-
Bug
-
Resolution: Unresolved
-
Major
-
rhos-18.0 FR 2 (Mar 2025)
-
None
-
5
-
False
-
-
False
-
?
-
rhos-ops-day1day2-upgrades
-
None
-
-
-
Pending Compose, RHOS Upgrades 2025 Sprint 12
-
2
-
Important
To Reproduce Steps to reproduce the behavior:
- Update guide chapter 3 recommends checking logs of
oc logs jobs/reboot-os-<deployment_name>-<nodeSet_name>to see whether a reboot is required on data plane nodes. - However, that job is not created/refreshed during a minor update procedure. It is only created when an EDPM deployment includes the "reboot-os" service, which is not the case during a minor update.
Expected behavior
- We should decide whether to reboot or not based on up-to-date info after the minor update procedure.
Bug impact
- Checking the logs can give an impression that a reboot is not required, while in fact a reboot can be required.
Known workaround
- A separate deployment could be created with the "reboot-os" EDPM service to while setting ansible variable `edpm_reboot_strategy: never` to report up-to-date information whether a reboot is needed.
Additional notes
- In the upcoming split update procedure, the reboot check should be part of the optional operating system update. If the admin chooses to use custom tooling for updating the operating system, they will need to do needs-reboot information tracking in a custom way too.
- depends on
-
OSPRH-20382 During edpm update run bootc variable are undefined.
-
- Review
-
- links to