| *** tkajinam is now known as Guest213 | 08:56 | |
| opendevreview | Jan Marchel proposed openstack/project-config master: Add new project to Nebulous: gui-controller https://review.opendev.org/c/openstack/project-config/+/909435 | 11:40 |
|---|---|---|
| opendevreview | Edward Hope-Morley proposed openstack/project-config master: Add ovn charms to charms-stable-maint purview https://review.opendev.org/c/openstack/project-config/+/909445 | 12:30 |
| opendevreview | Merged openstack/project-config master: Add new project to Nebulous: gui-controller https://review.opendev.org/c/openstack/project-config/+/909435 | 13:07 |
| frickler | has either zuul or gerrit recently become more strict regarding mergeability checks? I'm wondering why zuul refuses to run check jobs on https://review.opendev.org/c/openstack/requirements/+/879743, checking the executor log isn't really helpful, too | 14:36 |
| frickler | my guess is that the reason are the three merged patches at the bottom of the stack at https://review.opendev.org/c/openstack/aodh/+/896097/, but why do tests on the aodh patch itself run, then? | 14:37 |
| frickler | also it would be pretty nasty to have to rebase that whole stack every time a single patch is merged | 14:37 |
| frickler | corvus: ^^ maybe you can have a look and find something I missed? | 14:38 |
| fungi | i wonder if this is a change in behavior for cross-repo dependencies outside a shared queue | 15:03 |
| frickler | oh, I was assuming aodh to be part of the integrated queue, but you are right that it isn't | 15:14 |
| fungi | ildikov: just a heads up, i saw a (subscribed) spammer send junk to the kata-dev and starlingx-discuss mailing lists. i went ahead and deleted their posts from the archive as well as added them to the global bans in the mailman admin interface | 15:21 |
| fungi | looks like they signed up for an account on january 9 | 15:22 |
| fungi | clearly a throwaway gmail username | 15:22 |
| fungi | spam was sent through normal smtp from gmail, not posted through mailman's hyperkitty webui | 15:22 |
| corvus | frickler fungi: i suspect a race updating change information since that event was being processed at the same time in two different tenants (vexxhost and openstack). the dequeue and reported error happened after the data finished updating and zuul saw that the queue did not have all the necessary changes in it. it's definitely a bug in zuul (but we failed safe rather than ran incorrect tests). i don't see a trivial fix at the moment, but i do | 15:41 |
| corvus | suspect a recheck may help in this case and if it happens again. | 15:41 |
| fungi | corvus: noted, i'll try to keep an eye out for more occurrences and see how prevalent it is. do you think this came in with the circdep refactor work or has been lingering in there for longer? | 15:43 |
| corvus | fungi: i think the probability is small that the underlying issue originated with the refactor; i would assign a medium probability that the refactor might change some behavior that could cause it to manifest more easily. | 15:44 |
| fungi | thanks! | 15:45 |
| frickler | I did another recheck right now and got the same response from zuul | 15:54 |
| frickler | if there is not obvious bug seen, I would next rebase the aodh stack and see if that resolves the issue? | 15:55 |
| frickler | s/not/no/ | 15:56 |
| corvus | frickler: for good measure, try rechecking one more time | 15:56 |
| frickler | so it seems I needed to mention your name in the "recheck" comment as magic trigger ... now zuul is starting jobs on it ... ;-) | 16:02 |
| fungi | bwahahaha | 16:03 |
| SvenKieske | I'm having trouble re-editing a drafted message on a WIP Change: https://review.opendev.org/c/openstack/kolla-ansible/+/763011/6..9?tab=comments | 16:10 |
| SvenKieske | I'm getting a gerrit error: "An error occurred" "You might have not enough privileges. Error 404 (Not Found): Not found: 12efa4b1_bf27fcb0 Endpoint: /changes/*~*/revisions/*/drafts/*" | 16:11 |
| clarkb | probably need an acl on the project allowing people to mark (and unmark) changes work in progress outside of the owner | 16:12 |
| clarkb | my hunch is that gerrit is trying to preserve the wip state and can't because you are not the owner | 16:12 |
| SvenKieske | but I didn't edit the WIP state? I just edited a comment of mine? and now the comment ended up being posted twice..weird. | 16:13 |
| SvenKieske | anyway I was able to post the comment somehow, even if it got posted twice. | 16:14 |
| clarkb | SvenKieske: are you editing the files (including the commit message) of the change? or do you mean editing of a comment? I may have misunderstood initially and thought you were editing the files | 16:15 |
| clarkb | I wonder if it actually succeeded in writing the content to disk the first time but raised an error to the browser client then when you posted again it rerecorded | 16:17 |
| clarkb | I don't think I've seen that before. Not sure why that would happen | 16:17 |
| SvenKieske | clarkb: just editing a comment | 16:17 |
| SvenKieske | yeah, me too :) | 16:17 |
| SvenKieske | might very well be (the writing to disk part), because I'm really sure I did not post that comment two times :) | 16:18 |
| opendevreview | Clark Boylan proposed openstack/diskimage-builder master: Fetch compatibile dnf download command in functest setup https://review.opendev.org/c/openstack/diskimage-builder/+/909466 | 16:24 |
| fungi | heading out for a quick lunch, back in under an hour | 16:25 |
| clarkb | infra-root: these services appear to run local mariadbs: lodgeit, gitea, gerrit, mailman3, refstack, etherpad, and keycloak. Of those I believe keycloak is running the latest version and the others will need to go through the upgrade process we discussed last week | 16:50 |
| clarkb | lodgeit (paste) might be a good first candidate given the minimal impact there? Gerrit would be another if it didn't imply a service downtime as the database there is actually not very important anymore. | 16:51 |
| clarkb | Maybe we do those two in that order and then based on what we learn can proceed with the others. | 16:51 |
| opendevreview | Clark Boylan proposed opendev/system-config master: Prep the lodgeit mariadb container for upgrades https://review.opendev.org/c/opendev/system-config/+/909471 | 17:05 |
| opendevreview | Clark Boylan proposed opendev/system-config master: Update lodgeit mariadb container to mariadb 10.11 https://review.opendev.org/c/opendev/system-config/+/909472 | 17:05 |
| clarkb | looking for feedback on those. I think we may be able to do this in one change when we actually do the upgrade but I wanted to get some testing in first | 17:06 |
| corvus | clarkb: ++ to doing that in one change | 17:16 |
| fungi | clarkb: storyboard also runs a local db | 17:41 |
| fungi | though not in a container | 17:41 |
| clarkb | fungi: not with containers | 17:41 |
| fungi | zactly | 17:42 |
| clarkb | right it will need a different process | 17:42 |
| fungi | or switching to a container, sure | 17:42 |
| fungi | clarkb: anyway, your proposed plan seems good to me | 17:43 |
| *** dmitriis is now known as Guest251 | 17:48 | |
| opendevreview | James E. Blair proposed zuul/zuul-jobs master: DNM: test ansible 9 https://review.opendev.org/c/zuul/zuul-jobs/+/909475 | 17:54 |
| *** NeilHanlon is now known as nhanlon | 19:00 | |
| *** nhanlon is now known as NeilHanlon | 19:02 | |
| clarkb | alright I'm going to update the lodgeit chagnes a bit in order to hold a node that can be used to test a change that sets the env var and upgrades in the same go. And if that looks good I'll update things again so that we can use a single change. | 19:15 |
| opendevreview | Clark Boylan proposed opendev/system-config master: Prep the lodgeit mariadb container for upgrades https://review.opendev.org/c/opendev/system-config/+/909471 | 19:18 |
| fungi | i'd be inclined to just upgrade it, but that works too | 19:18 |
| fungi | lodgeit's use of the database is via an orm and about as simplistic as possible | 19:19 |
| clarkb | ya I'm more doing this to uinderstand the process better as a whole particularly for other services | 19:19 |
| clarkb | hold is in place now | 19:20 |
| fungi | makes sense, and yeah it's great for a testbed/dry-run | 19:20 |
| fungi | if something goes wrong with that upgrade, then it's really, really wrong | 19:20 |
| clarkb | also I'm probably going to be afk a bit today starting nowish. Kids are home from school due to the holidy and need to get them outside before cabin fever sets in. I'll get a meeting agenda sent out and continue to poke at the mariadb upgrade stuff as I can | 19:33 |
| fungi | good luck! | 19:42 |
| clarkb | fungi: should I keep the keycloak agenda item or do we think now is a reasonable time to drop it? We can add it back again as the next round of effort there starts up and has stuff to report | 21:16 |
| fungi | clarkb: we should be able to drop it for the moment, until we near the next milestone for the sso spec | 21:20 |
| clarkb | done. I've also started to add some of the more actionable bits from the preptg to the agenda. feel free to add items that I'm forgetting | 21:21 |
| fungi | thanks! | 21:23 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!