-
Bug
-
Resolution: Cannot Reproduce
-
Major
-
None
-
4.11
-
Critical
-
None
-
Rejected
-
False
-
-
-
Rel Note for Telco: TBD
Description of problem:
During ZTP, when a supermicro server is used, server always boots into existing HD instead of CD, even though the supermicro maintenance logs shows the bootorder instance was set to CD.
Version-Release number of selected component (if applicable):
hub cluster ocp version: 4.11.0
ACM version: 2.5.2
gitops version: 1.6.0
How reproducible: 100%
Steps to Reproduce:
1. Create siteconfig for a supermicro server.
2. Trigger ZTP from hub cluster
3. Wait for bmh in provisioned state, then wait for server to boot into CD and start install.
Actual results:
3. BMH becomes provsioned, but server booted into HD. Although I can see the iso was successfully attached to virtual media and supermicro mtc events indicates the boot order instance was set to CD.
Expected results:
Supermicro server boots into CD
Additional info:
- Logs, boot video, bmh: http://registry.ran-vcl01.ptp.lab.eng.bos.redhat.com:8080/images/OCPBUGS-369/
- This issue did not happen on same hub cluster with other spoke servers, e.g., Dell, ZT.
- The same Supermicro server installed successfully from another hub cluster with following versions: OCP: 4.9.21, ACM: 2.4.5, gitops: 1.5.5.
- mentioned on
[OCPBUGS-369] Supermicro server booted into HD instead of CD in initial boot during ZTP
https://issues.redhat.com/browse/OCPBUGS-369?focusedId=21812242&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-21812242 was supposed to be a NEEDINFO, adding it now.
rhn-support-yliu1 are you able to provide this information?
Thank you and sorry for the mixup.
GitLab CEE Bot
added a comment - Yang Liu mentioned this issue in a merge request of ran / ran-integration on branch get_livecd:
get livecd
GitLab CEE Bot
added a comment - Yang Liu mentioned this issue in a merge request of ran / ran-integration on branch get_livecd : get livecd I haven't asked explicitly so asking now - prior to upgrading firmware, please record versions of:
BIOS
BMC
Redfish
I'd like to compare these with the X12 in another environment which is being provisioned successfully without encountering this issue.
janders@redhat.com bootMode is set to UEFI. You can see the full bmh here: http://registry.ran-vcl01.ptp.lab.eng.bos.redhat.com:8080/images/OCPBUGS-369/OCPBUGS-369.bmh-cnfde25.yaml
There is no boot mode option in BMC.
Boot mode is UEFI in bios config. Screenshot added: http://registry.ran-vcl01.ptp.lab.eng.bos.redhat.com:8080/images/OCPBUGS-369/SM_boot.png
rhn-support-yliu1 thank you for the bug report.
Looking at logs, it seems to me there may be some confusion around boot modes:
08-19T18:22:13.598435153Z 2022-08-19 18:22:13.598 1 DEBUG ironic.drivers.modules.boot_mode_utils [req-139bda3f-de67-42b1-b3c7-b36b7fcbc414 - - - - -] Deploy boot mode is uefi for 5c3dfff2-36a7-4ef3-9afb-efbce536ef63. get_boot_mode_for_deploy /usr/lib/python3.6/site-packages/ironic/drivers/modules/boot_mode_utils.py:275ESC[00m
2022-08-19T18:22:13.598787518Z 2022-08-19 18:22:13.598 1 INFO ironic.drivers.modules.boot_mode_utils [req-139bda3f-de67-42b1-b3c7-b36b7fcbc414 - - - - -] Boot mode bios currently configured on node 5c3dfff2-36a7-4ef3-9afb-efbce536ef63 does not match the boot mode uefi requested for provisioning.Attempting to set node boot mode to uefi.ESC[00m
This is also reflected in the event log (entries 8 & 9 in the screenshot attached)
Can you confirm what boot mode is configured in the BIOS settings of the node?
Thank you,
Jacob
This bug needs information necessary for us to investigate and move forward. We have not received it in a month since asking so we’re going to close this bug.
If you’re able to provide the information and work with us at a later time, feel free to reopen the bug.