| juergbi | jjardon, valentind: we still have a scheduled CI pipeline for bst-1.2. does this make sense or shall we drop it? that branch doesn't appear to be maintained anymore | 05:06 |
|---|---|---|
| gitlab-br-bot | juergbi opened MR !1859 (juerg/ci-schedules->master: .gitlab-ci.yml: Don't run win32 and wsl tests in scheduled pipelines) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/1859 | 05:06 |
| *** BarbaraPalv has joined #buildstream | 06:07 | |
| *** BarbaraPalv has quit IRC | 06:11 | |
| *** tristan_ has quit IRC | 06:37 | |
| *** tristan_ has joined #buildstream | 06:37 | |
| *** ChanServ sets mode: +o tristan_ | 06:38 | |
| *** benschubert has joined #buildstream | 06:46 | |
| gitlab-br-bot | BenjaminSchubert approved MR !1858 (juerg/export-to-tar->master: Improve Directory.export_to_tar()) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/1858 | 06:48 |
| *** mohan43u has quit IRC | 06:58 | |
| *** mohan43u has joined #buildstream | 08:16 | |
| *** rdale has joined #buildstream | 08:17 | |
| gitlab-br-bot | marge-bot123 merged MR !1858 (juerg/export-to-tar->master: Improve Directory.export_to_tar()) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/1858 | 08:25 |
| *** santi has joined #buildstream | 08:25 | |
| *** tpollard has joined #buildstream | 08:31 | |
| *** lachlan has joined #buildstream | 08:56 | |
| *** slaf has quit IRC | 09:19 | |
| *** narispo has quit IRC | 09:33 | |
| *** slaf has joined #buildstream | 09:56 | |
| jjardon | juergbi: yeah let's drop it | 09:58 |
| juergbi | ok, will do | 09:58 |
| jjardon | juergbi: I have paused it now | 09:58 |
| juergbi | ah, you already did | 09:59 |
| jjardon | valentind: ^ | 09:59 |
| juergbi | ok | 09:59 |
| juergbi | jjardon: I assume !1859 makes sense to you as well, or am I missing something? | 09:59 |
| gitlab-br-bot | jjardon approved MR !1859 (juerg/ci-schedules->master: .gitlab-ci.yml: Don't run win32 and wsl tests in scheduled pipelines) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/1859 | 10:00 |
| jjardon | yup | 10:00 |
| valentind | Sorry for the delay. I agree. We can drop it. | 10:01 |
| jjardon | note the wsl are currently failing though: https://gitlab.com/BuildStream/buildstream/-/issues/1277 (make them allow to fail for now so master pipeline passes) | 10:02 |
| juergbi | jjardon: last run completed: https://gitlab.com/BuildStream/buildstream/-/jobs/503132884 | 10:21 |
| tristan_ | Is there existing instructions to install buildbox ? | 10:27 |
| tristan_ | I'm trying to generate example docs and it's not automatically installed by tox (obviously) | 10:28 |
| juergbi | tristan_: https://buildstream.build/buildbox_install.html | 10:28 |
| tristan_ | thanks :) | 10:28 |
| tristan_ | this binary tarball seems difficult to download :-/ | 10:36 |
| tristan_ | my connection appeared to be fine, I wonder if hosting has a problem | 10:36 |
| juergbi | it's a digitalocean 'space'. seems to work fine here | 10:37 |
| juergbi | tristan_: here is the non-CDN link, not sure whether that makes a difference: https://buildbox-casd-binaries.nyc3.digitaloceanspaces.com/buildbox-x86_64-linux-0.0.7-253cb8d6.tar.xz | 10:38 |
| tristan_ | that works much better for me here thanks | 10:39 |
| * tristan_ tried the other link 3 times and got stalled | 10:39 | |
| gitlab-br-bot | jjardon opened MR !1860 (jjardon/wsl->master: Revert ".gitlab-ci.yml: Allow WSL test to fail") on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/1860 | 10:43 |
| tristan_ | tox -e docs fails because bst-plugins-experimental does not have ostree ? | 10:43 |
| tristan_ | How can that fail in a tox setting ? | 10:44 |
| tristan_ | weird, I have a clear failure for the flatpak-autotools example, but the overall docs build is a success | 10:45 |
| * tristan_ will give examples some love | 10:46 | |
| WSalmon | git+https://gitlab.com/BuildStream/bst-plugins-experimental.git@be5ac19e5062bc23a46ed8ce7aa2958a2653c917#egg=bst_plugins_experimental[ostree] | 10:46 |
| WSalmon | what dose the egg bit do | 10:47 |
| WSalmon | dose it trigger https://gitlab.com/BuildStream/bst-plugins-experimental/-/blob/master/setup.py#L92 | 10:47 |
| benschubert | it should yes | 10:50 |
| juergbi | oh, 101 droplets :-/ | 10:57 |
| juergbi | it seems there is still an issue that droplets are not being shut down | 10:58 |
| juergbi | the oldest one is 7 hours old | 10:58 |
| *** lachlan has quit IRC | 11:15 | |
| gitlab-br-bot | marge-bot123 merged MR !1859 (juerg/ci-schedules->master: .gitlab-ci.yml: Don't run win32 and wsl tests in scheduled pipelines) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/1859 | 11:41 |
| *** lachlan has joined #buildstream | 12:00 | |
| *** lachlan has quit IRC | 12:08 | |
| benschubert | juergbi: if you are happy with !1837 I am too | 12:31 |
| gitlab-br-bot | MR !1837: Fix expiry of public data of artifact https://gitlab.com/BuildStream/buildstream/-/merge_requests/1837 | 12:31 |
| juergbi | benschubert: ok, let's get this merged. thanks for checking | 12:31 |
| benschubert | and thanks a lot to abderrahim[m] for fixing it! | 12:32 |
| juergbi | +1 | 12:34 |
| abderrahim[m] | :) | 12:34 |
| juergbi | after the merge of this and a few other pending fixes I think it would be good to tag 1.93.2 and then we can update the tag in bst-plugins-experiemntal | 12:38 |
| benschubert | Seems good | 12:40 |
| *** lachlan has joined #buildstream | 13:25 | |
| *** mohan43u has quit IRC | 13:37 | |
| *** slaf has quit IRC | 13:38 | |
| *** slaf has joined #buildstream | 13:39 | |
| *** mohan43u has joined #buildstream | 13:41 | |
| gitlab-br-bot | marge-bot123 closed issue #1264 (Staging fails with "Could not find file at ...") on buildstream https://gitlab.com/BuildStream/buildstream/-/issues/1264 | 14:18 |
| gitlab-br-bot | marge-bot123 merged MR !1837 (abderrahim/public-data->master: Fix expiry of public data of artifact) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/1837 | 14:18 |
| *** mohan43u has quit IRC | 14:34 | |
| *** mohan43u has joined #buildstream | 14:37 | |
| *** lachlan has quit IRC | 14:59 | |
| *** lachlan has joined #buildstream | 15:13 | |
| *** lachlan has quit IRC | 15:56 | |
| *** lachlan has joined #buildstream | 16:06 | |
| *** lachlan has quit IRC | 16:14 | |
| gitlab-br-bot | juergbi opened MR !1861 (juerg/job-sigterm->master: Fix SIGTERM handling in job processes) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/1861 | 16:25 |
| gitlab-br-bot | BenjaminSchubert approved MR !1861 (juerg/job-sigterm->master: Fix SIGTERM handling in job processes) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/1861 | 16:29 |
| gitlab-br-bot | marge-bot123 closed issue #1277 (CI for master is failing as it seems there is no WSL runners available) on buildstream https://gitlab.com/BuildStream/buildstream/-/issues/1277 | 16:34 |
| gitlab-br-bot | marge-bot123 merged MR !1860 (jjardon/wsl->master: Revert ".gitlab-ci.yml: Allow WSL test to fail") on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/1860 | 16:34 |
| gitlab-br-bot | marge-bot123 merged MR !1857 (abderrahim/options-1->bst-1: [bst-1] _project.py: resolve options before running the final assertions) on buildstream https://gitlab.com/BuildStream/buildstream/-/merge_requests/1857 | 16:35 |
| *** lachlan has joined #buildstream | 16:40 | |
| *** lachlan has quit IRC | 16:48 | |
| *** lachlan has joined #buildstream | 16:50 | |
| *** tpollard has quit IRC | 17:15 | |
| *** lachlan has quit IRC | 17:26 | |
| *** milloni has quit IRC | 17:27 | |
| *** milloni has joined #buildstream | 17:31 | |
| valentind | abderrahim[m], juergbi, With all the fixes, there are still some issues building freedesktop-sdk with bst2. | 17:37 |
| valentind | https://gitlab.com/freedesktop-sdk/freedesktop-sdk/-/jobs/504045104 | 17:37 |
| valentind | FileNotFoundError | 17:37 |
| juergbi | valentind: will take a look tomorrow. I've also seen a failing test case in CI with a rebased blob-not-found branch. maybe it's related | 17:40 |
| juergbi | it might also be because of the log file | 17:41 |
| juergbi | will try to fix that tomorrow as well | 17:41 |
| valentind | Interestingly, this was during a push. | 17:41 |
| juergbi | yes, so I think we fail push if the log file is not available | 17:42 |
| juergbi | we either have to consider the log file essential or make it optional also in push | 17:42 |
| juergbi | (and pull) | 17:42 |
| juergbi | although, this is part of send_directory for artifact_proto.files | 17:42 |
| juergbi | so this can't be the log file issue | 17:42 |
| valentind | Is not push a different log file? | 17:43 |
| valentind | pull is. | 17:43 |
| juergbi | I mean pushing the build log that is linked in the artifact blob | 17:43 |
| juergbi | *artifact proto | 17:43 |
| juergbi | valentind: is this already with buildbox-casd 0.0.7? | 17:43 |
| valentind | juergbi, Yes | 17:44 |
| juergbi | okhm, and this is pushing right after a fresh build | 17:44 |
| valentind | Yes | 17:44 |
| juergbi | so it would be really odd if expiry was the issue here | 17:45 |
| valentind | But it does not seem to happen for all builds. | 17:45 |
| juergbi | but don't know why else this would fail | 17:45 |
| valentind | Just one failed. | 17:45 |
| juergbi | will review the code, maybe something pops up | 17:45 |
| juergbi | (small) reproducible test case would be great, of course, but might be tricky | 17:45 |
| valentind | Also interesting thing, it does not manage to download files from our mirror. | 17:47 |
| valentind | From https://gitlab.com/freedesktop-sdk/mirrors/tarball-lfs-mirror/ | 17:47 |
| valentind | And this happens on several different machines that have completely different network. | 17:47 |
| valentind | So I think bst2 is the issue. Not sure why though. | 17:47 |
| valentind | For example: https://gitlab.com/freedesktop-sdk/mirrors/tarball-lfs-mirror/raw/master/sourceforge_download/giflib/giflib-5.2.1.tar.gz | 17:48 |
| valentind | It cannot download it. | 17:49 |
| *** santi has quit IRC | 17:49 | |
| *** santi has joined #buildstream | 17:50 | |
| juergbi | might be related to gitlab rejecting requests with empty user agent, which is the default of urllib, iirc | 17:52 |
| juergbi | valentind: I thought there was already fixed but might misremember | 17:52 |
| juergbi | don't know why bst1 is not affected, though, maybe using older version of urllib? | 17:53 |
| *** lachlan has joined #buildstream | 18:01 | |
| valentind | juergbi, It is the same version of python in bst1 and bst2 docker images. | 18:08 |
| juergbi | ah, and urllib is part of the standard library. odd | 18:09 |
| valentind | I do not seee anything about user agent. | 18:10 |
| valentind | In the git log | 18:10 |
| valentind | juergbi, Seems someon else complained: https://gitlab.com/BuildStream/buildstream/-/issues/1281 | 18:14 |
| valentind | Strange, it talks about bst 1.4 | 18:14 |
| juergbi | right, that's what I was referring to | 18:14 |
| valentind | We do not have that issue. | 18:14 |
| juergbi | are you sure it's actually fetching with bst 1.4 and not simply using the cached file right now? | 18:15 |
| *** lachlan has quit IRC | 18:15 | |
| valentind | juergbi, Yes. For example: https://storage.googleapis.com/gitlab-gprd-artifacts/5c/63/5c63681383a059a6dab32deb7a44d5cf535132c0c0c35105f4ce1334be7e8bef/2020_03_15/472007636/507572579/job.log?response-content-type=text%2Fplain%3B%20charset%3Dutf-8&response-content-disposition=inline&GoogleAccessId=gitlab-object-storage-prd@gitlab-production.iam.gserviceaccount.com&Signature=GC1nkGGkx%2B080BZpY8q%2Bx4vRNN9kxmOZRXb9BrZckav80sUosvJhGZx3zlFb%0AvVr%2FWgJ | 18:19 |
| valentind | rCbh7a%2BxFMZ9ykDpdOIeNlq%2BIGCvMhM5NAp93Cg4c8UtPAlgqpjSZ%0AyX5f%2B%2FaErCCXSIBOUFmy%2BguZgi5GgD%2FUnTRk10S0lbeAVw%2F3aZMVp9BO7DzR%0ABmdC%2FnnVALcUp83SxwmGJOgdwH4DIygMQHxWeTd6qvxSesIQObfvVziVziOF%0AIYIGCzyiIoiLyRen9%2BHhh9NLbF5stD2FQPLVlavsAPDBxA0kZ9GNm7DZHL%2FC%0AaCLKnYxfvB8umzykYde2UqWEcgdJaNLg3agncuWAXw%3D%3D&Expires=1586370550 | 18:19 |
| valentind | Ah what a terrible url. | 18:19 |
| valentind | https://gitlab.com/freedesktop-sdk/freedesktop-sdk/-/jobs/472007636 | 18:19 |
| valentind | Then if you download the complete raw, you can search for "tarball-lfs-mirror" | 18:20 |
| valentind | And there are multiple successes. | 18:20 |
| juergbi | ok, no idea what's the difference | 18:21 |
| juergbi | setting a custom User-Agent does seem sensible in general but the difference is odd | 18:22 |
| gitlab-br-bot | valentindavid opened issue #1285 (Some downloads from gitlab.com in master not working but work in 1.4) on buildstream https://gitlab.com/BuildStream/buildstream/-/issues/1285 | 18:22 |
| valentind | juergbi, wget -U "" works. | 18:23 |
| valentind | So the user agent is not useful. | 18:24 |
| valentind | Documentation says wget disable the user agent if it is an empty string. | 18:24 |
| juergbi | I misremembered, according to the gitlab issue, the default user agent is "Python-urllib/3.7" | 18:24 |
| valentind | juergbi, Right, it gives 403 | 18:25 |
| valentind | Ah... | 18:25 |
| valentind | Maybe bst-external has a fix. | 18:25 |
| valentind | No, it is the "tar" plugin. | 18:25 |
| valentind | juergbi, Is that something new that gitlab fails on? | 18:26 |
| valentind | Because maybe this log was too old. | 18:27 |
| juergbi | I don't think I have heard about the issue before last week, however, no idea how long it has existed | 18:27 |
| *** santi has quit IRC | 18:33 | |
| *** phildawson has quit IRC | 19:07 | |
| *** toscalix has joined #buildstream | 19:07 | |
| *** benschubert has quit IRC | 19:13 | |
| *** narispo has joined #buildstream | 19:47 | |
| *** narispo has quit IRC | 20:11 | |
| *** narispo has joined #buildstream | 20:11 | |
| *** narispo has quit IRC | 20:21 | |
| *** narispo has joined #buildstream | 20:21 | |
| *** rdale has quit IRC | 22:10 | |
| *** toscalix has quit IRC | 22:24 | |
| *** narispo has quit IRC | 22:32 | |
| *** narispo has joined #buildstream | 22:32 | |
| *** narispo has quit IRC | 22:47 | |
| *** narispo has joined #buildstream | 22:47 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!