In my kubernetes job, I wrote a wrong docker name, and I find that my job keeps in a state of ImagePullBackOff. I want the job release resources, and stop pulling docker image since the image name is wrong. I noticed that there is a imagePullPolicy configuration, but seems does not meet my request. I want the job always pull docker images, but has a retry count limit for pulling. Does kubernetes support this action?
You should use backoffLimit and activeDeadlineSeconds as generic controller to quit the job after specified number of failures.
Example
apiVersion: batch/v1
kind: Job
metadata:
name: myjob
namespace: default
spec:
backoffLimit: 5
activeDeadlineSeconds: 100
template:
metadata:
name: myjob
spec:
containers:
- name: nginx
image: wrong_image_name
restartPolicy: OnFailureBy setting the
activeDeadlineSecondsfield of the Job to a number of seconds. TheactiveDeadlineSecondsapplies to the duration of the job, no matter how many Pods are created.Once a Job reaches
activeDeadlineSeconds, all of its running Pods are terminated and the Job status will become type:Failed with reason: DeadlineExceeded