We have a pod in kubernetes and it is processing some jobs in there regularly. Some of these jobs take more than 2 or 3 minutes. When deploying this pod, we don't want to kill current processes on this pod. How to keep wait this pod's termination until all processes completed? I read some topics about 'terminationGracePeriodSeconds'. But is there any upper seconds limit about that property or how it works?
You can use wait command, which takes multiple resources and waits until the specified condition.
For example, assume you create a job called worker that print the word 'printsth' and pause for 3 seconds ten times.
$ kubectl create ns waitplayground
$ kubectl -n waitplayground \
create job worker \
--image centos:7 -- \
sh -c \
'for i in {1..10} ; do echo printsth ; sleep 3; done'Kubectl wait command returns something similar to that:
$ kubectl -n waitplayground \
wait --for=condition=complete --timeout=40s \
job/worker
job.batch/worker condition metI hope it will helps you.