Bug 1820698
| Summary: | After a failed introspection, and inspector goes on a second try, it might fail to set_boot_device if it's stuck in POST | ||
|---|---|---|---|
| Product: | Red Hat OpenStack | Reporter: | David Vallee Delisle <dvd> |
| Component: | openstack-ironic-inspector | Assignee: | Julia Kreger <jkreger> |
| Status: | CLOSED ERRATA | QA Contact: | mlammon |
| Severity: | low | Docs Contact: | |
| Priority: | low | ||
| Version: | 16.0 (Train) | CC: | achernet, bfournie, cswanson, dhill, dtantsur, eduen, hbrock, jkreger, jparoly, jslagle, mburns, pweeks, rpittau, slinaber, stendulker |
| Target Milestone: | beta | Keywords: | Reopened, Triaged |
| Target Release: | 17.0 | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Whiteboard: | |||
| Fixed In Version: | openstack-ironic-inspector-10.6.2-0.20220118051837.8f97076.el8ost | Doc Type: | If docs needed, set a value |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2022-09-21 12:09:43 UTC | Type: | Bug |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
|
Description
David Vallee Delisle
2020-04-03 15:45:03 UTC
As Ilya noted - this is up to proliantutils or iLO driver to make that decision...? Redfish does not require any specific power state when changing boot options I think the managed introspection functionality that merged during the last upstream development cycle (Ussuri), should effectively solve this as the item managing the power and boot mode settings is then just ironic with-in a workflow, at least as long as [inspector]require_managed_boot is set to True. The only way to realistically prevent this is for inspector to force the power state off in advance of trying to run, or the driver trying to assert power state off before changing the boot device. I guess the machine was already powered on when inspection was triggered? Depending on the code path, it looks like the call goes to inspector, inspector then attempts to ask ironic to set the network device to boot, and then reboot the node. I guess my disconnect is why is the node on even before this step? Patch uploaded upstream to address this. The actual process in this case is being driven by ironic-inspector. The previous focus on proliantutils was not correct as it is legitimately failing, just not with much clarity, although patches have been proposed upstream to improve that. low priority, no progress in the last year closing wontfix If this needs to be reconsidered, please re-open I noticed the noted patch will be in OSP17, linking appropriately and moving to modified state. Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Release of components for Red Hat OpenStack Platform 17.0 (Wallaby)), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHEA-2022:6543 |