IRC logs for #buildstream for Monday, 2019-10-14

« Sunday, 2019-10-13 Index Tuesday, 2019-10-15 »
*** tpollard has joined #buildstream07:48
*** bochecha has joined #buildstream08:11
*** benschubert has joined #buildstream08:18
*** tiagogomes has joined #buildstream08:26
*** santix has joined #buildstream08:39
gitlab-br-botjuergbi closed issue #909 (Artifact As A Proto) on buildstream https://gitlab.com/BuildStream/buildstream/issues/90908:43
*** jonathanmaw has joined #buildstream09:03
*** lachlan has joined #buildstream09:36
*** traveltissues has joined #buildstream09:42
gitlab-br-botmarge-bot123 merged MR !1624 (aevri/platform_win32->master: win32: _platform/win32: add support for win32) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/162409:58
tpollard\o/10:08
*** tpollard has quit IRC10:12
*** tpollard has joined #buildstream10:15
*** traveltissues has quit IRC10:16
*** traveltissues has joined #buildstream10:17
*** traveltissues has quit IRC10:30
*** traveltissues has joined #buildstream10:31
*** rdale has quit IRC10:32
*** rdale has joined #buildstream10:33
*** santix has quit IRC10:42
*** lachlan has quit IRC10:43
*** santix has joined #buildstream10:45
*** santix has quit IRC10:46
*** santix has joined #buildstream10:48
*** toscalix has joined #buildstream10:54
*** lachlan has joined #buildstream10:57
*** lachlan has quit IRC11:07
gitlab-br-botaevri opened (was WIP) MR !1638 (aevri/casdprocessmanager->master: cascache: refactor, extract CASDProcessManager) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/163811:11
*** lachlan has joined #buildstream11:16
*** cs-shadow has joined #buildstream11:17
*** traveltissues has quit IRC11:29
*** lachlan has quit IRC12:09
tlater[m]Does anyone know why we remove the artifactshare directory after a test is done?12:14
tlater[m]It's inside the @pytest.mark.datafiles-given directory, so surely it shouldn't be a problem to keep it around?12:15
*** santix has quit IRC12:35
KinnisonDon't want artifact caches to pollute from test to test12:42
tlater[m]Kinnison: My point is that that directory is subsequently removed by pytest12:51
tlater[m]Unless the test fails, in which case no other test runs inside it12:52
KinnisonOh I see12:54
Kinnisonerm12:54
Kinnisonno idea then :D12:54
benschubertIt would start keeping lots of files no?12:55
tlater[m]benschubert: You mean between runs? Not more than we already do12:56
tlater[m]Because everything else is persisted12:56
tlater[m]It's just a random shutil.rmtree() call at the end of our helpers for artifact caches.12:56
benschubertif it doesn't blow up storage then I'm fine :)12:56
* tlater[m] will double check if getting rid of that call breaks anything, then12:57
*** santix has joined #buildstream13:18
*** lachlan has joined #buildstream13:35
*** akvilebirgelyte_ has quit IRC13:47
*** lachlan has quit IRC13:47
*** akvilebirgelyte has joined #buildstream14:02
*** lachlan has joined #buildstream14:18
*** lachlan has quit IRC14:24
benschubertjuergbi: concerning CasCache, 'refs/head' is buildstream specific, not cas specific, correct? Is there a reason why it's in cas/ then?14:38
juergbibenschubert: yes. for artifact protos that should no longer be the case14:39
juergbishould only be used for backward compatibility in bst-artifact-server14:40
benschubertthen is there a reason it's in 'CasCache' and not in 'CasServer' ?14:40
benschubertother than 'we didn't move it' ?14:40
juergbino, we should move it14:41
juergbimaybe tlater[m] is already doing that in his branch14:41
benschuberttlater[m]: let me know if you are, otherwise I'll shoot a PR :)14:42
benschubertthanks juergbi14:42
tlater[m]benschubert: No, I've not gotten to untangling CASCache yet14:43
tlater[m]PR would be appreciated, if you've already got it :)14:43
benschuberttlater[m]: would i conflict with you by moving this to casserver?14:44
tlater[m]Probably not, at least not much14:44
benschubertso adding a new 'CasServerCache' class in casserver.py that inherits cascache and moves the few methods/variables there right?14:45
tlater[m]But well, I don't mind the conflict, it'll be easy to sort out on my end14:45
tlater[m]This patch isn't going to be huge14:46
tlater[m]Hrm, actually, if you're going to inherit from CASCache that might be a bit more of a headache14:46
benschubertok I won't touch it then14:46
benschubertI'll let you handle this14:46
juergbiI would have expected to simply move that logic, without introducing a new class14:47
tlater[m]Ok, I'll make sure to catch the superfluous CASCache methods then14:47
juergbita14:47
benschubertawesome14:47
*** toscalix has quit IRC14:49
*** lachlan has joined #buildstream14:50
*** lachlan has quit IRC14:55
*** lachlan has joined #buildstream15:10
*** lachlan has quit IRC15:15
*** lachlan has joined #buildstream15:17
*** lachlan has quit IRC15:22
*** santix has quit IRC15:22
*** lachlan has joined #buildstream15:23
*** santix has joined #buildstream15:24
*** santix has quit IRC15:26
*** lachlan has quit IRC15:26
*** santix has joined #buildstream15:27
*** santix has quit IRC15:30
tlater[m]juergbi: I have a test that abuses `CASCache.add_object` to put a file inside a cache, pretending that a remote cache is a localcas15:35
tlater[m]It uses `CaptureFiles`, but the file doesn't seem to appear on the other side15:35
tlater[m]Any idea how I could figure out why?15:35
juergbitlater[m]: try running casd with higher log level15:39
juergbibst --debug does that now but that might not help in the test case15:40
tlater[m]Hum, I guess I'd need to hardcode something in the test suite for that?15:41
tlater[m]Oh, right, we have that enum15:41
tlater[m]juergbi: I assume I'd see some kind of messages if a client tried to do anything with LogLevel.TRACE?16:01
tlater[m]It looks like the LocalCas isn't getting any connections, which is strange16:01
juergbitlater[m]: I think we're not fully consistent with request tracing in buildbox-casd, unfortunately, so can't say for sure otoh16:02
tlater[m]Fair enough, I'll need to dig a little then16:03
juergbitlater[m]: do you at least have a casd log file with the startup debug messages?16:03
tlater[m]Yup16:03
juergbiok. and you may also want to make sure it's for the right casd process as in the test suite sometimes multiple casd instances are running (local and remote)16:03
tlater[m]So I know the server is being started, if nothing else16:03
tlater[m]It seems to be, judging by the socket name16:04
tlater[m]I assume it can't reuse that?16:04
juergbisocket files should be in unique temp dirs, so that should be fine16:05
tlater[m]I suppose I'll go figure out what happened before my patches. This may all be a red herring...16:06
*** santix has joined #buildstream16:06
*** santix has quit IRC16:07
*** santix has joined #buildstream16:08
*** lachlan has joined #buildstream16:13
tlater[m]juergbi: Gah, I finally figured it out16:30
tlater[m]I accidentally changed the directory structure and ended up without a parent 'cas' directory :|16:30
tlater[m]Sorry for the noise16:31
juergbiah, that definitely explains it16:32
* tlater[m] has done similar things twice now16:32
tlater[m]It's really easy to miss this16:32
*** jonathanmaw has quit IRC17:03
*** lachlan has quit IRC17:08
*** lachlan has joined #buildstream17:11
benschubertjuergbi: I think I remember a place where we give RW access to CAS to the files in the directory to import, but I can't find where it is in BuildStream. Do you remember by any chance?17:13
*** santix has quit IRC17:15
*** lachlan has quit IRC17:26
*** tiagogomes has quit IRC17:56
*** lachlan has joined #buildstream18:01
*** lachlan has quit IRC18:09
*** lachlan has joined #buildstream18:31
*** lachlan has quit IRC19:40
*** lachlan has joined #buildstream19:42
*** lachlan has quit IRC20:21
*** cs-shadow has quit IRC21:14
*** Trevinho[m] has quit IRC21:14
*** segfault3[m] has quit IRC21:14
*** mohan43u has quit IRC21:14
*** asingh_[m] has quit IRC21:14
*** gimpnet-irc[m] has quit IRC21:14
*** theawless[m] has quit IRC21:14
*** cgmcintyre[m] has quit IRC21:14
*** Demos[m] has quit IRC21:14
*** benschubert has quit IRC21:14
*** tristan has quit IRC21:14
*** tchaik[m] has quit IRC21:14
*** ssssam[m] has quit IRC21:14
*** jswagner has quit IRC21:14
*** reuben640[m] has quit IRC21:14
*** tlater[m] has quit IRC21:14
*** rafaelff[m] has quit IRC21:14
*** mrmcq2u[m] has quit IRC21:14
*** nielsdg has quit IRC21:14
*** dbuch has quit IRC21:14
*** kailueke[m] has quit IRC21:14
*** abderrahim[m] has quit IRC21:14
*** m_22[m] has quit IRC21:14
*** doras has quit IRC21:14
*** mattiasb has quit IRC21:14
*** albfan[m] has quit IRC21:14
*** jjardon[m] has quit IRC21:14
*** skullone[m] has quit IRC21:14
*** verdre[m] has quit IRC21:14
*** waltervargas[m] has quit IRC21:14
*** awacheux[m] has quit IRC21:14
*** pro[m] has quit IRC21:14
*** krichter[m] has quit IRC21:14
*** dineshdb[m] has quit IRC21:14
*** connorshea[m] has quit IRC21:14
*** slaf has quit IRC21:14
*** Kinnison has quit IRC21:14
*** coldtom has quit IRC21:14
*** WSalmon has quit IRC21:14
*** valentind has quit IRC21:14
*** swick has quit IRC21:14
*** flatmush has quit IRC21:14
*** benbrown has quit IRC21:14
*** kapip has quit IRC21:14
*** aevri has quit IRC21:14
*** jjardon has quit IRC21:14
*** lchlan has quit IRC21:14
*** persia has quit IRC21:14
*** ironfoot has quit IRC21:14
*** benschubert has joined #buildstream21:14
*** tristan has joined #buildstream21:14
*** tchaik[m] has joined #buildstream21:14
*** ssssam[m] has joined #buildstream21:14
*** jswagner has joined #buildstream21:14
*** reuben640[m] has joined #buildstream21:14
*** tlater[m] has joined #buildstream21:14
*** rafaelff[m] has joined #buildstream21:14
*** mrmcq2u[m] has joined #buildstream21:14
*** nielsdg has joined #buildstream21:14
*** dbuch has joined #buildstream21:14
*** kailueke[m] has joined #buildstream21:14
*** abderrahim[m] has joined #buildstream21:14
*** m_22[m] has joined #buildstream21:14
*** doras has joined #buildstream21:14
*** mattiasb has joined #buildstream21:14
*** albfan[m] has joined #buildstream21:15
*** jjardon[m] has joined #buildstream21:15
*** skullone[m] has joined #buildstream21:15
*** verdre[m] has joined #buildstream21:15
*** waltervargas[m] has joined #buildstream21:15
*** awacheux[m] has joined #buildstream21:15
*** pro[m] has joined #buildstream21:15
*** krichter[m] has joined #buildstream21:15
*** dineshdb[m] has joined #buildstream21:15
*** connorshea[m] has joined #buildstream21:15
*** slaf has joined #buildstream21:15
*** Kinnison has joined #buildstream21:15
*** coldtom has joined #buildstream21:15
*** WSalmon has joined #buildstream21:15
*** valentind has joined #buildstream21:15
*** swick has joined #buildstream21:15
*** flatmush has joined #buildstream21:15
*** benbrown has joined #buildstream21:15
*** kapip has joined #buildstream21:15
*** aevri has joined #buildstream21:15
*** jjardon has joined #buildstream21:15
*** lchlan has joined #buildstream21:15
*** persia has joined #buildstream21:15
*** ironfoot has joined #buildstream21:15
*** irc.acc.umu.se sets mode: +oo jjardon ironfoot21:15
*** cs-shadow has joined #buildstream21:15
*** mohan43u has joined #buildstream21:15
*** Trevinho[m] has joined #buildstream21:17
*** Demos[m] has joined #buildstream21:29
*** segfault3[m] has joined #buildstream21:35
*** asingh_[m] has joined #buildstream22:10
*** gimpnet-irc[m] has joined #buildstream22:21
*** theawless[m] has joined #buildstream23:17
*** cgmcintyre[m] has joined #buildstream23:49
« Sunday, 2019-10-13 Index Tuesday, 2019-10-15 »

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