| *** d34dh0r5- is now known as d34dh0r53 | 14:00 | |
| gryf | which channel is the right one regarding devstack deployments? | 14:04 |
|---|---|---|
| gryf | or devstack development? | 14:05 |
| d34dh0r53 | can someone take a look at https://review.opendev.org/c/openstack/keystonemiddleware/+/884738 ? It seems like Zuul is running but not populating job results or voting on that issue. | 14:14 |
| fungi | gryf: #openstack-qa | 14:16 |
| fungi | d34dh0r53: sure | 14:16 |
| gryf | fungi, thanks. | 14:17 |
| fungi | d34dh0r53: i think we found a corner case/bug in zuul's gerrit driver (or in gerrit's api)... | 14:31 |
| fungi | 2023-07-19 17:09:50,068 ERROR zuul.GerritConnection: Exception: Received response 404: Multiple changes found for openstack%2Fkeystonemiddleware~stable%2Fussuri~Ice8e34440a5fe1baa370646ed70b5e085c4af70e | 14:31 |
| fungi | https://review.opendev.org/q/Ice8e34440a5fe1baa370646ed70b5e085c4af70e | 14:32 |
| fungi | change ids are supposed to be unique per project+branch combination | 14:32 |
| fungi | there are 3 distinct changes with that same change id on ksm's ussuri branch (two of which are abandoned) | 14:33 |
| d34dh0r53 | interesting | 14:33 |
| fungi | i'm guessing someone cherry-picked the change to that branch, then abandoned it and cherry-picked again, and then again... | 14:33 |
| fungi | rather than revising the original cherry-pick they proposed | 14:34 |
| d34dh0r53 | yeah, the abandon message on the two is that they were duplicates | 14:34 |
| fungi | i'm not entirely sure how to work around that for now other than abandoning 884738 and then proposing it again but regenerating the change-id for it (just delete the line in the commit message and then let git-review add a new random one) | 14:35 |
| fungi | it won't share the same change-id as the others, but it should work otherwise | 14:35 |
| fungi | looks like the first two changes were proposed 30 seconds apart, and the third 90 seconds after that | 14:37 |
| fungi | the timing (shortly after 02:00 utc) doesn't collide with any of our cronjobs for backups/db dumps or git pruning (we've seen reviews pushed right when backups kick off resulting in duplicates in the past) | 14:39 |
| fungi | but maybe there was something else making gerrit unresponsive around that time | 14:40 |
| fungi | unfortunately that was a couple of months ago so at best i might have some coarse-grained resource graphs from that timeframe | 14:40 |
| fungi | but i'm not having any luck bending cacti to my will. i think its custom time range widget is broken | 14:45 |
| frickler | fungi: ah, I also noticed that cacti issue recently, but didn't dig deeper yet. at least it seems reproducible then | 15:14 |
| fungi | d34dh0r53: i'm pretty sure there must have been some sort of lag or retry involved, because normally gerrit will prevent proposing a change with a duplicate change-id, so zuul is probably doing the best it can in that circumstance (the error actually came back from gerrit's rest api when zuul tried to comment on the change) | 15:18 |
| fungi | there's probably not much we can do from the zuul side of things to work around it | 15:18 |
| d34dh0r53 | fungi: ack, so should we abandon that one and re-backport? | 15:37 |
| fungi | d34dh0r53: yes, and like i said, clear out the commit-id in the commit message before re-proposing so that git-review will generate a new, different one that won't collide | 15:40 |
| d34dh0r53 | fungi: ack, I'll give that a try | 15:41 |
| fungi | er, the change-id in the commit message, i mean | 15:41 |
| d34dh0r53 | I knew what you meant :) | 15:41 |
| zul | Hi, Can I get another +2 on this one please https://review.opendev.org/c/openstack/project-config/+/888930 | 15:57 |
| fungi | zul: slittle's still out, right? i guess you're not expecting any further ml discussion about that proposal? i can single-core approve it on the assumption there's consensus from stx leadership | 16:03 |
| zul | I believe there is consensus from the stx leadership | 16:05 |
| fungi | zul: good enough for me. approved | 16:10 |
| opendevreview | Merged openstack/project-config master: Add starlingx/apt-ostree https://review.opendev.org/c/openstack/project-config/+/888930 | 16:21 |
| *** melwitt is now known as jgwentworth | 18:26 | |
| *** haleyb is now known as haleyb_out | 21:54 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!