Bug 2280606
| Summary: | nova-manage db purge fails on large datasets | ||
|---|---|---|---|
| Product: | Red Hat OpenStack | Reporter: | Alex Stupnikov <astupnik> |
| Component: | openstack-nova | Assignee: | melanie witt <mwitt> |
| Status: | CLOSED MIGRATED | QA Contact: | OSP DFG:Compute <osp-dfg-compute> |
| Severity: | medium | Docs Contact: | |
| Priority: | medium | ||
| Version: | 17.1 (Wallaby) | CC: | alifshit, dasmith, eglynn, jhakimra, kchamart, mwitt, sbauza, sgordon, vromanso |
| Target Milestone: | async | Keywords: | Patch, Triaged |
| Target Release: | 17.1 | ||
| Hardware: | All | ||
| OS: | All | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | If docs needed, set a value | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2025-01-14 20:41:54 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
Alex Stupnikov
2024-05-15 11:42:37 UTC
I had a look in the code and can confirm this bug. Depending on the environment it may be possible for the customer to work around the problem by specifying --before to 'nova-manage db purge' to manually limit the number of rows the command will attempt to delete in one transaction. And repeat the command with subsequent adjustments to --before until all rows are deleted or until the number of rows in the shadow tables has decreased enough for --all to be used again. It may be worth noting that there were enhancements made to the related 'nova-manage db archive_deleted_rows' command to better handle very large numbers of records in 16.2.6 [1] and 17.1.1 [2] because it might be possible to get around the problem by purging records indirectly with 'nova-manage db archive_deleted_rows --purge'. We will aim to get 'nova-manage db purge' fixed for 17.1 provided we have a z4 release. [1] https://bugzilla.redhat.com/show_bug.cgi?id=2170683 [2] https://bugzilla.redhat.com/show_bug.cgi?id=2170686 Moving to async since we're out of runway for 17.1.4 non-blockers, and the impact seems to be not too high, given the presence of the workaround of using --before. We'll continue the work to get this into 18 as tracked by https://issues.redhat.com/browse/OSPRH-13035, but for 17.1 this no longer fits into the inclusion criteria: https://groups.google.com/a/redhat.com/g/openstack-program/c/sn4Pso92fuM |