When a Job is manually started via the OpenShift Console's using "Start Job" functionality, its spec.backoffLimit is unexpectedly set to 6, even if the underlying Job definition or template is configured with backoffLimit: 0. 

In contrast, when the same Job is manually initiated via the oc CLI, it correctly adheres to the backoffLimit: 0 specified in its configuration.

This behavior creates an inconsistency, where the console's manual trigger overrides explicit configuration.

    4.19.Z

    100%

    1. Manually start the Job via the OpenShift Console. 
       Go to "Workloads" -> "Jobs" and click the "Start Job" 

    2. Note the name of the new Job run generated by the console and check the backofflimit paramater. (It is giving 6)
     
    

    The newly created Job run shows backoffLimit: 6, overriding the configured in the base definition.

    It should obey the defined backofflimit in the job. 

- From the CLI it is working as expected --> whatever the backofflimit is define job is obeying that

- From the console it is getting "backofflimit: 6"

1. Job created manually via cli.
 
$ oc get job image-pruner-29210400 -o yaml | grep -i backofflimit
  backoffLimit: 0


2. Job created manually via console

$ oc get job image-pruner-1752640975985 -o yaml | grep -i backofflimit
  backoffLimit: 6