When a run terminates (successful or unsuccessful), Sematic cleans up the corresponding k8s job. Sometimes when diagnosing problems with k8s, it can be helpful to look at job/pod states even after termination. We should allow a feature flag whereby you can disable job cleanup for debugging.
When a run terminates (successful or unsuccessful), Sematic cleans up the corresponding k8s job. Sometimes when diagnosing problems with k8s, it can be helpful to look at job/pod states even after termination. We should allow a feature flag whereby you can disable job cleanup for debugging.