IRC logs for #buildstream for Tuesday, 2018-10-09

« Monday, 2018-10-08 Index Wednesday, 2018-10-10 »
*** catonano has quit IRC00:38
*** tristan has quit IRC02:12
*** Prince781 has joined #buildstream02:21
*** catonano has joined #buildstream05:54
*** mohan43u has quit IRC06:20
*** abderrahim has quit IRC06:22
*** abderrahim has joined #buildstream06:22
*** catonano has quit IRC06:26
*** Prince781 has quit IRC06:28
*** mohan43u has joined #buildstream06:48
*** bochecha has joined #buildstream07:07
*** catonano has joined #buildstream07:16
gitlab-br-botbuildstream: merge request (willsalmon/outOfSourecBuild->master: Out of source builds) #776 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/77608:20
gitlab-br-botbuildstream: merge request (willsalmon/outOfSourecBuild->master: Out of source builds) #776 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/77608:21
*** tristan has joined #buildstream08:26
*** ChanServ sets mode: +o tristan08:26
*** toscalix has joined #buildstream08:40
*** abderrahim has quit IRC08:55
*** jonathanmaw has joined #buildstream08:55
*** abderrahim has joined #buildstream08:57
gitlab-br-botbuildstream: merge request (aevri/contributing_fixups->master: Minor fixups to contributing.rst) #866 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/86609:12
*** lachlan has joined #buildstream09:30
*** rdale has joined #buildstream09:31
gitlab-br-botbuildstream: merge request (aevri/contributing_fixups->master: Minor fixups to contributing.rst) #866 changed state ("merged"): https://gitlab.com/BuildStream/buildstream/merge_requests/86609:36
*** abderrahim has quit IRC09:40
gitlab-br-botbuildstream: issue #700 ("Document how to cleanup the bst cache: locally and in the server side") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/70009:45
*** tristan has quit IRC09:50
*** catonano has quit IRC09:52
gitlab-br-botbuildstream: merge request (lachlan/pickle-yaml-test-list-composite->master: WIP - Add YAM cache testing to yaml tests and further YAML cache refinements) #836 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/83609:57
gitlab-br-botbuildstream: issue #701 ("bst2html doc's") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/70110:00
gitlab-br-botbuildstream: merge request (lachlan/pickle-yaml-test-list-composite->master: WIP - Add YAM cache testing to yaml tests and further YAML cache refinements) #836 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/83610:00
*** alatiera_ has joined #buildstream10:02
*** tristan has joined #buildstream11:00
gitlab-br-botbuildstream: merge request (lachlan/pickle-yaml-test-list-composite->master: WIP - Add YAM cache testing to yaml tests and further YAML cache refinements) #836 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/83611:05
*** dtf has joined #buildstream11:24
*** mohan43u has quit IRC11:29
*** mohan43u has joined #buildstream11:44
*** lachlan has quit IRC11:48
*** lachlan has joined #buildstream11:53
*** dtf has joined #buildstream11:57
*** lachlan has quit IRC12:03
NexusI'm having some issues finding an answer for one of the fixes i'm attempting. It seems like, if the cascache manages to reach a server during a request, but never gets back a response, it hangs forever. Does anyone have an idea of how i can add a timeout to it?12:28
Nexuscode in question is buildstream/_artifactcache/cascache.py +112912:28
Nexus`remote_execution_pb2_grpc.CapabilitiesStub(self.channel).GetCapabilities(request)`12:29
skullmanhttps://grpc.io/blog/deadlines suggests there should be a method for setting a deadline12:31
* skullman would grep the API to see if there's anything with deadline in the name12:32
Nexusnot that i could find12:33
*** ChanServ sets mode: +o tristan12:34
tristanIt might be that it needs to be implemented with the protobufs and that12:34
Nexusprotobufs?12:35
tristanif I understand correctly, remote_execution_pb2_grpc is generated code right ?12:35
tristanAnd the server and local cas needs to support (implement) all of the APIs12:35
tristanhttps://docs.buildstream.build/CONTRIBUTING.html#protocol-buffers12:36
tristanNexus, I'm not familiar yet with this area, but I believe that any aspect of the protocol needs to be implemented by us, at least if we intend to use it12:37
Nexusi know about as much about this area as you do tbh, i'm just seeing what i can find that mightdo wat i think needs to be done12:38
tristanAha, I think https://grpc.io/blog/deadlines is not part of the CAS protocol by looking at that blog12:40
Nexusis juergbi around? I think he wrote this bit12:41
phildawsonNexus, I suspect he'll be busy with  bazelcon12:41
Nexusdarn12:42
phildawsonThough you never know, it's possible he'll turn up at some point.12:42
tristanNexus, google says https://stackoverflow.com/questions/43869397/how-do-you-set-a-timeout-in-pythons-grpc-library12:52
skullmanaw, the hooks in the exception API for the purposes of testing don't retain the task exception object12:53
Nexusyeah i foudn that one, i've been trying to translate it to our usecase but i'm not sure how12:53
tristanskullman, In which case ?12:54
tristanOhh the object12:54
tristanskullman, correct; we don't, because we never serialize complex objects (especially exceptions) through the IPC12:54
skullmanok, was looking to use it to see what exit code the element that failed was, I'll look for a better interface to do that12:55
skullmanit'll probably be in result.stderr, but that's a bit flaky12:56
* skullman will look at the message output formatting12:56
tristanAh... that is also clouded by another issue13:04
tristansec13:05
tristanhttps://gitlab.com/BuildStream/buildstream/issues/28613:05
skullmanthat's the issue I'm working on13:06
tristanI see that you commented there yeah :)13:06
skullmanI can distinguish the error code in my current version, I'm just trying to find a good way to check what it actually said. `assert "exitcode 42" in result.stderr` is what I've currently got13:07
tristanI see, but that will also require a newer version of bwrap correct ?13:07
tristanI guess we need a HAVE_BWRAP_WITH_EXITSTATUS or such in tests/testutils/site.py to conditionally run the test13:08
skullmancould probably make it more robust by changing the log format to "%{element}|%{message}" and check for "sandbox-bwrap/command-exit-42.bst|Command 'exit 42' failed with exitcode 42"13:08
skullmantristan: yeah, that's what I'm doing.13:08
skullmanthough it's called HAVE_BWRAP_JSON_STATUS since it's named after the new bwrap interface rather than the information provided by it13:09
jennisKinnison, so in addition to 'resolving elements', we also have 'resolving cached state'13:14
*** lachlan has joined #buildstream13:14
*** bochecha has quit IRC13:15
*** catonano has joined #buildstream13:23
Kinnisonjennis: yeah, I'm looking through where it might be called13:27
tristanskullman, nod - anyway I think it is not worth extending the testing framework since we probably dont need to assert exit status of sandbox commands in any other case13:35
*** lachlan has quit IRC13:45
*** lachlan has joined #buildstream13:53
*** tristan has quit IRC14:46
skullman:/ I'm currently checking for whether json-status-fd is supported by a version check, for consistency with how the rest of bwrap features are checked for, but that'll have to wait until the bubblewrap code is merged15:03
*** Prince781 has joined #buildstream15:06
*** Prince781 has quit IRC15:29
gitlab-br-botbuildstream: issue #702 ("Make `bst init` take an argument") changed state ("opened") https://gitlab.com/BuildStream/buildstream/issues/70215:48
gitlab-br-botbuildstream: merge request (cascache_timeouts->master: Adding timeout and retries to cascache) #867 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/86715:52
gitlab-br-botbuildstream: merge request (cascache_timeouts->master: WIP: Adding timeout and retries to cascache) #867 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/86715:52
gitlab-br-botbuildstream: merge request (willsalmon/outOfSourecBuild->master: Out of source builds) #776 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/77615:56
gitlab-br-botbuildstream: merge request (cascache_timeouts->master: WIP: Adding timeout and retries to cascache) #867 changed state ("closed"): https://gitlab.com/BuildStream/buildstream/merge_requests/86716:16
*** catonano has quit IRC16:18
gitlab-br-botbuildstream: merge request (richardmaw/distinguish-sandboxing-build-fail->master: WIP: Distinguish between bubblewrap sandboxing failure and command failure) #868 changed state ("opened"): https://gitlab.com/BuildStream/buildstream/merge_requests/86816:25
*** lachlan has quit IRC16:46
*** lachlan has joined #buildstream16:47
*** lachlan has quit IRC16:55
*** lachlan has joined #buildstream16:59
*** jonathanmaw has quit IRC17:01
*** bochecha has joined #buildstream17:02
*** lachlan has quit IRC17:03
*** alatiera_ has quit IRC17:05
*** tristan has joined #buildstream17:10
*** tristan has quit IRC17:11
*** lachlan has joined #buildstream17:12
*** tristan has joined #buildstream17:13
*** xjuan has joined #buildstream17:42
*** xjuan has quit IRC17:43
*** xjuan has joined #buildstream17:53
*** ChanServ sets mode: +o tristan17:59
*** tristan changes topic to "BuildStream 1.2.3 is out ! | https://gitlab.com/BuildStream/buildstream | Docs: https://buildstream.gitlab.io/buildstream | IRC logs: https://irclogs.baserock.org/buildstream | Mailing List: https://mail.gnome.org/mailman/listinfo/buildstream-list | Roadmap: https://wiki.gnome.org/Projects/BuildStream/Road"17:59
*** abderrahim has joined #buildstream18:20
*** toscalix has quit IRC18:22
*** catonano has joined #buildstream18:34
*** tristan has quit IRC18:41
*** tristan has joined #buildstream18:42
*** Prince781 has joined #buildstream19:28
*** Prince781 has quit IRC19:31
*** Prince781 has joined #buildstream19:34
*** catonano has quit IRC20:17
*** Prince781 has quit IRC20:33
*** Prince781 has joined #buildstream20:39
*** catonano has joined #buildstream20:55
*** Prince781 has quit IRC21:55
*** bochecha has quit IRC22:09
*** xjuan has quit IRC22:16
*** xjuan has joined #buildstream22:38
*** catonano has quit IRC22:48
*** tristan has quit IRC23:00
*** Prince781 has joined #buildstream23:02
*** rdale has quit IRC23:21
*** Prince781 has quit IRC23:34
« Monday, 2018-10-08 Index Wednesday, 2018-10-10 »

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