Hitting a bug and I want to diagnose whether Im hitting a memory issue, but the cloud run jobs/logs autodelete themselves on failure

You can set on the Work Pool or for a deployment in the

job_variables={"keep_job": True}