| corvus | the reboot on ze12 was stuck due to a paused job that never resumed; i'm not going to look into the forensics now, but i restarted it manually to let the playbook continue. | 15:37 |
|---|---|---|
| corvus | the image build jobs failed to upload to swift; and i have two suspects as to why: 1) the container did not exist. i'm not sure if it was accidentally deleted, or maybe it was automatically deleted because it was empty? 2) we have more than one project now; i wonder if we need to specify the project in our authentication (which doesn't use clouds.yaml -- this is the upload in the job) | 17:29 |
| corvus | i recreated the container and the jobs are running again | 17:31 |
| clarkb | I don't think any of the changes we made so far should have deleted any conatiners. And we didn't update the tenant yet so wouldn't be seeing the new empty tenant | 18:42 |
| corvus | i see two tenants/projects in the web ui -- i'm assuming one of them is the new one. so my hypothesis about authentication is that if we only had one before, then we may need to specify which tenant/project now | 18:55 |
| corvus | first job just finished, so i guess it was the missing container, not the auth theory. | 18:55 |
| fungi | i think you always have to specify a project name or id to interact with project-scoped resources, don't you? | 19:02 |
| fungi | but also that's a great point, do i need to add anything to dfw3 and sjc3 for zuul to be able to upload niz images to swift there before we merge 942231? | 19:04 |
| corvus | https://opendev.org/opendev/zuul-providers/src/branch/master/playbooks/opendev-build-diskimage-base/post-inner.yaml#L17-L23 | 19:23 |
| corvus | that's the auth information we're providing now; maybe the app credential is scoped to the project? but it shows up in both when i switch between them using the toggle at the top in the webui. | 19:24 |
| corvus | maybe it just defaults to the first? | 19:26 |
| Clark[m] | sorry I didn't followup on this my irc bouncer machine died according to the cloud provider. The server is back up but I'm not in a good spot to restore the client yet | 21:50 |
| Clark[m] | I'll have to do that tomorrow likely (I want to go through the steps of updating weechat since it died anyway) | 21:50 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!