| -@gerrit:opendev.org- Zuul merged on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: [zuul/zuul] 909280: Monitor and report executor inode usage https://review.opendev.org/c/zuul/zuul/+/909280 | 09:00 | |
| @jim:acmegating.com | zuul-maint: how does this look for a zuul release? commit dc7c896af2aaa5de2a15bb509ef46394ce23f7f5 (tag: 10.0.0) | 17:32 |
|---|---|---|
| @jim:acmegating.com | i believe that is what opendev is running; it's 2 commits behind current master (it should be the ansible 6->9 transition and omits the inode stuff) | 17:34 |
| @fungicide:matrix.org | commit dc7c896af2aaa5de2a15bb509ef46394ce23f7f5 Merge "Replace Ansible 6 with Ansible 9" | 17:46 |
| @fungicide:matrix.org | yeah, that merged friday before opendev's automated restart, next change merged yesterday | 17:47 |
| @fungicide:matrix.org | inspection of servers indicate it upgraded as scheduled, and the components page shows all services on a consistent version | 17:50 |
| @fungicide:matrix.org | lgtm! | 17:50 |
| @clarkb:matrix.org | I was just checking that myself | 17:51 |
| @clarkb:matrix.org | Agreed this looks good and that commit sha looks correct to me | 17:51 |
| @fungicide:matrix.org | it's been a couple of weekdays already and i haven't heard anyone scream. expected more fallout from the ansible 6 removal honestly, but maybe all the impact is in bitrotten branches nobody's paying attention to | 17:59 |
| @fungicide:matrix.org | only new thing that came up was the odd race condition yesterday with a cross-repo dependency outside a shared queue resulting in a "depends on a change that could not be merged" error, but that's the only change i saw hit it and it sounds like it's unlikely to be a recent regression | 18:01 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/nodepool] 909607: Change the AWS default image volume-type from gp2 to gp3 https://review.opendev.org/c/zuul/nodepool/+/909607 | 18:01 | |
| @clarkb:matrix.org | we set the default across opendev to ansible 8 too so old bitrotted stuff may have broken way back then | 18:03 |
| @mhuin:matrix.org | hi, quick question: do executors require a unique IP or would it be possible to run several executor processes (in distinct venvs, or containers) on the same host? | 19:39 |
| @clarkb:matrix.org | off the top of my head I think the main consideration there is the finger protocol stuff? But you can have each one listen on a different port? | 19:40 |
| @clarkb:matrix.org | all of the other connections are from the executor out (to zk and to test nodes) | 19:40 |
| @clarkb:matrix.org | that said the executors should scale up pretty well to fit the size of your host. | 19:41 |
| @mhuin:matrix.org | thanks Clark - It's probably better to have smaller, but more hosts than several executors on the same, beefy host, if only for availability concerns | 19:44 |
| @mhuin:matrix.org | but just as a thought experiment, I was wondering if the latter would be possible | 19:44 |
| @fungicide:matrix.org | it's also unclear why multiple executors on one host would be necessarily better than one (in my estimation, seems like it would be worse in almost every way) | 19:47 |
| @fungicide:matrix.org | the more resources you give an executor, the more simultaneous builds it should be capable of managing | 19:48 |
| @fungicide:matrix.org | i guess your concern is something other than scaling up capacity? like additional workload isolation? | 19:49 |
| @mhuin:matrix.org | okay so there's seemingly nothing to gain from multiple executors on a single host | 19:49 |
| @fungicide:matrix.org | i would go so far as to say that if running more than one executor on the same machine gets you better performance and scalability, then that probably indicates a bug/design defect in the software | 19:50 |
| @clarkb:matrix.org | ya I don't think multiple processes on the same host gives you significantly more redundancy than a single executor process | 19:59 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/nodepool] 909607: Change the AWS default image volume-type from gp2 to gp3 https://review.opendev.org/c/zuul/nodepool/+/909607 | 21:04 | |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed wip on behalf of Lukas Kranz: [zuul/zuul-jobs] 887917: prepare-workspace-git: Add ability to define synced pojects https://review.opendev.org/c/zuul/zuul-jobs/+/887917 | 21:46 | |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: | 22:57 | |
| - [zuul/zuul] 895867: Handle output from socket commands https://review.opendev.org/c/zuul/zuul/+/895867 | ||
| - [zuul/zuul] 909622: Add zuul-scheduler status command https://review.opendev.org/c/zuul/zuul/+/909622 | ||
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: | 23:28 | |
| - [zuul/zuul] 895867: Handle output from socket commands https://review.opendev.org/c/zuul/zuul/+/895867 | ||
| - [zuul/zuul] 909624: Add verbosity argument to zuul-executor verbose https://review.opendev.org/c/zuul/zuul/+/909624 | ||
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 909622: Add zuul-scheduler status command https://review.opendev.org/c/zuul/zuul/+/909622 | 23:28 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!