Last active
March 31, 2020 05:10
-
-
Save Jeffwan/d1a688a5bb85b684eed079fcd5f2a54e to your computer and use it in GitHub Desktop.
elastic_job.yaml
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
apiVersion: elastic.pytorch.org/v1alpha1 | |
kind: ElasticJob | |
metadata: | |
name: imagenet | |
namespace: elastic-job | |
spec: | |
# Use "etcd-service:2379" if you already apply etcd.yaml | |
rdzvEndpoint: "etcd-service:2379" | |
minReplicas: 1 | |
maxReplicas: 5 | |
replicaSpecs: | |
Worker: | |
replicas: 2 | |
restartPolicy: ExitCode | |
template: | |
apiVersion: v1 | |
kind: Pod | |
spec: | |
containers: | |
- name: elasticjob-worker | |
image: seedjeffwan/examples:0.1.0rc1 | |
imagePullPolicy: Always | |
args: | |
- "s3://torchelastic-shjiaxin-1h71m-s3bucket-m1b9b9pjldqw/petctl/shjiaxin/imagenet-job/main.py" | |
- "--input_path" | |
- "/data/tiny-imagenet-200/train" | |
- "--epochs" | |
- "10" | |
resources: | |
limits: | |
nvidia.com/gpu: 1 |
020-03-31T05:08:27.055Z DEBUG controller-runtime.controller Successfully Reconciled {"controller": "elasticjob", "request": "elastic-job/imagenet"}
time="2020-03-31T05:08:30Z" level=info msg="Reconciling for job imagenet"
time="2020-03-31T05:08:30Z" level=info msg="Ignoring inactive pod elastic-job/imagenet-worker-1 in state Failed, deletion time 2020-03-31 05:07:53 +0000 UTC"
time="2020-03-31T05:08:30Z" level=warning msg="The label index is not expected: 1, pod: elastic-job/imagenet-worker-1" job=elastic-job.imagenet replica-type=worker uid=24f9ba80-730c-11ea-8300-0e6ac959ddc0
time="2020-03-31T05:08:30Z" level=info msg="Deleting pod elastic-job/imagenet-worker-1, Job name: imagenet" job=elastic-job.imagenet uid=24f9ba80-730c-11ea-8300-0e6ac959ddc0
2020-03-31T05:08:30.805Z INFO controllers.ElasticJob Update on delete function: elastic-job-controller create object imagenet-worker-1
time="2020-03-31T05:08:30Z" level=warning msg="ReconcilePods error pods \"imagenet-worker-1\" not found"
time="2020-03-31T05:08:30Z" level=info msg="Reconcile ElasticJob error pods \"imagenet-worker-1\" not found" job=elastic-job.imagenet uid=24f9ba80-730c-11ea-8300-0e6ac959ddc0
2020-03-31T05:08:30.806Z ERROR controller-runtime.controller Reconciler error {"controller": "elasticjob", "request": "elastic-job/imagenet", "error": "pods \"imagenet-worker-1\" not found"}
github.com/go-logr/zapr.(*zapLogger).Error
/go/pkg/mod/github.com/go-logr/zapr@v0.1.0/zapr.go:128
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).reconcileHandler
/go/pkg/mod/sigs.k8s.io/controller-runtime@v0.3.0/pkg/internal/controller/controller.go:218
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).processNextWorkItem
/go/pkg/mod/sigs.k8s.io/controller-runtime@v0.3.0/pkg/internal/controller/controller.go:192
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).worker
/go/pkg/mod/sigs.k8s.io/controller-runtime@v0.3.0/pkg/internal/controller/controller.go:171
k8s.io/apimachinery/pkg/util/wait.JitterUntil.func1
/go/pkg/mod/k8s.io/apimachinery@v0.15.11-beta.0/pkg/util/wait/wait.go:152
k8s.io/apimachinery/pkg/util/wait.JitterUntil
/go/pkg/mod/k8s.io/apimachinery@v0.15.11-beta.0/pkg/util/wait/wait.go:153
k8s.io/apimachinery/pkg/util/wait.Until
/go/pkg/mod/k8s.io/apimachinery@v0.15.11-beta.0/pkg/util/wait/wait.go:88
2020-03-31T05:08:30.806Z DEBUG controller-runtime.manager.events Warning {"object": {"kind":"ElasticJob","namespace":"elastic-job","name":"imagenet","uid":"24f9ba80-730c-11ea-8300-0e6ac959ddc0","apiVersion":"elastic.pytorch.org/v1alpha1","resourceVersion":"4061536"}, "reason": "FailedDeletePod", "message": "Error deleting: pods \"imagenet-worker-1\" not found"}
time="2020-03-31T05:08:31Z" level=info msg="Reconciling for job imagenet"
time="2020-03-31T05:08:31Z" level=info msg="ElasticJob=imagenet, ReplicaType=Worker expected=1, running=1, succeeded=0 , failed=0" job=elastic-job.imagenet uid=24f9ba80-730c-11ea-8300-0e6ac959ddc0
time="2020-03-31T05:08:31Z" level=info msg="ElasticJob imagenet is running." job=elas
Why we delete pod and get event and reconcile it again? Did we miss the logic to handle terminating pods?
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
pytorch/elastic#63 (comment)