| *** rdale has quit IRC | 00:35 | |
| *** slaf has quit IRC | 01:38 | |
| *** slaf has joined #buildstream | 01:41 | |
| *** slaf has joined #buildstream | 01:41 | |
| *** nimish2711 has joined #buildstream | 02:53 | |
| *** nimish2711 has quit IRC | 03:43 | |
| *** nimish2711 has joined #buildstream | 03:43 | |
| *** nimish2711 has quit IRC | 03:58 | |
| *** nimish2711 has joined #buildstream | 03:58 | |
| *** nimish2711 has quit IRC | 04:03 | |
| *** nimish2711 has joined #buildstream | 04:04 | |
| *** nimish2711 has quit IRC | 04:24 | |
| *** nimish2711 has joined #buildstream | 04:30 | |
| *** nimish2711 has quit IRC | 04:45 | |
| *** nimish2711 has joined #buildstream | 04:46 | |
| *** nimish2711 has quit IRC | 05:04 | |
| *** nimish2711 has joined #buildstream | 05:09 | |
| *** mohan43u has quit IRC | 05:18 | |
| *** mohan43u has joined #buildstream | 05:18 | |
| *** phildawson_ has joined #buildstream | 07:18 | |
| *** phildawson has quit IRC | 07:20 | |
| *** benbrown has quit IRC | 07:20 | |
| *** benbrown_ has joined #buildstream | 07:20 | |
| jennis | Lots of test suite problems over the last few days | 07:26 |
|---|---|---|
| jennis | I've seen tests/sources/zip.py::test_use_netrc[HTTP] hang for a few pipelines now | 07:26 |
| jennis | And lots of "Preparation failed", e.g. https://gitlab.com/BuildStream/buildstream/-/jobs/202372222 | 07:27 |
| jennis | Does anyone have any idea what's going on? | 07:27 |
| jennis | s/test suite problems/CI problems | 07:29 |
| *** toscalix has joined #buildstream | 07:29 | |
| *** bochecha has joined #buildstream | 07:59 | |
| *** nimish2711 has quit IRC | 08:16 | |
| *** nimish2711 has joined #buildstream | 08:17 | |
| *** nimish2711 has quit IRC | 08:30 | |
| *** nimish2711_ has joined #buildstream | 08:31 | |
| *** nimish2711_ is now known as nimish2711 | 08:31 | |
| *** rdale has joined #buildstream | 08:32 | |
| benschubert | all jobs stuck because we don't have runners? wait what?https://gitlab.com/BuildStream/buildstream/-/jobs/202922924 | 08:32 |
| *** nimish2711 has quit IRC | 08:42 | |
| tpollard | :S | 08:51 |
| benschubert | who's responsible for runners? | 08:53 |
| tpollard | jjardon or valentind maybe? | 08:54 |
| valentind | Let me have a look. | 08:54 |
| valentind | DO runner seems down | 08:56 |
| valentind | gitlab-runner crashed. | 09:02 |
| valentind | I will cleanup all the runner images and then restart it. | 09:02 |
| benschubert | thanks valentind ! | 09:02 |
| tpollard | cheers valentind | 09:03 |
| valentind | I do not know what happened. But I suppose some network issue at DO disconnected the runner from the builders. | 09:04 |
| valentind | Still not working | 09:06 |
| *** jonathanmaw has joined #buildstream | 09:12 | |
| valentind | docker-machine has too many opened files | 09:12 |
| valentind | benschubert, tpollard, it should be working now. | 09:15 |
| valentind | Something wrong with docker-machine. There was a lot of processes for it that were hanging. I rebooted the runner. | 09:15 |
| benschubert | valentind: thanks a lot for the quick fix! | 09:15 |
| tpollard | maybe it will remedy the other intermittent failures too | 09:17 |
| benschubert | should we have a periodic reboot of the runners? like once a week might actually "fix" some issues | 09:20 |
| valentind | benschubert, just ping me when you get another issue so I look at it. | 09:21 |
| benschubert | sure, hope I won't need too ;) | 09:21 |
| valentind | I might decide to make periodic reboot like I do with the arm builders. But I will do it only if I know the reason. | 09:21 |
| benschubert | agreed | 09:21 |
| gitlab-br-bot | jennis approved MR !1307 (phil/move-integration-cache-to-testing->master: Make Integration cache fixture available in testing package) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1307 | 09:33 |
| gitlab-br-bot | BenjaminSchubert approved MR !1307 (phil/move-integration-cache-to-testing->master: Make Integration cache fixture available in testing package) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1307 | 09:34 |
| *** bochecha has quit IRC | 09:52 | |
| *** phildawson_ has quit IRC | 09:52 | |
| *** hergertme has quit IRC | 09:52 | |
| *** bochecha_ has joined #buildstream | 09:52 | |
| *** bochecha_ is now known as bochecha | 09:52 | |
| tlater[m] | o/ I'd like to set up a cache for the buildstream docs | 09:52 |
| tlater[m] | Do we have any sort of infrastructure that would be able to run that? | 09:52 |
| *** phildawson_ has joined #buildstream | 09:52 | |
| *** hergertme has joined #buildstream | 09:53 | |
| tlater[m] | See https://gitlab.com/BuildStream/buildstream/merge_requests/1300 for context :) | 09:53 |
| benschubert | valentind: I thing I was too optimistic https://gitlab.com/BuildStream/buildstream/-/jobs/202922924 :/ | 09:55 |
| gitlab-br-bot | tpollard opened MR !1314 (tpollard/artifactmetadata->master: Simplify metadata within Artifact class) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1314 | 10:06 |
| gitlab-br-bot | jonathanmaw approved MR !1314 (tpollard/artifactmetadata->master: Simplify metadata within Artifact class) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1314 | 10:27 |
| tpollard | cheers jonathanmaw | 10:31 |
| valentind | benschubert, still have this too many open files issue. | 10:31 |
| tpollard | juergbi: would you like to look at that MR before I marge it or not? | 10:32 |
| valentind | benschubert, I have found where docker-machine keeps state. I have cleaned up. Now I am removing runners by hand, and I will then restart gitlab-runner | 10:37 |
| valentind | benschubert, It is working now | 10:54 |
| valentind | benschubert, you might need to restart jobs that are hanging. But wait a bit because builders are still being spawned so it might look that hanging, but it just takes time to start. | 10:54 |
| valentind | benschubert, you should look at the duration before restarting a job. If has been less than 15 minutes, let it run a bit. If it is longer restart it. | 10:55 |
| benschubert | valentind: thanks! | 10:56 |
| *** tristan has joined #buildstream | 10:57 | |
| juergbi | tpollard: looks good to me at a quick glance | 11:11 |
| tpollard | juergbi: great, will set it to marge! | 11:13 |
| tlater[m] | valentind: Are you the person to ask about buildstream runners? | 11:16 |
| valentind | tlater[m], yes | 11:16 |
| *** toscalix has quit IRC | 11:17 | |
| tlater[m] | Any chance we could fit another cache for documentation builds in there? | 11:18 |
| tlater[m] | We'd probably need a cache whose artifacts are always up-to-date with master, and one for the current stable release | 11:19 |
| tlater[m] | With the former being updated by CI | 11:20 |
| tlater[m] | Although, if it still simply maps cache keys, perhaps that's not necessary since we could simply retain old artifacts. | 11:22 |
| gitlab-br-bot | marge-bot123 merged MR !1314 (tpollard/artifactmetadata->master: Simplify metadata within Artifact class) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1314 | 11:22 |
| gitlab-br-bot | BenjaminSchubert approved MR !1313 (fixed-bug-for-pip-test->master: tests/testutil/python_repo.py: fixed executable path when running pip) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1313 | 11:24 |
| valentind | tlater[m], I am not surw what documentation builds you are talking about. | 11:28 |
| tlater[m] | valentind: Essentially, an MR I'm currently writing is going to make the buildstream tutorial builds take a lot of time | 11:29 |
| tlater[m] | We don't want that, so we want to use a buildstream artifact cache for the buildstream docs tutorials | 11:29 |
| valentind | I could add a cache server on DO if you want. | 11:30 |
| tlater[m] | That would be great :) | 11:30 |
| valentind | Are the artifacts big? | 11:30 |
| tlater[m] | Let me check... | 11:30 |
| valentind | tlater[m], maybe create an issue and assigned to me. I am not sure I will do that today. Maybe Monday. | 11:31 |
| tlater[m] | valentind: Sure - thanks in any case | 11:32 |
| gitlab-br-bot | tlater opened issue #1011 (Create artifact cache for buildstream tutorial builds) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1011 | 11:41 |
| *** lachlan has joined #buildstream | 11:46 | |
| *** jjardon has quit IRC | 12:01 | |
| *** gokcennurlu has quit IRC | 12:01 | |
| *** albfan has quit IRC | 12:01 | |
| *** albfan has joined #buildstream | 12:03 | |
| *** gokcennurlu has joined #buildstream | 12:04 | |
| *** jjardon has joined #buildstream | 12:04 | |
| *** ChanServ sets mode: +o jjardon | 12:04 | |
| *** lachlan has quit IRC | 12:22 | |
| *** lachlan has joined #buildstream | 12:38 | |
| *** lachlan has quit IRC | 12:43 | |
| *** lachlan has joined #buildstream | 13:08 | |
| *** lachlan has quit IRC | 13:18 | |
| *** lachlan has joined #buildstream | 13:28 | |
| *** lachlan has quit IRC | 13:32 | |
| tpollard | do we have criteria for adding a new distro target to the CI? | 13:33 |
| benschubert | tpollard: I think we usually add if we find some problems with BuildStream on one distro | 13:35 |
| * tlater[m] thinks the criteria were "someone wants to devote time to ensuring that BuildStream works on the distro" | 13:44 | |
| tlater[m] | Usually that comes hand-in-hand with finding a problem on that distro | 13:45 |
| tpollard | I just wondered with us having 2 fedora releases | 13:45 |
| tpollard | I know quite a few people are jumping ship from ubuntu 18.04 to 19.04 (5.0 kernel, python3.7, new gnome etc) | 13:45 |
| tlater[m] | Ah, that's probably more of an artefact of the process by which the distros were added. | 13:46 |
| tpollard | but I think it's fine to just target latest LTS | 13:46 |
| benschubert | tpollard: fedora doesn't do LTS, and has a 13 months support for every version. with a new version every 6 months, that means mainly 2 version to support :) | 13:46 |
| tlater[m] | We originally started testing fedora, and fedora only, because Sam Thursfield liked fedora. | 13:46 |
| tlater[m] | Hence it's been around the longest, and old versions simply haven't cycled out. | 13:46 |
| tlater[m] | Or we need to support both versions, I suppose (: | 13:47 |
| benschubert | tlater[m]: we cycled F27 or 28 if I remember correctly | 13:47 |
| * tlater[m] didn't quite remember if that actually happened, but it was certainly discussed | 13:48 | |
| benschubert | plus buildstream was apckaged for fedora, would be good not to break it but that's an aside | 13:48 |
| tlater[m] | I think we have packages for all tested distros. There's a pkgbuild for arch and a debian package. | 13:49 |
| tlater[m] | Well, "packages", I guess | 13:49 |
| tpollard | cool :) | 13:51 |
| *** lachlan has joined #buildstream | 13:58 | |
| *** lachlan has quit IRC | 14:10 | |
| gitlab-br-bot | marge-bot123 merged MR !1307 (phil/move-integration-cache-to-testing->master: Make Integration cache fixture available in testing package) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/1307 | 14:19 |
| *** lachlan has joined #buildstream | 14:33 | |
| *** tristan has quit IRC | 14:42 | |
| * tlater[m] wonders if tox works with nested virtualenvs... | 14:53 | |
| tlater[m] | It's a shame every piece of python tooling has its own virtualenv management | 14:53 |
| *** tristan has joined #buildstream | 15:20 | |
| jennis | cs-shadow, did you notice that the contrib script for showing dependency relationships was duplicating nodes? | 15:54 |
| jennis | Here's the output for a project of 5 'fanned' levels: https://ibb.co/yg58Byy | 15:56 |
| *** tpollard has quit IRC | 15:58 | |
| cs-shadow | jennis: didn't notice that. Would you be able to share the source for the project as well so that i can try it out? | 16:04 |
| jennis | yeah, it's just the bst benchmarks repo | 16:13 |
| jennis | one sec | 16:13 |
| jennis | cs-shadow: if you're able, install this tool: https://gitlab.com/BuildStream/benchmarks/tree/master/contrib/bstgen | 16:14 |
| jennis | and then I just did `bstgen fan -l 3 project` and that'll create a 'project' directory and write the project there | 16:15 |
| cs-shadow | thanks! Will try to see if i get the same thing. I remember deduplicating the nodes at some point but it's been a while since I looked at it last | 16:15 |
| jennis | Ahh fair enough, it's a really cool little script though! | 16:16 |
| jennis | ahh, i used `-l 5` to get that graph, not 3 | 16:17 |
| *** lachlan has quit IRC | 16:21 | |
| *** lachlan has joined #buildstream | 16:26 | |
| lachlan | For info, tpollard s changes under tpollard/artifactmetadata have made quite a differrence to one of the benchmarking tests https://buildstream.gitlab.io/buildstream-benchmarking/test-debian-like-project/public/average_kb_master_Show_of_Debian_like_project_for_x86_64_20190426-010809.html | 16:48 |
| *** lachlan has quit IRC | 16:57 | |
| *** lachlan has joined #buildstream | 16:58 | |
| *** lachlan has quit IRC | 17:04 | |
| *** lachlan has joined #buildstream | 17:07 | |
| *** jjardon has quit IRC | 17:24 | |
| *** jjardon has joined #buildstream | 17:25 | |
| *** ChanServ sets mode: +o jjardon | 17:25 | |
| *** lachlan has quit IRC | 17:32 | |
| *** jonathanmaw has quit IRC | 17:34 | |
| *** lachlan has joined #buildstream | 18:04 | |
| *** lachlan has quit IRC | 18:24 | |
| *** nimish2711 has joined #buildstream | 18:45 | |
| *** nimish2711 has quit IRC | 18:49 | |
| *** nimish2711 has joined #buildstream | 18:50 | |
| *** nimish2711 has quit IRC | 18:55 | |
| *** nimish2711 has joined #buildstream | 18:55 | |
| *** lachlan has joined #buildstream | 19:04 | |
| *** nimish2711 has quit IRC | 19:05 | |
| *** nimish2711 has joined #buildstream | 19:11 | |
| *** nimish2711 has quit IRC | 19:21 | |
| *** nimish2711 has joined #buildstream | 19:22 | |
| *** lachlan has quit IRC | 19:25 | |
| *** rdale has quit IRC | 19:40 | |
| *** nimish2711 has quit IRC | 19:44 | |
| *** nimish2711 has joined #buildstream | 19:46 | |
| gitlab-br-bot | bochecha opened issue #1012 (Traceback when tracking an element) on buildstream https://gitlab.com/BuildStream/buildstream/issues/1012 | 19:57 |
| *** nimish2711 has quit IRC | 20:21 | |
| *** cs-shadow has quit IRC | 21:12 | |
| *** bochecha has quit IRC | 23:45 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!