {
if pytorchJob.Annotations != nil {
if suspend, exist := pytorchJob.Annotations[suspendInQueue]; exist && suspend == "true" {
return true
}
}
return false
}

func isSucceeded(status common.JobStatus) bool {
return hasCondition(status, common.JobSucceeded)
}
Expand Down