[FLINK-38858] Reuse jobid after failed submissions to guard against job duplication

[gyfora]

What is the purpose of the change

Currently some session job submission errors such as timeouts can very easily lead to duplicate submissions as jobids are incorrectly rotated even if the operator assumes that submission failed.

Brief change log

• Do not rotate job ids when submission failed
• Separate client and job submission timeout with larger default value
• Unit tests

Verifying this change

Unit tested + tested in prod environments

Does this pull request potentially affect one of the following parts:

• Dependencies (does it add or upgrade a dependency): no
• The public API, i.e., is any changes to the CustomResourceDescriptors: no
• Core observer or reconciler logic that is regularly executed: yes

Documentation

• Does this pull request introduce a new feature? no

[csviri]

LGTM

[gaborgsomogyi]

LGTM