Following the instructions in the parent ticket, we need to upgrade clouddb1021 to bookworm.
We will need to keep the /srv volume during the reimaging process.
Avoid the beginning of the month, which is when the sqoop jobs make the most use of this server.
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Resolved | Request | VRiley-WMF | T368518 decommission clouddb1021 | ||
| Resolved | fnegri | T365424 Upgrade clouddb* hosts to Bookworm | |||
| Declined | BTullis | T365450 Upgrade clouddb1021 to bookworm |
By default /srv is preserved
modules/profile/data/profile/installserver/preseed.yaml:
'clouddb1*': - reuse-parts.cfg - partman/custom/reuse-db.cfg
I have started work on T365453: Bring an-redacteddb1001 into service to replace clouddb1021 before this ticket.
It's quite possible that we will be able to decom clouddb1021 instead of upgrading it to bookworm, but I'll see how the other ticket progresses.
We are about to bring an-redacteddb1021 into service, so I will decline this ticket in favour of a decom ticket for clouddb1021 instead. T368518: decommission clouddb1021