| @jim:acmegating.com | okay, i get it. getCachedChanges goes to ZK, so we're doing a get_children and a get call for every item in the cache (about 6500 in opendev) every time we call onChangeUpdated. that does take about 14 seconds. | 02:30 |
|---|---|---|
| @jim:acmegating.com | Clark: ^ smoking gun fyi | 02:30 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] WIP: Don't clear dependent change cache https://review.opendev.org/c/zuul/zuul/+/810539 | 03:14 | |
| -@gerrit:opendev.org- Ian Wienand proposed: | 07:03 | |
| - [zuul/zuul] github: more complete mocking for app setup https://review.opendev.org/c/zuul/zuul/+/810553 | ||
| - [zuul/zuul] github: handle suspended apps https://review.opendev.org/c/zuul/zuul/+/810554 | ||
| -@gerrit:opendev.org- Ian Wienand proposed: | 07:08 | |
| - [zuul/zuul] github: more complete mocking for app setup https://review.opendev.org/c/zuul/zuul/+/810553 | ||
| - [zuul/zuul] github: handle suspended apps https://review.opendev.org/c/zuul/zuul/+/810554 | ||
| -@gerrit:opendev.org- Simon Westphahl proposed on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: [zuul/zuul] WIP: Don't clear dependent change cache https://review.opendev.org/c/zuul/zuul/+/810539 | 08:10 | |
| -@gerrit:opendev.org- Ian Wienand proposed: | 09:55 | |
| - [zuul/zuul] github: more complete mocking for app setup https://review.opendev.org/c/zuul/zuul/+/810553 | ||
| - [zuul/zuul] github: handle suspended apps https://review.opendev.org/c/zuul/zuul/+/810554 | ||
| -@gerrit:opendev.org- Simon Westphahl proposed: | 12:38 | |
| - [zuul/zuul] wip: Make QueueItem a Zookeeper object https://review.opendev.org/c/zuul/zuul/+/809414 | ||
| - [zuul/zuul] wip: Store pipeline state in Zookeeper https://review.opendev.org/c/zuul/zuul/+/810658 | ||
| @icey:matrix.org | is there any way to diagnose why a job is stuck Queued for a long time? I've noticed that my deploy is sometimes taking 10 minutes to start running a job even when there's nothing else running | 13:32 |
| -@gerrit:opendev.org- Matthieu Huin https://matrix.to/#/@mhuin:matrix.org proposed: | 13:52 | |
| - [zuul/zuul] Add authentication-realm attribute to tenants https://review.opendev.org/c/zuul/zuul/+/735586 | ||
| - [zuul/zuul] web UI: user login with OpenID Connect https://review.opendev.org/c/zuul/zuul/+/734082 | ||
| - [zuul/zuul] web UI: allow a privileged user to dequeue a change https://review.opendev.org/c/zuul/zuul/+/734850 | ||
| - [zuul/zuul] web UI: allow a privileged user to re-enqueue a change https://review.opendev.org/c/zuul/zuul/+/736772 | ||
| - [zuul/zuul] Web UI: allow a privileged user to request autohold https://review.opendev.org/c/zuul/zuul/+/768115 | ||
| - [zuul/zuul] Web UI: add Autoholds, Autohold page https://review.opendev.org/c/zuul/zuul/+/768199 | ||
| - [zuul/zuul] REST API, Web UI: add pipelines' manager, triggers data in status https://review.opendev.org/c/zuul/zuul/+/736968 | ||
| - [zuul/zuul] web UI: allow a privileged user to promote a change https://review.opendev.org/c/zuul/zuul/+/781858 | ||
| - [zuul/zuul] Web UI: Add "Create Autohold Request" form, improve API error messages https://review.opendev.org/c/zuul/zuul/+/802559 | ||
| - [zuul/zuul] Example Docker compose: keycloak integration https://review.opendev.org/c/zuul/zuul/+/769943 | ||
| @jim:acmegating.com | icey: mouseover the 'queued' status in the web ui or check the scheduler log | 13:53 |
| @icey:matrix.org | mousing over doesn't seem to work (4.x+ only?); nothing jumps out at me in the scheduler log | 13:55 |
| @icey:matrix.org | although I do notice that the executors are _constantly_ `zuul.ExecutorServer: Updating repo` | 13:55 |
| -@gerrit:opendev.org- Matthieu Huin https://matrix.to/#/@mhuin:matrix.org proposed: | 14:21 | |
| - [zuul/zuul] Zuul-web: Add authentication-realm attribute to tenants https://review.opendev.org/c/zuul/zuul/+/735586 | ||
| - [zuul/zuul] zuul-web: add pipelines' manager, triggers data in status https://review.opendev.org/c/zuul/zuul/+/736968 | ||
| - [zuul/zuul] web UI: user login with OpenID Connect https://review.opendev.org/c/zuul/zuul/+/734082 | ||
| - [zuul/zuul] web UI: allow a privileged user to dequeue a change https://review.opendev.org/c/zuul/zuul/+/734850 | ||
| - [zuul/zuul] web UI: allow a privileged user to re-enqueue a change https://review.opendev.org/c/zuul/zuul/+/736772 | ||
| - [zuul/zuul] Web UI: allow a privileged user to request autohold https://review.opendev.org/c/zuul/zuul/+/768115 | ||
| - [zuul/zuul] Web UI: add Autoholds, Autohold page https://review.opendev.org/c/zuul/zuul/+/768199 | ||
| - [zuul/zuul] web UI: allow a privileged user to promote a change https://review.opendev.org/c/zuul/zuul/+/781858 | ||
| - [zuul/zuul] Web UI: Add "Create Autohold Request" form, improve API error messages https://review.opendev.org/c/zuul/zuul/+/802559 | ||
| - [zuul/zuul] Example Docker compose: keycloak integration https://review.opendev.org/c/zuul/zuul/+/769943 | ||
| - [zuul/zuul] Web UI: Show pipeline types as icons https://review.opendev.org/c/zuul/zuul/+/810699 | ||
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: | 16:26 | |
| - [zuul/zuul] WIP: Don't clear dependent change cache https://review.opendev.org/c/zuul/zuul/+/810539 | ||
| - [zuul/zuul] Add a test for github dependency updates https://review.opendev.org/c/zuul/zuul/+/810726 | ||
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: | 16:27 | |
| - [zuul/zuul] Add a test for github dependency updates https://review.opendev.org/c/zuul/zuul/+/810726 | ||
| - [zuul/zuul] WIP: Don't clear dependent change cache https://review.opendev.org/c/zuul/zuul/+/810539 | ||
| -@gerrit:opendev.org- Jeremy Stanley proposed: [zuul/zuul-jobs] Deprecate EOL Python releases and OS versions https://review.opendev.org/c/zuul/zuul-jobs/+/810299 | 17:42 | |
| @mhuin:matrix.org | hello zuul-maint, this change just needs a workflow+1: https://review.opendev.org/c/zuul/zuul/+/806201 | 17:45 |
| @jim:acmegating.com | oh that changed, gimme a bit to rereview pls | 17:46 |
| @mhuin:matrix.org | no worries | 17:47 |
| @mhuin:matrix.org | On a side note I've also re-ordered the admin GUI patch chain so that the first 2 patches are modifications to zuul-web and the rest is purely GUI stuff. The zuul-web changes are small and it'd be good to merge them, so that it's not necessary anymore to rebuild the images to use the demo compose | 17:50 |
| @mhuin:matrix.org | (and the topic isn't random anymore! https://review.opendev.org/q/topic:%22GUI_admin%22+(status:open%20OR%20status:merged) ) | 17:50 |
| @mhuin:matrix.org | The changes didn't pass the CI but I suspect some flakiness, I'll recheck tomorrow | 17:51 |
| @jim:acmegating.com | mhu: 806201+1 but i left one more js/css/something puzzle on that in a comment if you feel like taking a look. i don't think it's a critical UX thing (the benefit of the switch outweighs the weirdness). but i don't want to deprive you of the opportunity to figure that one out if you have some time. :) | 17:56 |
| @jim:acmegating.com | * mhu: 806201+3 but i left one more js/css/something puzzle on that in a comment if you feel like taking a look. i don't think it's a critical UX thing (the benefit of the switch outweighs the weirdness). but i don't want to deprive you of the opportunity to figure that one out if you have some time. :) | 17:56 |
| @mhuin:matrix.org | corvus: thanks! | 17:57 |
| @fungicide:matrix.org | any feel for how much consensus we should seek on https://review.opendev.org/809899 prior to approving? my appeal to the mailing list a week ago didn't generate any responses (negative or otherwise). should we announce this more broadly? | 18:20 |
| @fungicide:matrix.org | i mean didn't generate any responses on the mailing list, of course we have a bunch of positive votes on the change itself | 18:21 |
| @jim:acmegating.com | fungi: seems pretty good to me -- only thing i'd suggest is maybe checking in with tristanC or ianw to make sure they've seen it | 18:37 |
| @fungicide:matrix.org | can do | 18:50 |
| @fungicide:matrix.org | looks like one of them just +2'd it, i'll follow up with the other next week, as i gather it's a national holiday there now | 18:51 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: | 20:46 | |
| - [zuul/zuul] Add a test for github dependency updates https://review.opendev.org/c/zuul/zuul/+/810726 | ||
| - [zuul/zuul] Add a test for dequeue-on-new-patchest https://review.opendev.org/c/zuul/zuul/+/810772 | ||
| @jpew:matrix.org | Is the ssh-agent running the executor forwarded to the worker node? If not can I run an ssh-agent on the worker node? (I want the worker to be able to ssh into another host) | 20:51 |
| @jim:acmegating.com | jpew: it is not forwarded; yau can take a look at the multinode setup roles in zuul-jobs for some potential help setting that sort of thing up (including copying the per-build ssh keys to all hosts) | 20:56 |
| @jim:acmegating.com | * jpew: it is not forwarded; you can take a look at the multinode setup roles in zuul-jobs for some potential help setting that sort of thing up (including copying the per-build ssh keys to all hosts) | 20:57 |
| @jpew:matrix.org | Ah, so I should copy the per-build SSH key to the target node (using a common key on the executor) and remove it after the job instead of trying to add the common key to the worker? | 20:58 |
| @jim:acmegating.com | jpew: well, it's one of many options. :) also, note there is a per-project ssh key too, which may be helpful, depending on circumstances. | 21:01 |
| @jim:acmegating.com | https://zuul-ci.org/docs/zuul/reference/jobs.html#ssh-keys | 21:02 |
| @jpew:matrix.org | That key is still only available on the executor? | 21:16 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: | 21:35 | |
| - [zuul/zuul] Add a test for dequeue-on-new-patchest https://review.opendev.org/c/zuul/zuul/+/810772 | ||
| - [zuul/zuul] Remove onChangeUpdated method https://review.opendev.org/c/zuul/zuul/+/810539 | ||
| @jim:acmegating.com | jpew: yes, and the private key is inaccessible. but it's long-lived, so if you're about to do something like generate a key and install it on a long-running host and also the executor, then the per-project key may be similar enough to use that. | 21:36 |
| @jim:acmegating.com | note also that storing a private key in a zuul secret and then writing it out in the job is also an option; i think there are roles for that too | 21:37 |
| @jpew:matrix.org | @corvus Ya, I can do that easily enough. I was sort of hoping there might be a more-secure (trickier) solution, but looks like not. | 21:38 |
| @fungicide:matrix.org | the trickier solutions are not necessarily any more secure. at least for our (opendev's) case, we have to be very careful what we give job nodes access to because they often run entirely untrusted code as root | 22:27 |
| @fungicide:matrix.org | so for the most part we don't want them to have access to anything else besides maybe other ephemeral nodes in the same multi-node build context | 22:29 |
| @fungicide:matrix.org | which is where the multinode setup roles in zuul-jobs originated | 22:29 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] Only refresh deps if change messages have changed https://review.opendev.org/c/zuul/zuul/+/810787 | 22:46 | |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: | 23:21 | |
| - [zuul/zuul] Remove onChangeUpdated method https://review.opendev.org/c/zuul/zuul/+/810539 | ||
| - [zuul/zuul] Only refresh deps if change messages have changed https://review.opendev.org/c/zuul/zuul/+/810787 | ||
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!