| @clarkb:matrix.org | so I think not using the pass through proxy has helped but we're still hitting the limits (which I half expected) | 00:03 |
|---|---|---|
| @clarkb:matrix.org | https://zuul.opendev.org/t/zuul/build/51e2ec57ce13411c940ccc99d98d2496 this one failed trying to fetch the registry container image | 00:04 |
| @jim:acmegating.com | Clark: should we give it some more time before we call that assesment final? especially with the ambiguity around the actual limits and what might trigger them that you noticed... | 00:09 |
| @clarkb:matrix.org | oh good point ya maybe we should observe for a bit longer | 00:12 |
| @clarkb:matrix.org | the success rate does seem higher though. I rechecked 3 changes and they all have at least two jobs doing container things and only one job failed | 00:12 |
| @clarkb:matrix.org | I approved the buildest registry location configuration change in hopes that it iwll land now | 00:36 |
| @clarkb:matrix.org | all the jobs but one passed in check so I'm hopeful and that will reduce docker pulls (since zuul-registry is in quay now) | 00:37 |
| @clarkb:matrix.org | still failed. We can try again tomorrow. | 00:55 |
| @mnaser:matrix.org | Is it possible that Podman uses something else for pulls ? | 01:25 |
| @clarkb:matrix.org | I'm not sure I understand the question | 03:45 |
| @clarkb:matrix.org | Podman uses the libcontainer skopeo related stuff | 03:45 |
| -@gerrit:opendev.org- Benedikt Löffler proposed: [zuul/nodepool] 935881: Set priority in create_fleet call https://review.opendev.org/c/zuul/nodepool/+/935881 | 11:50 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!