Human deployers and automated processes have hit the problem described in T438148 several times over the last couple of weeks. We've reached the point where /srv just needs to be bigger on deploy1003 and friends. It's currently 277G. I suggest increasing by >= 30GB.
Description
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Resolved | dancy | T438148 Image building is running out of space in `/srv` | |||
| Resolved | Clement_Goubert | T438232 Request to increase the size of /srv on production deployment servers |
Event Timeline
Ack. Setting to Medium as I understand there's a workaround mentioned in the other task
We can allocate about 80GB more, but that's it, the vg would be full afterwards. I will allocate 50GB as a first step (so we keep some margin on the vg) and we'll go from there.
Mentioned in SAL (#wikimedia-operations) [2026-09-29T10:17:28Z] <claime> deploy2003:~$ sudo lvextend -L +50GB --resizefs /dev/vg0/srv # T438232
Mentioned in SAL (#wikimedia-operations) [2026-09-29T10:17:53Z] <claime> deploy1003:~$ sudo lvextend -L +50GB --resizefs /dev/vg0/srv # T438232
Mentioned in SAL (#wikimedia-operations) [2026-09-29T10:19:19Z] <claime> deploy2002:~$ sudo lvextend -L +50GB --resizefs /dev/vg0/srv # T438232
cgoubert@deploy1003:~$ sudo lvextend -L +50GB --resizefs /dev/vg0/srv Size of logical volume vg0/srv changed from <281.92 GiB (72171 extents) to <331.92 GiB (84971 extents). Logical volume vg0/srv successfully resized. resize2fs 1.47.0 (5-Feb-2023) Filesystem at /dev/mapper/vg0-srv is mounted on /srv; on-line resizing required old_desc_blocks = 36, new_desc_blocks = 42 The filesystem on /dev/mapper/vg0-srv is now 87010304 (4k) blocks long. cgoubert@deploy1003:~$ df -h /srv Filesystem Size Used Avail Use% Mounted on /dev/mapper/vg0-srv 326G 182G 128G 59% /srv
Done on all deploy servers.
Thanks @Clement_Goubert! Combined with the daily scap clean-images, I think 50GB will be plenty.