| @blaisep-sureify:matrix.org | Hi corvus , I am going to look into the issue | 13:06 |
|---|---|---|
| https://opendev.org/zuul/zuul-operator/commit/5bf96c1e30d034bd6f0c8eccf1103b8ed2b338ff | ||
| it may have to do with https://kubernetes.io/docs/reference/using-api/deprecation-guide/#customresourcedefinition-v122 | ||
| @clarkb:matrix.org | Blaise Pabon: see https://review.opendev.org/c/zuul/nodepool/+/836860 you should be able to do something similar on the zuul operator side | 13:15 |
| @blaisep-sureify:matrix.org | Thank you Clark , (I'm new here... old just about everywhere else) | 14:58 |
| shall I try to submit a pr for the operator CRD file? | ||
| @clarkb:matrix.org | Blaise Pabon: yes I think you can push an update to Gerrit to fix the operator. If you remove the minikube pin at the same time then CI should test it for you | 15:02 |
| @blaisep-sureify:matrix.org | > <@clarkb:matrix.org> Blaise Pabon: see https://review.opendev.org/c/zuul/nodepool/+/836860 you should be able to do something similar on the zuul operator side | 15:04 |
| Oh, I also noticed in | ||
| https://opendev.org/zuul/zuul-operator/src/commit/5bf96c1e30d034bd6f0c8eccf1103b8ed2b338ff/.zuul.yaml#L13 | ||
| that it uses `runtime: docker` | ||
| should I change that to something modern, like `podman` or `containerd` ? | ||
| @clarkb:matrix.org | > <@blaisep-sureify:matrix.org> Oh, I also noticed in | 16:17 |
| > https://opendev.org/zuul/zuul-operator/src/commit/5bf96c1e30d034bd6f0c8eccf1103b8ed2b338ff/.zuul.yaml#L13 | ||
| > that it uses `runtime: docker` | ||
| > should I change that to something modern, like `podman` or `containerd` ? | ||
| I'm not sure. I've tried to find what uses the container_runtime var and haven't had any luck. It may be for running a registry via docker-compose and not related to the minikube k8s install. | ||
| @clarkb:matrix.org | I wonder if it is meant to be kubernetes_runtime not container_runtime. That defaults to docker though so same behavior currnetly now if so | 16:19 |
| @hanson76:matrix.org | > <@clarkb:matrix.org> Hanson: ^ I'm not sure if that is a complete fix. Feel free to update the change as necessary if you like. I don't have a kubernetes to test with beyond the functional test job that zuul will run for us | 19:53 |
| That plus https://review.opendev.org/c/zuul/nodepool/+/836875/1 seems to work. | ||
| @jpew:matrix.org | Hmm, interesting. If you run multiple nodepool-launchers, `min-ready` is per instances, so you actually get `min-ready` * number_of_nodepool-lauchers nodes sitting around. Is this intentional? | 20:04 |
| @jpew:matrix.org | * Hmm, interesting. If you run multiple nodepool-launchers, `min-ready` is per instance, so you actually get `min-ready` * number_of_nodepool-lauchers nodes sitting around. Is this intentional? | 20:04 |
| @clarkb:matrix.org | I don't know how intentional it was at the time nodepool was converted to zookeeper but it is known and why we set min ready on a single launcher. But we also set min ready to low values like 1 so the impact is low | 20:07 |
| @jim:acmegating.com | not intentional as such, but also not considered a bug or anti-design -- the protocol is designed to require zero cooperation, so they occasionally race each other and don't really care. | 20:11 |
| @jim:acmegating.com | in addition to what Clark suggested, https://zuul-ci.org/docs/nodepool/latest/configuration.html#attr-labels.max-ready-age may be useful | 20:13 |
| -@gerrit:opendev.org- Clark Boylan proposed: [zuul/nodepool] 837035: Update DIB to version 3.20.2 https://review.opendev.org/c/zuul/nodepool/+/837035 | 21:42 | |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 835464: Add a blob store and store large secrets in it https://review.opendev.org/c/zuul/zuul/+/835464 | 22:15 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!