| Ukikie | vorlon, Laney: Hey, would you guys like to try out ubot3 in here? It's a clone like ubot5, but has support for gitlab/github issues as well as commits from those and cgit (though that can be turned off.) | 00:14 |
|---|---|---|
| teward | ummmmm.... 20.04.1's zsync for server ISOs seems... wrong/bad... | 01:04 |
| teward | zsync says it can't parse it and it's straight binary data | 01:04 |
| teward | neither of the zsync files work | 01:05 |
| teward | for 20.04.1 | 01:05 |
| teward | is that normal? | 01:05 |
| teward | nevermind sounds like an HTTPS problem >.> | 01:08 |
| -queuebot:#ubuntu-release- New binary: xfce4-settings [s390x] (hirsute-proposed/universe) [4.15.2-2ubuntu1] (xubuntu) | 02:35 | |
| -queuebot:#ubuntu-release- New binary: xfce4-settings [ppc64el] (hirsute-proposed/universe) [4.15.2-2ubuntu1] (xubuntu) | 02:36 | |
| -queuebot:#ubuntu-release- New binary: xfce4-settings [amd64] (hirsute-proposed/universe) [4.15.2-2ubuntu1] (xubuntu) | 02:37 | |
| -queuebot:#ubuntu-release- New binary: xfce4-settings [arm64] (hirsute-proposed/universe) [4.15.2-2ubuntu1] (xubuntu) | 02:39 | |
| -queuebot:#ubuntu-release- New binary: xfce4-settings [armhf] (hirsute-proposed/universe) [4.15.2-2ubuntu1] (xubuntu) | 02:45 | |
| -queuebot:#ubuntu-release- New binary: libcommonmark-perl [ppc64el] (hirsute-proposed/none) [0.290000-1] (no packageset) | 05:17 | |
| -queuebot:#ubuntu-release- New binary: wtype [ppc64el] (hirsute-proposed/none) [0.2-1] (no packageset) | 05:17 | |
| -queuebot:#ubuntu-release- New binary: libnet-dns-native-perl [ppc64el] (hirsute-proposed/none) [0.22-1] (no packageset) | 05:17 | |
| -queuebot:#ubuntu-release- New binary: wtype [s390x] (hirsute-proposed/none) [0.2-1] (no packageset) | 05:17 | |
| -queuebot:#ubuntu-release- New binary: aravis [s390x] (hirsute-proposed/universe) [0.8.1-1] (no packageset) | 05:18 | |
| -queuebot:#ubuntu-release- New binary: fcitx5-chewing [s390x] (hirsute-proposed/none) [5.0.1-1] (no packageset) | 05:18 | |
| -queuebot:#ubuntu-release- New binary: fcitx5-chewing [ppc64el] (hirsute-proposed/none) [5.0.1-1] (no packageset) | 05:18 | |
| -queuebot:#ubuntu-release- New binary: libnet-dns-native-perl [amd64] (hirsute-proposed/none) [0.22-1] (no packageset) | 05:18 | |
| -queuebot:#ubuntu-release- New binary: aravis [ppc64el] (hirsute-proposed/universe) [0.8.1-1] (no packageset) | 05:19 | |
| -queuebot:#ubuntu-release- New binary: php-league-mime-type-detection [amd64] (hirsute-proposed/universe) [1.5.1+ds-1] (no packageset) | 05:19 | |
| -queuebot:#ubuntu-release- New binary: libcommonmark-perl [amd64] (hirsute-proposed/universe) [0.290000-1] (no packageset) | 05:19 | |
| -queuebot:#ubuntu-release- New binary: wtype [amd64] (hirsute-proposed/universe) [0.2-1] (no packageset) | 05:19 | |
| -queuebot:#ubuntu-release- New binary: libcommonmark-perl [s390x] (hirsute-proposed/universe) [0.290000-1] (no packageset) | 05:20 | |
| -queuebot:#ubuntu-release- New binary: libnet-dns-native-perl [s390x] (hirsute-proposed/universe) [0.22-1] (no packageset) | 05:20 | |
| -queuebot:#ubuntu-release- New binary: aravis [amd64] (hirsute-proposed/universe) [0.8.1-1] (no packageset) | 05:21 | |
| -queuebot:#ubuntu-release- New binary: node-make-error [amd64] (hirsute-proposed/universe) [1.3.6-1] (no packageset) | 05:22 | |
| -queuebot:#ubuntu-release- New binary: fcitx5-chewing [amd64] (hirsute-proposed/universe) [5.0.1-1] (no packageset) | 05:23 | |
| -queuebot:#ubuntu-release- New binary: ruby-terser [amd64] (hirsute-proposed/universe) [1.0.2+dfsg-1] (no packageset) | 05:23 | |
| -queuebot:#ubuntu-release- New binary: wtype [riscv64] (hirsute-proposed/universe) [0.2-1] (no packageset) | 05:26 | |
| -queuebot:#ubuntu-release- New binary: libcommonmark-perl [riscv64] (hirsute-proposed/universe) [0.290000-1] (no packageset) | 05:28 | |
| -queuebot:#ubuntu-release- New binary: libnet-dns-native-perl [riscv64] (hirsute-proposed/universe) [0.22-1] (no packageset) | 05:28 | |
| -queuebot:#ubuntu-release- New binary: fcitx5-chewing [riscv64] (hirsute-proposed/universe) [5.0.1-1] (no packageset) | 05:30 | |
| -queuebot:#ubuntu-release- New binary: fcitx5-chewing [armhf] (hirsute-proposed/universe) [5.0.1-1] (no packageset) | 05:35 | |
| -queuebot:#ubuntu-release- New binary: wtype [armhf] (hirsute-proposed/universe) [0.2-1] (no packageset) | 05:35 | |
| -queuebot:#ubuntu-release- New binary: libcommonmark-perl [arm64] (hirsute-proposed/universe) [0.290000-1] (no packageset) | 05:37 | |
| -queuebot:#ubuntu-release- New binary: wtype [arm64] (hirsute-proposed/universe) [0.2-1] (no packageset) | 05:37 | |
| -queuebot:#ubuntu-release- New binary: libcommonmark-perl [armhf] (hirsute-proposed/universe) [0.290000-1] (no packageset) | 05:37 | |
| -queuebot:#ubuntu-release- New binary: libnet-dns-native-perl [arm64] (hirsute-proposed/universe) [0.22-1] (no packageset) | 05:37 | |
| -queuebot:#ubuntu-release- New binary: fcitx5-chewing [arm64] (hirsute-proposed/universe) [5.0.1-1] (no packageset) | 05:38 | |
| -queuebot:#ubuntu-release- New binary: libnet-dns-native-perl [armhf] (hirsute-proposed/universe) [0.22-1] (no packageset) | 05:39 | |
| -queuebot:#ubuntu-release- New binary: aravis [armhf] (hirsute-proposed/universe) [0.8.1-1] (no packageset) | 05:44 | |
| -queuebot:#ubuntu-release- New binary: aravis [arm64] (hirsute-proposed/universe) [0.8.1-1] (no packageset) | 05:44 | |
| -queuebot:#ubuntu-release- New binary: aravis [riscv64] (hirsute-proposed/universe) [0.8.1-1] (no packageset) | 05:46 | |
| -queuebot:#ubuntu-release- New binary: xfce4-settings [riscv64] (hirsute-proposed/universe) [4.15.2-2ubuntu1] (xubuntu) | 05:50 | |
| -queuebot:#ubuntu-release- New: accepted aravis [arm64] (hirsute-proposed) [0.8.1-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted aravis [riscv64] (hirsute-proposed) [0.8.1-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted fcitx5-chewing [armhf] (hirsute-proposed) [5.0.1-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted libcommonmark-perl [arm64] (hirsute-proposed) [0.290000-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted libcommonmark-perl [riscv64] (hirsute-proposed) [0.290000-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted libnet-dns-native-perl [armhf] (hirsute-proposed) [0.22-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted ruby-terser [amd64] (hirsute-proposed) [1.0.2+dfsg-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted wtype [armhf] (hirsute-proposed) [0.2-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted xfce4-settings [riscv64] (hirsute-proposed) [4.15.2-2ubuntu1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted aravis [armhf] (hirsute-proposed) [0.8.1-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted fcitx5-chewing [riscv64] (hirsute-proposed) [5.0.1-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted libnet-dns-native-perl [arm64] (hirsute-proposed) [0.22-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted wtype [arm64] (hirsute-proposed) [0.2-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted fcitx5-chewing [arm64] (hirsute-proposed) [5.0.1-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted libnet-dns-native-perl [riscv64] (hirsute-proposed) [0.22-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted libcommonmark-perl [armhf] (hirsute-proposed) [0.290000-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted wtype [riscv64] (hirsute-proposed) [0.2-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted aravis [amd64] (hirsute-proposed) [0.8.1-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted aravis [s390x] (hirsute-proposed) [0.8.1-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted bazel-bootstrap [arm64] (hirsute-proposed) [3.4.0+ds-2] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted bazel-bootstrap [s390x] (hirsute-proposed) [3.4.0+ds-2] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted fcitx5-chewing [ppc64el] (hirsute-proposed) [5.0.1-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted libcommonmark-perl [amd64] (hirsute-proposed) [0.290000-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted libcommonmark-perl [s390x] (hirsute-proposed) [0.290000-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted libnet-dns-native-perl [ppc64el] (hirsute-proposed) [0.22-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted node-make-error [amd64] (hirsute-proposed) [1.3.6-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted wtype [amd64] (hirsute-proposed) [0.2-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted aravis [ppc64el] (hirsute-proposed) [0.8.1-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted bazel-bootstrap [ppc64el] (hirsute-proposed) [3.4.0+ds-2] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted fcitx5-chewing [s390x] (hirsute-proposed) [5.0.1-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted libnet-dns-native-perl [amd64] (hirsute-proposed) [0.22-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted php-league-mime-type-detection [amd64] (hirsute-proposed) [1.5.1+ds-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted wtype [s390x] (hirsute-proposed) [0.2-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted xfce4-settings [arm64] (hirsute-proposed) [4.15.2-2ubuntu1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted xfce4-settings [ppc64el] (hirsute-proposed) [4.15.2-2ubuntu1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted bazel-bootstrap [amd64] (hirsute-proposed) [3.4.0+ds-2] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted libcommonmark-perl [ppc64el] (hirsute-proposed) [0.290000-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted wtype [ppc64el] (hirsute-proposed) [0.2-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted xfce4-settings [armhf] (hirsute-proposed) [4.15.2-2ubuntu1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted fcitx5-chewing [amd64] (hirsute-proposed) [5.0.1-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted xfce4-settings [amd64] (hirsute-proposed) [4.15.2-2ubuntu1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted libnet-dns-native-perl [s390x] (hirsute-proposed) [0.22-1] | 07:40 | |
| -queuebot:#ubuntu-release- New: accepted xfce4-settings [s390x] (hirsute-proposed) [4.15.2-2ubuntu1] | 07:41 | |
| doko | how did the autopkg queues go down from 3000-6000 to <1000 within two hours? | 10:26 |
| Laney | I really purged those jobs I said I purged yesterday and actually didn't | 10:27 |
| doko | Laney: please could you also give back all the /unkown ones? | 10:33 |
| Laney | I think all the in progress or regression perls should be retried, will do that shortly | 10:34 |
| doko | ta | 10:35 |
| Laney | doing | 10:38 |
| doko | the queues look "better" again, 4k+ ;p | 11:34 |
| -queuebot:#ubuntu-release- Unapproved: busybox (groovy-proposed/main) [1:1.30.1-4ubuntu9 => 1:1.30.1-4ubuntu9.1] (core, i386-whitelist) | 12:18 | |
| juliank | Laney: Seems armhf caught up now, so I'll retire the additional lxd remote | 12:29 |
| Laney | okey | 12:32 |
| doko | I wouldn't mind if the armhf queue gets empty earlier, then you can retry armhf stuff which fails under load | 12:32 |
| doko | mwhudson: the execuses_by_team report now has some very long "Requires:" lines ... maybe you could only list the first few ones, and then refer to update_excuses? | 12:34 |
| juliank | Laney: I removed it from lxd remotes and ran juju set, will the workers go away automatically eventually? | 12:34 |
| juliank | I should probably send them SIGHUP | 12:34 |
| Laney | unlikely | 12:35 |
| Laney | well, they go away if the socat one does iirc | 12:35 |
| Laney | stop, reset-failed, disable that one, then see what happens | 12:35 |
| doko | do you know how the i386 hints are normally handled? | 12:35 |
| Laney | might hav eto do that for the workers too | 12:35 |
| Laney | that stuff is buggy | 12:35 |
| Laney | doko: by vorlon usually, not sure if there's a script | 12:36 |
| juliank | If I keep the lxd worker instance around as stopped, does that affect our quota? | 12:40 |
| Laney | yep | 12:40 |
| juliank | OK, all arm stuff should be back to its usual self | 12:41 |
| juliank | doko: failures under load should be ignored not retried, full load is what we're aiming for anyway | 12:42 |
| juliank | Like once we have continuous baseline retesting, workers will be fully loaded all the time | 12:43 |
| Laney | juliank: which one did you delete, 45.21? | 12:43 |
| juliank | yes | 12:43 |
| Laney | units are still there | 12:43 |
| Laney | probably need to smite them manually | 12:44 |
| Laney | think the new implementation fixes that ... I started rebasing that the other day | 12:44 |
| juliank | Laney: They are disabled however, it should just be systemd remnants that need cleaning | 12:44 |
| juliank | well confusing stuff | 12:45 |
| juliank | they certainly were stopped but now are running again | 12:45 |
| doko | juliank: you're aiming for overload, not full load | 12:45 |
| Laney | make sure to disable them | 12:45 |
| -queuebot:#ubuntu-release- Unapproved: busybox (focal-proposed/main) [1:1.30.1-4ubuntu6.2 => 1:1.30.1-4ubuntu6.3] (core, i386-whitelist) | 12:46 | |
| juliank | Seem gone now | 12:47 |
| juliank | The worker processes got stuck | 12:47 |
| juliank | doko: well we want full load on our configuration, but that is overload of the quota | 12:48 |
| juliank | :D | 12:48 |
| rbalint | Laney, has there been a mass badtest hinting for the i386 false passes due to using amd64? | 12:57 |
| rbalint | Laney, if not i prep one | 12:57 |
| Laney | no, that would be welcome, it should be only bumping of previous hints I guess | 12:57 |
| rbalint | Laney, no, there are many missing hints see https://code.launchpad.net/~rbalint/britney/+git/hints-ubuntu/+merge/393510 | 13:04 |
| Laney | rbalint: please scan the file, some of those are already in there and should be bumped, and in general I would prefer force-reset-test for these | 13:11 |
| rbalint | Laney, why? | 13:11 |
| Laney | why to which part? | 13:11 |
| Laney | if the latter, because we *accidentally* ran these and this is closest to pretending that didn't happen | 13:12 |
| rbalint | why preferring force-reset-test over badtesting all/i386 | 13:12 |
| rbalint | Laney, even triggering the tests that did not pass in focal on i386 is a waste of resources | 13:13 |
| Laney | even if that's right, force-badtest vs force-reset-test doesn't help there | 13:15 |
| rbalint | Laney, they are dead, Jim :-) | 13:15 |
| Laney | they are executed anyway | 13:15 |
| rbalint | Laney, not now, but i'll propose not triggering badtests by default in a different topic | 13:15 |
| rbalint | Laney, for all arches | 13:16 |
| rbalint | Laney, but running periodic full-archive tests in an additional queue for tests not triggered for a long time | 13:17 |
| Laney | if we get to that point, we would want to evaluate the hints then | 13:17 |
| Laney | right | 13:17 |
| Laney | don't go rogue on that plan but implement it in consultation with the rest of us | 13:18 |
| Laney | it's been discussed with juliank and vorlon for a long time | 13:18 |
| rbalint | Laney, i can't is don't have a separate queue :-) | 13:18 |
| rbalint | Laney, the right step now is marking the false positive tests on i386 consistently, so the next steps will be easier | 13:18 |
| Laney | I don't think it holds that these ones should be picked for skipping forever because we accidentally ran them in this short period, the analysis needs to be more sophisticated than that | 13:20 |
| rbalint | Laney, bumping reset tests is just an ad-hoc patch over the false positives | 13:20 |
| Laney | please let's just use reset-test, and discuss your other idea in a measured way | 13:20 |
| Laney | rather than mixing the two things | 13:20 |
| Laney | gotta go eat lunch now | 13:21 |
| rbalint | Laney, LP: #1903867 | 14:09 |
| ubot5 | Launchpad bug 1903867 in Auto Package Testing "False positive test results in Hirsute on i386" [Undecided,New] https://launchpad.net/bugs/1903867 | 14:09 |
| rbalint | Laney, i'm away for the afternoon, feel free to bump the tests | 14:10 |
| Laney | rbalint: cheers | 15:39 |
| Laney | and if you want to work on those other ideas, maybe schedule calls and we can chat about how that would be best to do | 15:39 |
| juliank | ugh | 15:53 |
| juliank | I'm building stuff in autopkgtest and it's not building in parallel | 15:53 |
| juliank | ah now that I complain it's finishing | 15:53 |
| vorlon | Laney, doko: there's some scripting (present as a comment in the ubuntu-release hint file) to handle i386 hints that are auto-detectable (but that britney doesn't correctly auto-detect), i.e. the only binary on i386 is arch: all. Otherwise it's manual. | 16:08 |
| Laney | nod | 16:11 |
| Laney | meanwhile I'll assume rbalint's list is correct and hint all of those :-) | 16:11 |
| vorlon | :) | 16:19 |
| -queuebot:#ubuntu-release- Unapproved: grub2 (hirsute-proposed/main) [2.04-1ubuntu36 => 2.04-1ubuntu36] (core) | 17:05 | |
| -queuebot:#ubuntu-release- Unapproved: grub2 (hirsute-proposed/main) [2.04-1ubuntu36 => 2.04-1ubuntu36] (core) | 17:25 | |
| -queuebot:#ubuntu-release- New binary: ruby-simpleidn [amd64] (hirsute-proposed/universe) [0.1.1-1] (no packageset) | 17:25 | |
| -queuebot:#ubuntu-release- New binary: ycm-cmake-modules [amd64] (hirsute-proposed/universe) [0.12.0-1] (no packageset) | 17:25 | |
| -queuebot:#ubuntu-release- Unapproved: darktable (groovy-proposed/universe) [3.2.1-2build2 => 3.2.1-3~ubuntu0.20.10.1] (ubuntustudio) | 17:44 | |
| Eickmeyer[m] | ^ SRU bug 1903892 | 17:45 |
| ubot5 | bug 1903892 in darktable (Ubuntu Groovy) "[SRU] darktable "export as web gallery" is broken" [High,Fix committed] https://launchpad.net/bugs/1903892 | 17:45 |
| === ijohnson is now known as ijohnson|lunch | ||
| -queuebot:#ubuntu-release- New binary: llvm-toolchain-11 [armhf] (focal-proposed/universe) [1:11.0.0-2~ubuntu20.04.1] (i386-whitelist) | 18:24 | |
| mwhudson | doko: i dispute the idea that because i worked on the first version of this script, i am responsible for it forever | 18:38 |
| === ijohnson|lunch is now known as ijohnson | ||
| doko | mwhudson: heh ... | 19:13 |
| rbalint | Laney, thanks for the hints | 19:20 |
| rbalint | Laney, could you please check my hints again? https://code.launchpad.net/~rbalint/britney/+git/hints-ubuntu/+merge/393626 | 19:21 |
| rbalint | Laney, they contained other hints than just the i386 ones | 19:21 |
| -queuebot:#ubuntu-release- New binary: llvm-toolchain-11 [arm64] (focal-proposed/universe) [1:11.0.0-2~ubuntu20.04.1] (i386-whitelist) | 20:18 | |
| rbalint | vorlon, Laney, juliank, doko about 5-10% of load reduction would be the result of skipping never passing tests LP: #1903913 | 21:04 |
| ubot5 | Launchpad bug 1903913 in Auto Package Testing "Don't run tests that never pass (or don't run tests with force-badtest)" [Undecided,New] https://launchpad.net/bugs/1903913 | 21:04 |
| juliank | rbalint: we have to try them eventually otherwise we won't notice once they start working | 21:05 |
| rbalint | juliank, please see the bug, also there is not much to loose not noticing that quickly | 21:05 |
| juliank | rbalint: force-badtest ignoring is ok IMO, though we should try running the tests at least once a month or so | 21:06 |
| rbalint | juliank, yes, i'm proposing that | 21:07 |
| juliank | Well if we run them at least once a month, we can ignore all always failed tests | 21:07 |
| juliank | Without having to rely on continuous baseline retesting | 21:07 |
| rbalint | juliank, yes | 21:07 |
| juliank | Just have britney only schedule the test for an always-failed if it had no results for a month, essentially | 21:07 |
| rbalint | retesting everything is ~1200h on amd64 and ~3000h on arm64 | 21:08 |
| rbalint | juliank, yes, and having a separate queue for that | 21:08 |
| juliank | rbalint: not a separate queue, no | 21:08 |
| rbalint | juliank, why? | 21:08 |
| rbalint | then which queue | 21:08 |
| juliank | The separate queue for the baseline retest | 21:08 |
| juliank | But I'm talking about the normal tests | 21:09 |
| juliank | e.g. foo has always failed | 21:09 |
| juliank | Bar triggers foo gets scheduled if we haven't run foo tests in a month | 21:09 |
| juliank | This avoids having to deal with the complicated continuous baseline retesting bits and extra queues | 21:09 |
| juliank | As an intermediate solution until we have baseline retesting, at which point we don't need that :( | 21:10 |
| juliank | Um :) | 21:10 |
| rbalint | if this is easier to implement, i like that | 21:10 |
| rbalint | bileto can always skip those | 21:10 |
| juliank | Depends on whether britney has the data for it | 21:11 |
| rbalint | juliank, imo the force-badtest data is good for that | 21:11 |
| juliank | That one is easy I guess | 21:11 |
| rbalint | do we have a plan? ;-) | 21:12 |
| juliank | But also doing that for any test that is always failed and not run in a month would be more effective. | 21:12 |
| juliank | The question here is whether britney knows when a test last ran | 21:12 |
| juliank | Or whether it only knows that the test always failed | 21:12 |
| rbalint | autopkgtest.db does not seem to have dates | 21:12 |
| juliank | I think it downloads the database from autopkgtesr | 21:13 |
| juliank | Hmm | 21:13 |
| juliank | Id expect there to be a timestamp given that we see dates in the histor | 21:13 |
| rbalint | juliank, hm, it is encoded in run_id | 21:14 |
| rbalint | so that would work :-) | 21:14 |
| juliank | Now write a merge proposal :) | 21:15 |
| juliank | Also I have to have another look at britney triggers | 21:16 |
| rbalint | i hopes someone more familiar with britney would do it ;-) | 21:16 |
| juliank | Because it adds triggers that are in release pocket and that makes stuff noisy | 21:16 |
| juliank | Like tests had 20 triggers for perl migration but all I looked at ex perl where in release pocket | 21:17 |
| juliank | And due to how this is implemented in autopkgtest, it might even lead to false results, certainly the version was ignored somewhat in the past, so it might pick proposed versions instead of the specified release one | 21:18 |
| rbalint | juliank, i can just agree :-) LP: #1897317 | 21:20 |
| ubot5 | Launchpad bug 1897317 in Auto Package Testing "Extra triggers may cause test failures, but retrying without them could help" [Undecided,New] https://launchpad.net/bugs/1897317 | 21:20 |
| rbalint | vorlon, doko unfortunately i need to upload glibc again, because the -prof changes FTBFS on riscv64, unless we can push in a binary version from bileto | 21:22 |
| vorlon | rbalint: well pushing binary versions from bileto is certainly allowed; it just doesn't avoid retriggering auotpkgtests | 21:22 |
| rbalint | vorlon, i mean a binary version only for riscv64 | 21:23 |
| vorlon | I don't follow how you would do that | 21:23 |
| vorlon | is it the same source version as in the archive? | 21:23 |
| rbalint | vorlon, i'd fix the source in bileto with the same version, build only for riscv64, copy the binaries | 21:24 |
| vorlon | yeah no :) | 21:24 |
| rbalint | vorlon, i can do it with the same source, but patched dpkg | 21:24 |
| vorlon | you can't copy binaries unless the source matches, and the source won't match | 21:24 |
| vorlon | why wouldn't that be solved then by publishing the patched dpkg to the main archive and retrying the riscv64 build there? | 21:25 |
| rbalint | because that dpkg would just add a hack to fix up the riscv build in glibc :-) | 21:26 |
| rbalint | it is not failing tests, it is a mistake in the packaging of glibc | 21:26 |
| rbalint | definitely cheating | 21:27 |
| rbalint | but i don't really want to do that, i'd rather have the test skipping logic in place as discussed above then land glibc normally with 5-10% less triggered tests | 21:29 |
| vorlon | rbalint: yeah, please no bileto builds that depend on build-deps that aren't going to the archive | 21:30 |
| vorlon | as convenient as it might be | 21:30 |
| vorlon | anyway, it's riscv64, so even publishing source to the main archive and building on all archs and running tests will probably finish before riscv64 build completes ;P | 21:31 |
| rbalint | vorlon, the riscv build became quite quick with the dpkg really passing nocheck, it is on par with arm64 :-) | 21:32 |
| rbalint | vorlon, so how do you like juliank's ideas about the test skipping logic for fixing LP: #1903913 ? | 21:33 |
| ubot5 | Launchpad bug 1903913 in Auto Package Testing "Don't run tests that never pass (or don't run tests with force-badtest)" [Undecided,New] https://launchpad.net/bugs/1903913 | 21:33 |
| rbalint | vorlon, i think they would be quick and efficient with an immedate return on next glibc upload | 21:34 |
| vorlon | rbalint: I'll have to look at that later, sorry | 21:35 |
| rbalint | vorlon, thanks, i'm EOD-ing anyway | 21:36 |
| tjaalton | new intel-microcode breaks dell systems https://bugs.launchpad.net/ubuntu/+source/intel-microcode/+bug/1903883 | 21:37 |
| ubot5 | Ubuntu bug 1903883 in intel-microcode (Ubuntu) "XPS 13 9310 Tiger Lake Unable to boot 20.10 after intel-microcode update 3.20201110.0ubuntu0.20.10.1" [Critical,Confirmed] | 21:37 |
| juliank | LG | 21:45 |
| vorlon | bdmurray, Laney: fwiw I fixed cron.NBS by cowboying archive-cruft-check back to python2.7; it depends on apt_pkg.version_compare for sorting, and I don't know of a python3-compatible key= equivalent | 22:07 |
| mwhudson | vorlon: the | 22:18 |
| mwhudson | vorlon: there's functools.cmp2key or something | 22:19 |
| mwhudson | vorlon: https://docs.python.org/3.8/library/functools.html#functools.cmp_to_key | 22:19 |
| mwhudson | (not the finest bit of python 3 design work) | 22:20 |
| vorlon | mwhudson: ah cheers | 22:41 |
| vorlon | mwhudson: looks like that does the trick | 22:54 |
| -queuebot:#ubuntu-release- New binary: ruby-simpleidn [amd64] (hirsute-proposed/universe) [0.1.1-2] (no packageset) | 23:17 | |
| -queuebot:#ubuntu-release- New binary: dcmtk [s390x] (hirsute-proposed/universe) [3.6.5-1] (ubuntustudio) | 23:27 | |
| -queuebot:#ubuntu-release- New binary: dcmtk [ppc64el] (hirsute-proposed/universe) [3.6.5-1] (ubuntustudio) | 23:35 | |
| -queuebot:#ubuntu-release- New binary: dcmtk [amd64] (hirsute-proposed/universe) [3.6.5-1] (ubuntustudio) | 23:40 | |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!