[Bug] [admin]When submitting the task, the jobmanager.archive.fs.dir was incorrectly set.

[jiangwwwei]

Search before asking

• I had searched in the issues and found no similar issues.

What happened

I configured this parameter in the cluster config, but when submitting the job, the parameter from the job config was retrieved and then directly assigned the constant address of the RS protocol.

When submitting to the cluster, the configuration jobmanager.archive.fs.dir = rs:/tmp/flink-job-archive was used. However, Flink itself does not support the RS protocol, resulting in an error that prevents the job from starting.

What you expected to happen

Independently configure the HistoryServer archive address.

How to reproduce

1.enabling the HistoryServer
2.do not configure jobmanager.archive.fs.dir in the task configuration;
3.instead, set jobmanager.archive.fs.dir in the cluster configuration.
4. Then, submit the job to the Kubernetes cluster.

Anything else

No response

Version

1.2.0

Are you willing to submit PR?

• Yes I am willing to submit a PR!

Code of Conduct

• I agree to follow this project's Code of Conduct

[github-actions]

Hello @jiangwwwei, this issue is about K8S, so I assign it to @gaoyan1998 and @zackyoungh. If you have any questions, you can comment and reply.

你好 @jiangwwwei, 这个 issue 是关于 K8S 的，所以我把它分配给了 @gaoyan1998 和 @zackyoungh。如有任何问题，可以评论回复。

[gaoyan1998]

Just go to the configuration center and turn off dinky's built-in history service.