15:30:01 #startmeeting Performance Team 15:30:02 Meeting started Tue Jan 23 15:30:01 2018 UTC and is due to finish in 60 minutes. The chair is DinaBelova. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:30:03 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:30:05 The meeting name has been set to 'performance_team' 15:30:12 hi! 15:30:19 hello ilyashakhat o/ 15:31:00 let'ssee if others are around 15:31:05 hi 15:31:07 o/ 15:31:08 tovin07 o/ 15:31:14 hey :) 15:31:25 it looks like we may start then :) 15:31:37 msimonin will probably join later 15:31:40 #topic Action Items 15:31:48 okay, last time we had two action items 15:32:02 one on myself with reviewing osprofiler review queue 15:32:27 I have done several reviews 15:32:36 it looks like I have to go through some of them again 15:32:41 due to the new patches 15:32:45 so will do it again 15:33:02 tovin07 the other one was on you about https://bugs.launchpad.net/osprofiler/+bug/1743586 15:33:03 Launchpad bug 1743586 in osprofiler "Nova operations fail when profiling is enabled" [Undecided,New] 15:33:38 tovin07 anything to share? 15:34:17 i published a patch for this 15:34:33 actually this one 15:34:35 #link https://review.openstack.org/#/c/535219/ 15:34:46 yep, was trying to find the link :D 15:34:47 thanks 15:35:17 tovin07 did you have a chance to verify this patch? 15:35:42 I tried with my local devstack 15:36:05 ilyashakhat can you check it again 15:36:09 tovin07_, what is the root cause of this issue? 15:36:28 as I see 15:36:45 boris-42 want to trace greenthread too 15:36:52 however, in some cases 15:37:19 there is no new greenthread 15:38:17 you can see in my comment (in the patch or in the bug in launchpad) 15:39:41 we are using threading.local variable - is it suppose to be "forked" when a new green thread is started? 15:40:12 my observation shows that it remains the same for all "green" threads 15:41:27 tovin07_, what do you mean under "there is no new greenthread"? 15:41:47 i mean that 15:42:49 the new thread that spawn-ed is just normal thread 15:43:03 hmm 15:43:33 then our threading.local variable should be different for that new thread, isn't it? 15:43:56 because, if it is green, its class will be eventlet.corolocal.local not thread._local 15:44:23 right 15:45:27 you can see here https://github.com/eventlet/eventlet/blob/master/eventlet/green/thread.py#L113 15:46:16 ok, the fix resolves the issue on devstack 15:46:49 the question now is now to deliver it (since we already passed the freeze day for non-client libs) 15:46:59 now -> how 15:48:17 well, yeah... 15:48:33 I suspect we missed the date and nothing can be done much here 15:49:05 but it means that profiling will be broken in Q 15:49:32 yeah.... tovin07_ can you please reach oslo core team? and speak on this? 15:49:44 so we can release our lib now? 15:49:58 as an exception? 15:50:10 yes, i will try 15:50:31 tovin07_ thanks 15:50:51 #info tovin07_ will try to reach oslo core team on releasing osprofiler now 15:51:15 okay, so let's go ahead 15:51:16 #topic Current status 15:51:17 ilyashakhat please go ahead 15:51:46 I've rebased patch with `trace show` command 15:52:01 I mean `trace list` 15:52:22 so it doesn't refer to ceilometer anymore 15:52:37 yeah, I've seen it 15:52:39 thanks 15:53:18 that's it from my side 15:53:44 ack, thank you 15:53:53 tovin07_ - anything we did not mentioned yet to share? 15:54:01 ah no 15:54:07 except 15:54:46 will any one plan to join Vancouver Summit 15:54:55 just curious :D 15:55:15 heh, good for you :D 15:55:16 I guess I on't be able to join 15:55:17 but I'm glad you will :) 15:55:26 okay, then thank you folks 15:55:27 #topic Open Discussion 15:55:37 tovin07_ ilyashakhat anything to share additionally? 15:55:49 nope 15:55:53 no 15:56:07 I've verified the patch for Nova - it works 15:56:42 cool 15:56:49 tovin07_ then please speak to core team 15:56:54 i don't see gcb online now 15:57:14 let's merge it first 15:57:18 i will reach him tomorrow 15:57:23 ack 15:57:37 i'll do it now 15:57:43 anyway, thanks for your tiem 15:57:47 please have a nice evening 15:58:00 have a nice day! 15:58:08 #endmeeting