10 Oct 2021 |
proletarius101 | In reply to @proletarius101:matrix.org Or maybe we just dynamically generate ci jobs which will also bypass the issue I can explore this if you think it's the best way | 08:10:15 |
linsui | Generate a CI job for every build? | 08:12:51 |
linsui | After upgrading to bulleye we'll drop many apt packages. Maybe that also helps. | 08:14:03 |
proletarius101 | In reply to @rdfg77:kde.org Generate a CI job for every build? I mean in each fdroid build jobs, generate sub jobs for each version number | 08:41:22 |
proletarius101 | That's a gitlab feature | 08:41:29 |
proletarius101 | In reply to @rdfg77:kde.org After upgrading to bulleye we'll drop many apt packages. Maybe that also helps. Yeah indeed | 08:41:36 |
proletarius101 | Debian dependencies takes ~3GB | 08:42:13 |
proletarius101 | In reply to @proletarius101:matrix.org I mean in each fdroid build jobs, generate sub jobs for each version number That's what we do on server | 08:44:25 |
proletarius101 | https://docs.gitlab.com/ee/ci/pipelines/parent_child_pipelines.html#dynamic-child-pipelines | 08:45:44 |
| outsidecontext_ joined the room. | 10:35:48 |
| mimi89999 left the room. | 14:32:55 |
| mimi89999 joined the room. | 14:37:13 |
linsui | I guess we need to email Ciaran to fix https://monitor.f-droid.org/builds/log/app.organicmaps/21100707#site-footer | 14:53:41 |
jochensp | linsui: yes, can you do that? | 15:03:15 |
jochensp | linsui: for the future: https://gitlab.com/fdroid/fdroidserver/-/merge_requests/1026 | 15:17:03 |
linsui | In reply to @_oftc_jochensp:matrix.org linsui: yes, can you do that? Done.👌 | 15:37:39 |
jochensp | thx :) | 15:37:47 |
Izzy | Looks like issuebot on RFP has some problems with docker: https://gitlab.com/fdroid/rfp/-/jobs/1666049183 There are again issues where it didn't show up for 5 days now. | 21:13:17 |
Izzy | "Cannot connect to the Docker daemon" | 21:13:45 |
Izzy | Funny. That time frame matches the day when I removed the bot label from Zulip (RFP 386) – and Zulip is the last thing mentioned in the pipeline logs. Let's see if putting the label back (and thus having Zulip skipped) fixes it. | 21:40:05 |
Licaon_Kter[xmpp] | It hates email based group chat software? :) | 21:41:08 |
Izzy | Yupp. Now it works again. | 21:48:05 |
Izzy | So please do not remove the bot label again from that RFP… Uh-oh. Lichess now looks like the next one. I already see the indicator: "bash: npm: command not found" (that was the last thing Zulip showed). | 21:51:24 |
Izzy | So maybe someone with the proper experience can take a look why the bot runs "npm install" without having installed npm first? | 21:52:35 |
Izzy | That said, it might well be something different. The very last log entry (the one it stalls on) is "Running /builds/fdroid/rfp/modules/active-hostnames.py" | 21:57:54 |
Izzy | Bingo: "Cannot connect to the Docker daemon at tcp://10.10.24.205:2376" | 22:10:26 |
Izzy | It will try for another 10..20 min now, and then fail. | 22:11:23 |
Izzy | Done: failed. It's cleaning up now. So issuebot ran once now, and will fail until either fixed or lichess is also skipped. It will fail whenever there's an APK successfully generated so it tries to scan for hostnames. | 22:16:35 |
11 Oct 2021 |
| zx2c4_ joined the room. | 01:27:33 |
proletarius101 | do we have a way to * 5 in the version code of UpdateCheckdata ? | 06:08:18 |