Hi All, I wanted to ask if there is a way to retain the

high-availability

configmap if the job crashes? I am running flink version 1.17. Thanks

It should be retained always unless the job goes into a terminal Failed or Finished state

Could u explain, what you mean by terminal Failed?

Jobstate == Failed , It means there are no more retries

Yes, I am currently using

faliure-rate

restart strategy but I want to keep the HA configmaps if job goes into the Failed state.

This is how it works, if the restart strategy is exhausted job will be terminally failed and HA deleted. I suggest using some infinite restart strategy to avoid the deletion

I did change it to

fixed-delay

but its will still create an issue for us due to the nature of the job if the application keeps in restarting.

If HA will work until job is not going to terminal state, now question is, how should we recover flink state after terminal failed? Considering that we are using flink operator and upgradeMode: last-state. Also we enabled checkpointing.