Saturday, 2025-08-23

opendevreviewOpenStack Proposal Bot proposed openstack/project-config master: Normalize projects.yaml  https://review.opendev.org/c/openstack/project-config/+/95799502:33
*** jroll06 is now known as jroll009:03
fungiit's gotten to the project.tarballs volume, just two relatively small ones to go after that finishes13:40
fungiit finished!16:14
fungiso mirror.logs needs corresponding ro replicas on afs01.dfw and afs02.dfw, while service and user just need ro replicas added to afs02.dfw16:18
fungii'll work on adding those in a bit and then move the rw volumes for them16:19
fungidocs-old can be ignored, i think16:19
fungiused `vos addisite ...` to add ro replicas of mirror.logs to both servers and followed that up with a `vos release mirror.logs`16:23
fungidid similar to add second ro sites for service and user, now i'll move the rw volumes for all three to afs02.dfw16:25
clarkblooks like the weekly zuul upgrade completed overnight and that means we're running with the refactored pbr hook in zuul16:30
clarkbseems to still produce a working website16:30
funginow we've got rw volumes for mirror.logs, service and user on afs02.dfw with ro replicas for all of those on both afs01.dfw and afs02.dfw16:30
fungithough it looks like the rw volume moves failed for these: mirror.debian, mirror.ubuntu, mirror.ubuntu-ports (rw volumes are still on afs01.dfw according to `vos listvldb`)16:33
fungii'll try them again16:33
clarkbmaybe that explains why debian security started so soon16:33
fungimirror.debian is in progress again now16:34
fungii didn't see any new kdc03 backup failures so seems likethe borg reinstall worked16:34
fungihopefully mirror.debian will finish before i check out for the night and then i'll restart the other two stragglers in series17:28
fungigrafana says /vicepa on afs02.dfw is now slightly higher than on afs01.dfw, but only by +0.6% so i don't think the problem is space17:30
fungialso mirror.debian hasn't updated yet (patch for the key addition hasn't merged) meaning it's 14 days stale, so i don't think churn explains the move failure either17:31

Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!