| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 824077: Add a zk-shell debug script https://review.opendev.org/c/zuul/zuul/+/824077 | 00:23 | |
| -@gerrit:opendev.org- Zuul merged on behalf of Clark Boylan: [zuul/zuul-jobs] 824057: Update ansible-lint pin to fix rich dep https://review.opendev.org/c/zuul/zuul-jobs/+/824057 | 06:59 | |
| -@gerrit:opendev.org- Albin Vass proposed: [zuul/nodepool] 823978: Add option of configurating imagePullSecrets for openshift drivers https://review.opendev.org/c/zuul/nodepool/+/823978 | 07:57 | |
| -@gerrit:opendev.org- Albin Vass proposed: [zuul/nodepool] 823978: Add option of configurating imagePullSecrets for openshift drivers https://review.opendev.org/c/zuul/nodepool/+/823978 | 08:58 | |
| -@gerrit:opendev.org- Albin Vass proposed: [zuul/nodepool] 823978: Add option of configurating imagePullSecrets for openshift drivers https://review.opendev.org/c/zuul/nodepool/+/823978 | 09:10 | |
| @avass:vassast.org | Could we create a new nodepool release? openshiftpods is broken in 4.3.0 and the fix was merged 4 months ago: https://review.opendev.org/c/zuul/nodepool/+/796524 | 13:13 |
|---|---|---|
| -@gerrit:opendev.org- Albin Vass proposed: [zuul/nodepool] 823978: Add option of configuring imagePullSecrets for openshift drivers https://review.opendev.org/c/zuul/nodepool/+/823978 | 13:19 | |
| @kuprince:matrix.org | HI Guys. | 14:08 |
| I have a multiple static nodes defined in nodeset. I also defined groups of nodes in the nodeset. Now I want to run the playbooks on single host out of multiple hosts present in the group, and the host should be choosen at random, based on node instances availability in zuul. | ||
| @kuprince:matrix.org | how can we do it? any idea? | 14:08 |
| @tristanc_:matrix.org | Prince Kumar: perhaps you can try using https://docs.ansible.com/ansible/2.9/plugins/lookup/random_choice.html | 14:20 |
| @kuprince:matrix.org | > <@tristanc_:matrix.org> Prince Kumar: perhaps you can try using https://docs.ansible.com/ansible/2.9/plugins/lookup/random_choice.html | 14:25 |
| can we select hosts based on this? | ||
| @jim:acmegating.com | Albin Vass: yep, will work on that today. | 14:31 |
| @avass:vassast.org | corvus: cool thanks. I did some work to be able to work with the master branch in the meantime but a release is still a good idea | 15:41 |
| @clarkb:matrix.org | Prince Kumar: if you want the job to execute against a single random node you would ask for a single node from a label that has a number of valid entities associated with it. Then nodepool will only give you the one node. If you want the job to run against many nodes but then run a playbook against only one node then you would do something within ansible to select one node out of the inventory you have been given | 15:50 |
| @mhuin:matrix.org | hi all, quick question: is the zookeeper connection mandatory when running zuul tenant-config-check ? | 17:24 |
| @mhuin:matrix.org | I mean, right now it is, as in if the zookeeper config is incorrect in zuul.conf zuul tenant-conf-check returns an error or loops on attemps to connect to zookeeper | 17:25 |
| @mhuin:matrix.org | but could the tenants config be checked without the need to connect to zookeeper? | 17:26 |
| @mhuin:matrix.org | (zuul 4.11.0 question btw) | 17:26 |
| @tristanc_:matrix.org | mhu: https://review.opendev.org/c/zuul/zuul/+/824058 seems to work on py36, so I think it is not necessary. Not sure why there is a timeout on py38 though. | 17:33 |
| @jim:acmegating.com | mhu: tristanC I can't think of a reason it should be required; i think that just got overlooked in a refactor. regarding the 3.8 timeout -- i've been seeing more timeouts on other changes, and i'm starting to suspect some dependency may have changed over the break and destabilized things. | 17:42 |
| @jim:acmegating.com | (specifically for 3.8) | 17:43 |
| @clarkb:matrix.org | re next steps for zuulv5 the major stuff is the gearman rpc cleanups? I don't know that I'll start reviewing those today as I clear out my backlog, but hopefully I can get to them tomorrow | 17:45 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 824218: Pin tzlocal to avoid warnings https://review.opendev.org/c/zuul/zuul/+/824218 | 17:58 | |
| @jim:acmegating.com | Clark: i think 2 things: 1) i've been working with swest & co on some changes to zk -- they found some inefficiencies that compound enough to cause operational problems. i think that's the first priority. then i think the second is landing the gearmanectomy stack. | 17:59 |
| @jim:acmegating.com | re #1 -- i think the actual changes are gtg, but i'm fighting weird test failures (open files / out of disk space...) | 17:59 |
| @clarkb:matrix.org | all of those are under the sos hashtag too I assume? If so I'll be sure to start with them | 17:59 |
| @jim:acmegating.com | yeah | 18:00 |
| @jim:acmegating.com | as soon as i set an autohold it started passing | 18:02 |
| @jim:acmegating.com | if it passes one more time, then i'm going to assume the out-of-disk-space was a fluke and it's only the too-many-open-files problem that's a real bug | 18:03 |
| @jim:acmegating.com | zuul-maint: how's this for a nodepool tag? commit e95b146d10936a9243bdd5639454c3bdd9b62aec (HEAD -> master, tag: 4.4.0, origin/master, refs/changes/81/821881/6) | 18:19 |
| @clarkb:matrix.org | The release notes look accurate and mostly complete too | 18:21 |
| @fungicide:matrix.org | e95b146 looks like the tip of master, we have 26 non-merge commits since 4.3.0, 6 new features added to the release notes (mostly azure-related), a few fixes and dependency updates but i agree those likely fall below the bar for needing release notes. proposal lgtm, thanks! | 18:41 |
| @clarkb:matrix.org | corvus: we're putting the foundation's semi weekly newsletter and for Zuul content the nodepool releas would be good. Do you think you'll push that today? | 20:36 |
| @jim:acmegating.com | Clark: yep, just after i finish lunch :) | 20:38 |
| @clarkb:matrix.org | cool, I'll be sure to get it in the newsletter then | 20:39 |
| -@gerrit:opendev.org- Clark Boylan proposed: [zuul/zuul-jobs] 824235: Remove centos-8 testing https://review.opendev.org/c/zuul/zuul-jobs/+/824235 | 20:53 | |
| @jim:acmegating.com | pushed 4.4.0 | 20:54 |
| @fungicide:matrix.org | thanks! | 21:00 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: | 21:25 | |
| - [zuul/zuul] 823587: Add some ZK debug scripts https://review.opendev.org/c/zuul/zuul/+/823587 | ||
| - [zuul/zuul] 824077: Add a zk-shell debug script https://review.opendev.org/c/zuul/zuul/+/824077 | ||
| @fungicide:matrix.org | interesting, it looks like the zuul-publish-tox-docs build triggered in the release pipeline doesn't update the tag versions in release notes | 21:26 |
| @clarkb:matrix.org | ya we need another change to land now iirc | 21:27 |
| @fungicide:matrix.org | well, maybe not the whole story... the docs preview site archived with the build results did update it | 21:28 |
| @fungicide:matrix.org | https://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_143/81645435cdb15da3d35cbfab00fd33f0b61cd45d/release/zuul-publish-tox-docs/143392b/docs/releasenotes.html | 21:28 |
| @fungicide:matrix.org | leaves me wondering if something's broken updates to that volume in afs | 21:29 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: | 21:29 | |
| - [zuul/zuul] 823739: Offload FrozenJob secrets https://review.opendev.org/c/zuul/zuul/+/823739 | ||
| - [zuul/zuul] 824244: DNM: remove file limit https://review.opendev.org/c/zuul/zuul/+/824244 | ||
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed on behalf of Simon Westphahl: [zuul/zuul] 823782: Compress ZKObject data https://review.opendev.org/c/zuul/zuul/+/823782 | 21:29 | |
| @fungicide:matrix.org | aha! got it | 21:31 |
| @jim:acmegating.com | fungi, Clark i think everything is as expected; the release job did update the tagged version of the docs: https://zuul-ci.org/docs/nodepool/4.4.0/releasenotes.html | 21:31 |
| @jim:acmegating.com | the 'master' version will update when a change lands to master | 21:31 |
| @fungicide:matrix.org | the versopn-specific upload worked, but master branch is what's displayed by default | 21:32 |
| @fungicide:matrix.org | yep, that | 21:32 |
| @jim:acmegating.com | the tagged version is the link i send out in announcement emails | 21:32 |
| @fungicide:matrix.org | cool | 21:32 |
| @jim:acmegating.com | and happily, the list of versions in the RTD version selector is updated in the tagged version as well, so we can expect that to work correctly on the master once a change lands there | 21:32 |
| @fungicide:matrix.org | and yeah, so Clark is also correct, the default one only gets updated when the master branch moves, because it's the copy specifically for the master branch | 21:33 |
| @jim:acmegating.com | i approved a simple nodepool change | 21:35 |
| -@gerrit:opendev.org- Zuul merged on behalf of Clark Boylan: [zuul/zuul-jobs] 824235: Remove centos-8 testing https://review.opendev.org/c/zuul/zuul-jobs/+/824235 | 21:41 | |
| -@gerrit:opendev.org- Tobias Henkel proposed: [zuul/zuul] 824245: Don't join the command thread if we stop via command socket https://review.opendev.org/c/zuul/zuul/+/824245 | 21:45 | |
| @tobias.henkel:matrix.org | corvus, Clark just stumbled accross this when doing some lifecycle testing with the scheduler ^ | 21:46 |
| @tobias.henkel:matrix.org | and the scheduler doesn't seem to react to sigterm when starting with -f | 21:47 |
| @fungicide:matrix.org | looks like the nodepool image tags on dockerhub are updated now | 21:48 |
| @jim:acmegating.com | release announcement sent | 21:55 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!