IRC logs for #buildstream for Monday, 2020-02-24

*** narispo has quit IRC01:11
*** narispo has joined #buildstream01:11
*** tristan has joined #buildstream01:39
*** traveltissues has joined #buildstream06:49
*** narispo has quit IRC06:54
*** narispo has joined #buildstream06:54
*** narispo has quit IRC07:14
*** narispo has joined #buildstream07:15
*** narispo has quit IRC07:22
*** narispo has joined #buildstream07:23
*** santi has joined #buildstream09:30
*** bochecha has joined #buildstream10:00
*** jonathanmaw has joined #buildstream10:06
*** tpollard has joined #buildstream10:07
*** rdale has joined #buildstream10:10
*** santi has quit IRC10:32
*** santi has joined #buildstream10:32
*** lachlan has joined #buildstream10:43
*** lachlan has quit IRC10:57
*** narispo has quit IRC11:00
*** narispo has joined #buildstream11:01
*** lachlan has joined #buildstream11:18
*** lachlan has quit IRC11:24
*** santi has quit IRC11:33
*** santi has joined #buildstream11:34
*** benschubert has joined #buildstream11:34
benschubertjuergbi: what would be your preference on preventing the 'buildbox-casd" timeout at the start? I'm hitting that more than 20% of the time, so I'll spend some time fixing it, just want to know what you think would be best11:35
benschuberthttps://www.irccloud.com/pastebin/eNFTy6Jx/11:36
juergbibenschubert: based on the discussion last week it seems the cause in your case might be different from what is described in #122211:37
gitlab-br-botIssue #1222: Frequent failures  with"timeout waiting for buildbox-casd" https://gitlab.com/BuildStream/buildstream/issues/122211:37
juergbias it happens for you even though `du` is relatively quick11:38
juergbiwhich means we first have to figure out the cause11:38
juergbi(maybe casd is simply much slower than `du` for some reason)11:38
benschubertah fair enough11:38
benschubertOk, guess I'll brush up my C++ and go look at casd :) thanks11:38
juergbibenschubert: as a first step I would remove/increase the timeout and then do some tracing / checking the casd log11:40
juergbii.e., figure out whether it's related to disk usage calculation or not11:40
benschubertI haven't profiled C++ for a while, is 'perf' still the best for this or do you know any handy tool? :)11:40
*** tme5 has joined #buildstream11:44
juergbiit's been a while for me as well. I think I've used kcachegrind in the past. sysprof might also be useful11:45
juergbithere might be better tools now11:46
*** lachlan has joined #buildstream11:52
*** lachlan has quit IRC11:55
*** lachlan has joined #buildstream12:17
benschubertcheers12:20
*** lachlan has quit IRC12:20
*** lachlan has joined #buildstream12:40
*** lachlan has quit IRC12:44
*** lachlan has joined #buildstream12:45
*** lachlan has quit IRC12:49
*** tme5 has quit IRC12:57
*** lachlan has joined #buildstream14:19
*** lachlan has quit IRC14:23
*** lachlan has joined #buildstream14:40
*** lachlan has quit IRC14:58
*** lachlan has joined #buildstream15:03
*** lachlan has quit IRC15:06
*** santi has quit IRC15:51
*** santi has joined #buildstream15:51
*** lachlan has joined #buildstream16:14
*** lachlan has quit IRC16:19
*** traveltissues has quit IRC16:41
*** lachlan has joined #buildstream16:43
*** phildawson-ct has joined #buildstream16:48
*** phildawson_ has quit IRC16:49
*** tme5 has joined #buildstream16:54
*** lachlan has quit IRC16:55
*** lachlan has joined #buildstream17:01
*** lachlan has quit IRC17:05
*** lachlan has joined #buildstream17:14
*** lachlan has quit IRC17:18
*** lachlan has joined #buildstream17:40
*** lachlan has joined #buildstream17:41
*** lachlan has quit IRC17:45
*** tme5 has quit IRC17:46
*** lachlan has joined #buildstream17:57
*** lachlan has quit IRC18:09
*** santi has quit IRC18:10
*** jonathanmaw has quit IRC18:46
*** lachlan has joined #buildstream19:23
*** lachlan has quit IRC19:29
*** narispo has quit IRC21:18
*** narispo has joined #buildstream21:18
*** narispo has quit IRC21:26
*** narispo has joined #buildstream21:27

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