I made a wrong click by accident that deleted the instance behind https://codesearch.wmcloud.org/
Now it's gone. There might be a chance to get it back from backups but it's not simple.
https://wikitech.wikimedia.org/wiki/Portal:Cloud_VPS/Admin/Instance_backups#Restoring
Trying to recreate it (it was on bookworm, T429496) so either that or try on trixie right away.
| Subject | Author | Repo | Branch | Lines +/- | |
|---|---|---|---|---|---|
| codesearch: ensure docker-cli is installed if on trixie or newer | Dzahn | operations/puppet | production | +4 -0 |
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Resolved | Dzahn | T429828 re-create codesearch instance / upgrade codesearch to trixie | |||
| Resolved | Dzahn | T429967 Codesearch suddenly uses the default Hound UI |
Mentioned in SAL (#wikimedia-cloud) [2026-06-22T19:29:51Z] <mutante> deleted instance codesearch9 by accident :( - creating instance codesearch10 on trixie - T429828 T429496
Mentioned in SAL (#wikimedia-cloud) [2026-06-22T19:56:44Z] <mutante> creating instance codesearch11 on bookworm - hound services failed to start on trixie - verify if it's distro related - T429828 T429496
Change #1304895 had a related patch set uploaded (by Dzahn; author: Dzahn):
[operations/puppet@production] codesearch: ensure docker-cli is installed if on trixie or newer
Change #1304895 merged by Dzahn:
[operations/puppet@production] codesearch: ensure docker-cli is installed if on trixie or newer
Mentioned in SAL (#wikimedia-cloud) [2026-06-22T20:57:58Z] <mutante> - re-created on trixie! oof:)) should be working - https://codesearch.wmcloud.org/search/ T429828
what was needed to fix it, in addition to puppet:
Since we don't want to use /usr/local/sbin/wmcs-prepare-cinder-volume to re-mount the cinder volume, because that would also format it:
blkid -o export /dev/sdb | grep -E '^(UUID|TYPE)='
Add this line: UUID=a81014b1-7b19-44b6-ab4d-b22a30ddb637 /srv ext4 rw,discard,rw,relatime 0 2 to /etc/fstab to ensure it survives the next reboot.