IRC logs for #buildstream for Tuesday, 2019-08-06

*** bochecha has quit IRC00:08
*** bochecha has joined #buildstream00:08
*** bochecha has quit IRC00:18
*** bochecha has joined #buildstream00:18
*** tristan has quit IRC00:22
*** bochecha has quit IRC00:23
*** bochecha has joined #buildstream00:24
*** bochecha has quit IRC00:29
*** bochecha has joined #buildstream00:29
*** bochecha has quit IRC00:39
*** bochecha has joined #buildstream00:39
*** bochecha has quit IRC00:44
*** bochecha has joined #buildstream00:45
*** bochecha has quit IRC00:50
*** bochecha has joined #buildstream00:50
*** bochecha has quit IRC01:00
*** bochecha has joined #buildstream01:00
*** bochecha has quit IRC01:06
*** bochecha has joined #buildstream01:06
*** bochecha has quit IRC01:16
*** bochecha has joined #buildstream01:16
*** bochecha has joined #buildstream01:17
*** bochecha has quit IRC01:32
*** bochecha has joined #buildstream01:32
*** bochecha has quit IRC01:37
*** bochecha has joined #buildstream01:37
*** bochecha has joined #buildstream01:38
*** bochecha has quit IRC01:43
*** bochecha has joined #buildstream01:43
*** bochecha has quit IRC01:53
*** bochecha has joined #buildstream01:53
*** bochecha has quit IRC01:59
*** bochecha has joined #buildstream01:59
*** tristan has joined #buildstream02:04
*** bochecha has quit IRC02:04
*** bochecha has joined #buildstream02:04
*** tristan has quit IRC02:07
*** tristan has joined #buildstream02:09
*** bochecha has quit IRC02:09
*** bochecha has joined #buildstream02:10
*** tristan has quit IRC02:10
*** bochecha has quit IRC02:11
*** bochecha has joined #buildstream02:12
*** bochecha has quit IRC02:17
*** bochecha has joined #buildstream02:17
*** bochecha has quit IRC02:22
*** bochecha has joined #buildstream02:23
*** bochecha has quit IRC02:33
*** bochecha has joined #buildstream02:33
*** bochecha has quit IRC02:38
*** bochecha has joined #buildstream02:39
*** tristan has joined #buildstream02:39
*** tristan has quit IRC02:42
*** bochecha has quit IRC02:44
*** bochecha has joined #buildstream02:44
*** bochecha has quit IRC02:49
*** bochecha has joined #buildstream02:49
*** bochecha has quit IRC02:51
*** bochecha has joined #buildstream02:51
*** tristan has joined #buildstream02:52
*** bochecha has quit IRC02:56
*** bochecha has joined #buildstream02:57
*** bochecha has quit IRC03:02
*** bochecha has joined #buildstream03:02
*** bochecha has quit IRC03:02
*** bochecha has joined #buildstream03:02
*** tristan has quit IRC03:07
*** bochecha has quit IRC03:07
*** bochecha has joined #buildstream03:08
*** bochecha has quit IRC03:13
*** bochecha has joined #buildstream03:13
*** bochecha has joined #buildstream03:13
*** bochecha has quit IRC03:18
*** bochecha has joined #buildstream03:18
*** bochecha has quit IRC03:24
*** bochecha has joined #buildstream03:24
*** bochecha has quit IRC03:29
*** bochecha has joined #buildstream03:29
*** tristan has joined #buildstream03:33
*** tristan has quit IRC03:34
*** bochecha has quit IRC03:49
*** bochecha has joined #buildstream03:50
*** bochecha has quit IRC03:55
*** bochecha has joined #buildstream03:55
*** bochecha has quit IRC04:00
*** bochecha has joined #buildstream04:00
*** bochecha has quit IRC04:05
*** bochecha has joined #buildstream04:06
*** tristan_ has quit IRC04:09
*** bochecha has quit IRC04:11
*** bochecha has joined #buildstream04:11
*** bochecha has quit IRC04:16
*** bochecha has joined #buildstream04:16
*** narispo has quit IRC04:17
*** bochecha has quit IRC04:22
*** bochecha has joined #buildstream04:22
*** bochecha has quit IRC04:45
*** bochecha has joined #buildstream04:45
*** bochecha has quit IRC04:50
*** bochecha has joined #buildstream04:51
*** tristan has joined #buildstream05:06
*** bochecha has quit IRC05:16
*** bochecha has joined #buildstream05:16
*** bochecha has quit IRC05:16
*** bochecha has joined #buildstream05:17
*** bochecha has quit IRC05:22
*** bochecha has joined #buildstream05:22
*** bochecha has quit IRC05:23
*** bochecha has joined #buildstream05:23
*** bochecha has quit IRC05:28
*** bochecha has joined #buildstream05:28
*** bochecha has quit IRC05:33
*** bochecha has joined #buildstream05:34
*** tristan_ has joined #buildstream05:36
*** bochecha has quit IRC05:39
*** bochecha has joined #buildstream05:39
*** bochecha has quit IRC05:49
*** bochecha has joined #buildstream05:49
*** bochecha has quit IRC05:55
*** bochecha has joined #buildstream05:55
*** bochecha has joined #buildstream06:06
*** tristan_ has quit IRC06:07
*** bochecha has quit IRC06:16
*** bochecha has joined #buildstream06:16
*** tristan_ has joined #buildstream06:16
*** bochecha has quit IRC06:21
*** bochecha has joined #buildstream06:21
*** tristan_ has quit IRC06:21
*** bochecha has quit IRC06:21
*** bochecha has joined #buildstream06:22
*** bochecha has quit IRC06:27
*** bochecha has joined #buildstream06:27
*** narispo has joined #buildstream06:40
*** bochecha has quit IRC06:47
*** bochecha has joined #buildstream06:47
*** bochecha has quit IRC06:52
*** bochecha has joined #buildstream06:53
*** tristan has quit IRC06:58
*** tristan has joined #buildstream07:00
*** bochecha has quit IRC07:03
*** bochecha has joined #buildstream07:03
*** bochecha has quit IRC07:08
*** bochecha has joined #buildstream07:08
*** bochecha has quit IRC07:08
*** bochecha has joined #buildstream07:09
gitlab-br-botjjardon closed issue #1097 (bst 1.x overnigth tests are failing: SyntaxError: invalid syntax) on buildstream https://gitlab.com/BuildStream/buildstream/issues/109707:11
*** bochecha has quit IRC07:14
*** bochecha has joined #buildstream07:14
*** bochecha has quit IRC07:19
*** bochecha has joined #buildstream07:19
*** bochecha has quit IRC07:24
*** bochecha has joined #buildstream07:25
*** bochecha has quit IRC07:40
*** bochecha has joined #buildstream07:40
*** bochecha has joined #buildstream07:40
*** kapip has joined #buildstream07:49
*** bochecha has quit IRC07:50
*** bochecha has joined #buildstream07:51
*** bochecha has quit IRC07:56
*** bochecha has joined #buildstream07:56
*** bochecha has quit IRC08:01
*** bochecha has joined #buildstream08:01
*** bochecha has quit IRC08:11
*** bochecha has joined #buildstream08:12
*** bochecha has joined #buildstream08:13
qinustyAnyone know if the details in https://buildgrid.gitlab.io/buildgrid/using_buildstream.html#project-configuration are still accurate for configuring buildgrid with buildstream (master) ?08:14
*** bochecha has quit IRC08:18
*** bochecha has joined #buildstream08:18
*** bochecha has quit IRC08:23
*** bochecha has joined #buildstream08:23
*** bochecha has quit IRC08:28
*** bochecha has joined #buildstream08:29
*** bochecha has quit IRC08:39
*** bochecha has joined #buildstream08:39
*** bochecha has quit IRC08:39
*** bochecha has joined #buildstream08:40
*** bochecha has quit IRC08:45
*** bochecha has joined #buildstream08:45
*** bochecha has quit IRC08:50
*** bochecha has joined #buildstream08:50
jjardontristan: probably you already noticed, but only in case: https://gitlab.com/BuildStream/buildstream/issues/1098 (bst-1.2 branch is fine)08:51
*** bochecha has quit IRC08:56
*** bochecha has joined #buildstream08:56
*** jonathanmaw has joined #buildstream09:15
jjardonwho has the access to https://pypi.org/project/BuildStream/ ? the release there is 1.2.7 but latest is 1.2.809:17
jjardonlaurence: ^ ?09:17
jennisjjardon, I *think* that's cs-shadow and tristan ^09:17
jjardonour buildstream-docker-images depend on that to get the latest version of buildstream, so currently we are shipping 1.2.7 on them instead 1.2.809:19
cs-shadowjjardon: Trisatn and I have access to it at the moment. Unless anyone beats me to it, I'll update it this evening09:19
cs-shadow(my pypi creds or on another laptop)09:19
jjardoncs-shadow: thanks a lot!09:19
jjardoncs-shadow: did you notice the nigthly images failing in the buildstream-docker-images repo? https://gitlab.com/BuildStream/buildstream-docker-images/-/jobs/26662877909:21
*** bochecha has quit IRC09:21
*** bochecha has joined #buildstream09:21
jjardonI wonder if we should allow the nigthly one to fail, as It's possible something is broken in bst master and we are holding all the rest to be released09:21
*** bochecha has quit IRC09:23
*** bochecha has joined #buildstream09:28
*** lachlan has joined #buildstream09:30
cs-shadowjjardon: i think that was a one off. My worry with allowing it to fail is that we will release something that doesn't work. Although this one was one off, we fail much more often due to Docker daemon issues on the runners. I think improving those runners would do much more in terms of imporoving stabilityy09:43
jjardoncs-shadow: I see, I have rerun the same pipeline 3 times and still failing though09:49
jjardononly the nigthlies, sometimes the -slim one, sometimes the normal variant09:49
jjardoncs-shadow: If we allow to fail and make the push depend on them, it will not be released, isnt it?09:50
cs-shadowjjardon: yeah, I've seen this before. I am yet to find the real root cause but it happens when the build step messes up something. When this has happened in the past, I have restarted the build job and then the test job09:51
cs-shadowdoing that now, let's see if that fixes it09:52
jjardonok, thanks!09:52
cs-shadowin future, i'd like to add an easy way of retrying all steps for a given image but that'll require some thought :)09:52
*** lachlan has quit IRC10:03
jjardoncs-shadow: very nice work on that repo btw; pity docker is a bit flaky10:08
*** lachlan has joined #buildstream10:11
gitlab-br-botcoldtom opened MR !1523 (coldtom/cython-req->master: Add cython as a dev requirement) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/152310:21
*** lachlan has quit IRC10:30
gitlab-br-botmarge-bot123 closed issue #1087 (Reduce cache_keys_unstable manipulation) on buildstream https://gitlab.com/BuildStream/buildstream/issues/108710:33
gitlab-br-botmarge-bot123 merged MR !1492 (traveltissues/setstability->master: Add Element.__update_cache_keys_stability method) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/149210:33
qinustyAny ideas what could be causing "Configured remote does not have the BuildStream capabilities service. Please check remote configuration." in relation to remote caches when buildstream (master) is configured to use buildgrid for remote exec?10:55
*** raoul has joined #buildstream11:01
*** toscalix has joined #buildstream11:17
*** lachlan has joined #buildstream11:18
*** lachlan has quit IRC11:26
*** dftxbs3e has quit IRC11:27
*** lachlan has joined #buildstream11:39
*** lachlan has quit IRC11:43
tlater[m]qinusty: We have extensions on top of normal CAS that we need for artifact servers.12:39
tlater[m]So you can't run a CAS server, you need to run bst-artifact-server12:39
* tlater[m] is currently working on fixing that12:39
qinustyOkay, is disabling the artifact cache is fine for remote execution?12:40
tlater[m]MR hopefully up by end of the week :)12:40
qinustys/cache is/cache12:40
tlater[m]Hm, I'm not sure. I think you need to upload artifacts to the artifact cache, and the current intention for the workflow is that you run a bst-artifact-server as well as your remote execution server. But I'm probably not the person to ask.12:41
jjardontlater[m]: what your fix is about? Only a CAS will be needed for remote caching?12:41
tlater[m]I think juergbi would know, but he appears to be absent12:41
tlater[m]Maybe jennis?12:41
tlater[m]jjardon: I think you should still need to run an artifact server to keep track of the artifacts, but at least you can use a normal CAS server with it.12:45
tpollardwe have some tests in the suite that use buildgrid afaik, so I'd look at how they're configured12:45
tpollardor at least a ci stage for it12:45
jjardontlater[m]: ah ok, so you fix will allow buildstream to use a generic CAS (buildfarm, buildbarn ...) instead only supporting BuildGrid; is that right?12:47
juergbiqinusty: artifact server should not be required for RE12:50
tlater[m]jjardon: Yep12:50
jjardonawesome, thanks!12:50
jjardonCan I haver reviews of https://gitlab.com/BuildStream/buildstream-docker-images/merge_requests/122/ , please?12:56
jjardoncs-shadow: ^12:57
cs-shadowjjardon: sure, left one comment on the MR12:59
jjardoncs-shadow: thanks! pushed a new version13:01
cs-shadowjjardon: ace. Thanks!13:02
*** lachlan has joined #buildstream13:35
*** narispo has quit IRC13:38
*** narispo has joined #buildstream13:38
gitlab-br-botmarge-bot123 merged MR !1403 (jennis/junctions_and_remotes->master: Revert default junction behaviour and add new junction config options) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/140313:45
*** lachlan has quit IRC13:45
*** lachlan has joined #buildstream13:46
laurencenb all - as a reminder, the monthly meeting is due to start in 10 mins. that's over on #buildstream-meetings13:50
laurencealso as a reminder, I have a clash and cannot facilitate this one I'm afraid - someone else needs to step in please13:51
*** narispo has quit IRC13:57
*** narispo has joined #buildstream13:57
qinustycheers tlater[m], juergbi. On]ce I've confirmed this is the case, I'll update the buildgrid docs for getting started with buildstream. I'm currently getting a slightly different issue where buildgrid doesn't appear to be returning an output directory when building autotools.14:23
qinustyI've pasted some snippets here https://paste.gnome.org/paxutrw1f/kb5lje/raw14:23
juergbiqinusty: ah, the hosttools worker is not usable with buildstream. buildstream requires sandboxing, or at least a chroot environment14:26
qinustyThat's something to work with :D I'll chase that down. Thanks juergbi14:27
juergbithe buildgrid buildbox worker should be using buildbox-fuse14:27
qinustyWe're using --buildbox-run=/usr/local/bin/buildbox-run-hosttools currently. I'll have to spin something up specifically for buildstream14:28
juergbiafaict, you're using the buildbox-worker, which doesn't support sandboxing/buildstream yet (but hopefully will soon)14:28
juergbithe 'old' buildgrid buildbox worker should work with buildbox-fuse, though14:28
juergbiprobably a bit confusing during the transition phase14:28
juergbithe worker/bot setup described here should still work, though: https://buildgrid.gitlab.io/buildgrid/using_buildstream.html14:29
juergbihave you been following that documentation or something else?14:30
qinustyI'm currently trying to get buildstream integrated into https://gitlab.com/remote-apis-testing/remote-apis-testing. Using that guide as a basis, however attempting to work with the buildgrid infra in the kubernetes configs14:35
qinustyIt's becoming apparently I'll have to tweak a few more things on the buildgrid side14:36
qinustys/apparently/apparent14:36
qinustyIs there an issue for "which doesn't support sandboxing/buildstream yet (but hopefully will soon)" or a WIP MR? juergbi14:43
juergbithis is related: https://gitlab.com/BuildGrid/buildbox/buildbox-worker/issues/514:44
juergbibuildbox-run-userchroot could also be used instead but userchroot is not commonly available14:44
juergbiand FUSE support, technically not a requirement, but provides better performance: https://gitlab.com/BuildGrid/buildbox/buildbox-casd/issues/814:45
*** hatari7 has joined #buildstream14:46
* tlater[m] is trying to figure out why we have this ArtifactRemoteSpec class: https://gitlab.com/BuildStream/buildstream/blob/master/src/buildstream/_artifactcache.py#L3514:48
tlater[m]Is there a special reason why we alias CASRemoteSpec instead of just reusing that class?14:48
tlater[m]-> trying to refactor that part of the code a bit...14:48
*** hatari7 has quit IRC14:48
coldtomhi, i've installed master from git, but running any command tells me i have severely malformed yaml in the default userconfig.yaml14:49
coldtomany idea why this is happening?14:50
jennis:(14:50
qinustyI briefly had this when installing between different versions? I think uninstalling buildstream fixed it14:50
qinustyand reinstalling, clearly14:50
jennispip3 uninstall ruamel.yaml.clib14:50
jennispip3 install ruamel.yaml.clib14:50
jennisFixes it...14:50
qinustywhich likely did that ^14:50
tpollardtlater[m]: seems redundant to me14:51
tlater[m]tpollard: Ta, just wanted a sanity check14:51
jennisI'd imagine it's to do with this: https://gitlab.com/BuildStream/buildstream/merge_requests/1506 in combination with our cython changes14:51
juergbitlater[m]: ArtifactCacheSpec might be separate for historic reasons, it used to be more than just an alias14:51
coldtomta jennis14:52
tlater[m]Yeah, I figured. It just struck me as a particularly odd alias14:52
*** lachlan has quit IRC15:06
*** connorshea[m] has quit IRC15:40
*** asingh_[m] has quit IRC15:40
*** rafaelff[m] has quit IRC15:40
*** segfault3[m] has quit IRC15:40
*** Trevinho[m] has quit IRC15:40
*** benschubert has quit IRC15:40
*** benschubert has joined #buildstream15:40
*** jjardon[m] has quit IRC15:40
*** reuben640[m] has quit IRC15:40
*** kailueke[m] has quit IRC15:40
*** theawless[m] has quit IRC15:40
*** tlater[m] has quit IRC15:40
*** dbuch has quit IRC15:40
*** mattiasb has quit IRC15:40
*** krichter[m] has quit IRC15:40
*** m_22[m] has quit IRC15:41
*** tchaik[m] has quit IRC15:41
*** skullone[m] has quit IRC15:41
*** persia has quit IRC15:41
*** nielsdg has quit IRC15:41
*** persia has joined #buildstream15:41
*** lachlan has joined #buildstream15:41
*** connorshea[m] has joined #buildstream15:56
*** phildawson_ has joined #buildstream15:59
*** phildawson has quit IRC16:01
*** lachlan has quit IRC16:09
*** lachlan has joined #buildstream16:20
*** lachlan has quit IRC16:26
*** asingh_[m] has joined #buildstream16:42
*** ChanServ sets mode: +o tristan16:47
tristanjjardon, I see. ok I'll try to build that branch locally, it's reproduced in CI (I retried the job you linked to because it looks like a spurious error I sometimes was seeing in tests at some point)16:48
jjardonyeah, I though it was a hiccup but seems It's always reproducible16:52
*** lachlan has joined #buildstream16:54
*** rafaelff[m] has joined #buildstream16:54
tristanjjardon, https://gitlab.com/BuildStream/bst-external/merge_requests/97 <-- thats the bst-external pipeline that takes too long16:54
tristanunblocking that would let me add the patch to gnome-build-meta16:55
*** phildawson_ has quit IRC16:55
tristanjjardon, one of the changes though caused cache keys to break, so 1.4 will *not* have the same cache keys16:55
tristan(so I brought in abderrahim[m]'s wishlist item of using json for cache keys since they already got broke)16:55
tristanjjardon, I am guessing that some of the tests will work differently because of that; i.e. nothing will be cached anywhere when tests run16:56
*** phildawson_ has joined #buildstream16:56
tristanpossible that that 1.2 branch just doesnt hit the same problem because it doesnt build from the bootstrap image16:56
*** phildawson_ has quit IRC17:01
*** phildawson_ has joined #buildstream17:02
*** jonathanmaw has quit IRC17:04
*** Trevinho[m] has joined #buildstream17:06
*** phildawson_ has quit IRC17:08
*** phildawson_ has joined #buildstream17:11
jjardontristan: there is a separate job to run without cache, let me check17:12
jjardontristan: yeah, still green: https://gitlab.com/BuildStream/buildstream/pipelines/7482205717:12
*** phildawson_ has quit IRC17:17
*** phildawson_ has joined #buildstream17:17
*** segfault3[m] has joined #buildstream17:21
*** phildawson_ has quit IRC17:28
*** phildawson_ has joined #buildstream17:29
*** phildawson_ has quit IRC17:31
*** phildawson_ has joined #buildstream17:32
*** lachlan has quit IRC17:33
jjardontristan: bst-external should be green now: https://gitlab.com/BuildStream/bst-external17:35
*** toscalix has quit IRC17:40
*** phildawson_ has quit IRC17:40
*** jjardon[m] has joined #buildstream17:42
*** phildawson_ has joined #buildstream17:45
*** phildawson_ has quit IRC17:49
*** reuben640[m] has joined #buildstream17:55
*** phoenix has joined #buildstream17:57
*** phoenix has quit IRC17:57
*** kailueke[m] has joined #buildstream18:00
*** tlater[m] has joined #buildstream18:31
*** dbuch has joined #buildstream18:45
*** mattiasb has joined #buildstream18:45
*** tristan_ has joined #buildstream18:49
*** ChanServ sets mode: +o tristan_18:53
*** dftxbs3e has joined #buildstream18:55
*** krichter[m] has joined #buildstream18:55
*** theawless[m] has joined #buildstream19:16
*** m_22[m] has joined #buildstream19:17
*** tchaik[m] has joined #buildstream19:28
*** skullone[m] has joined #buildstream19:28
*** phoenix has joined #buildstream20:15
*** nielsdg has joined #buildstream20:33
*** tristan_ has quit IRC21:05
*** tristan_ has joined #buildstream21:06
*** phoenix has quit IRC21:25
*** phoenix_ has joined #buildstream21:26
*** phoenix_ is now known as phoenix21:26
*** phoenix has quit IRC21:49
*** ChanServ sets mode: +o tristan_22:09

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!