-
Bug
-
Resolution: Won't Do
-
Normal
-
None
-
4.16
-
None
-
Quality / Stability / Reliability
-
False
-
-
None
-
Important
-
None
-
All
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
-
None
Description of problem:
The `on-prem-resolv-prepender.service` remains in an "activating" state and keeps executing in a loop (`until` loop) even after `/usr/local/bin/resolv-prepender.sh` completes successfully (exit code 0) and the DNS is correctly set in `/etc/resolv.conf`.
Version-Release number of selected component (if applicable):
OpenShift Container Platform 4.16.33 Component: Machine Config Operator / platform-baremetal
How reproducible:
Always
Steps to Reproduce:
1. Install or upgrade to OCP 4.16 in a Bare Metal IPI environment where this systemd unit exists. 2. Allow the system to boot and check the status of `on-prem-resolv-prepender.service`. 3. Observe that the service remains in a looping "activating" state even if the script inside completes successfully.
Actual results:
The service keeps running unnecessarily in a bash loop, continuing to invoke podman and DNS configuration logic.
Expected results:
The service should stop once the script completes successfully with code 0, as no further action is needed.
Additional info:
- The service was originally introduced as a workaround for OCPBUGS-27162, which was fixed in OCP 4.16. - This behavior could consume unnecessary resources and mislead operators. - Recommend updating or disabling the unit post-upgrade, or refining the script logic to ensure a one-shot behavior.