lissyx | hello, | 10:22 |
---|---|---|
lissyx | bandali, are you still on holidays ? | 10:23 |
bandali | hey lissyx, no i got back on monday, but now seb is away on holidays this week :p shall we skip meeting today or do you still want to meet? | 13:20 |
bittin_guadec | and other people being at GUADEC | 13:20 |
lissyx | bandali, nah I have a conflicting meeting (about snap though) | 13:20 |
lissyx | bandali, I was wondering if you had time to look into the github nightly failure :( | 13:21 |
bandali | lissyx, ack. as for nightly failure, not yet sorry, but i plan doing so today. it *seems* like it might be a memory-related issue? my first thought for a workaround is to try switching nightly away from using rustup like the other branches and see if it helps | 13:24 |
lissyx | it's OOM | 13:24 |
lissyx | I'm not sure how rustup is related ? | 13:24 |
lissyx | you need to use the appropriate rustc version | 13:25 |
lissyx | so you need it | 13:25 |
bandali | hmm i was going by https://github.com/canonical/firefox-snap/commit/2078640f64eae9f56d1a528335b39a0eca7848f5 | 13:26 |
-ubottu:#ubuntu-desktop- Commit 2078640 in canonical/firefox-snap "Do not use rustup to pull the rust toolchain, to work around https://forum.snapcraft.io/t/armhf-builds-on-launchpad-timing-out/31008." | 13:26 | |
lissyx | this was a network issue I think | 13:26 |
bandali | i think it's worth a try if anything, but yeah it might indeed not do much help for this case | 13:26 |
bandali | hm | 13:26 |
lissyx | the current issue is textbook OOM | 13:26 |
bandali | right. i mean otherwise i'd probably start with a revert of my commit bumping it to core22 and then try to figure out what could be causing it | 13:27 |
* bandali sighs | 13:27 | |
lissyx | I'm not sure it's also related to the core22 switch | 13:27 |
lissyx | 7GB of RAM is really low | 13:28 |
lissyx | pretty sure the only viable solution at taht point is reducing build parallelism on github actions | 13:28 |
bandali | right.. if https://docs.github.com/en/actions/using-github-hosted-runners/about-github-hosted-runners#supported-runners-and-hardware-resources is right, then we're already at only 2 cores, and the only step down would be a single-core build | 13:33 |
bandali | worth a shot i guess | 13:33 |
bandali | lissyx, btw i'm still hoping to switch stable from core20 to core22 this week, especially if i can get the nightly core22 to build. does that work for y'all? | 13:36 |
bandali | lissyx, or would you prefer i wait until after 116.0 stable is out? | 13:36 |
lissyx | you switched core22 on nightly recently right? | 13:38 |
lissyx | we should probably wait for it to ride beta as well no? | 13:38 |
bandali | lissyx, yeah but afaik we haven't had a successful nightly build after the switch. we do have them for beta too, though | 13:38 |
lissyx | either way I will be unavailable starting july 28th | 13:38 |
lissyx | well, not unreachable ... | 13:39 |
bandali | right.. any preference/suggestion for how to go about it? | 13:39 |
lissyx | bandali, for whaat it is worth, my work of doing builds on treeherder landed and cron jobs are arriving as well | 13:39 |
lissyx | we would get at least some build coverage | 13:39 |
lissyx | for the timeline, I'd say it's up to you | 13:40 |
bandali | i see. yeah i'd have more confidence once we've had successful nightly core22 builds as well and nothing is on fire | 13:42 |
bandali | plus, seb is away this week, so i can't have him quickly do something on the store side if absolutely necessary | 13:42 |
bandali | so i think i'll postpone the switch to core22 until at least next week | 13:43 |
bandali | re landing those and having build coverage: nice :) | 13:43 |
lissyx | bandali, ok so either the limiting in: "SNAPCRAFT_BUILD_ENVIRONMENT_MEMORY=6G SNAPCRAFT_BUILD_ENVIRONMENT_CPU=2 snap run snapcraft --use-lxd" does not work, or my local LXD behaves differently than GitHub Actions :( | 13:55 |
bandali | lissyx, huh :/ i'll try to investigate later today | 14:22 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!