15:31:48 #startmeeting Performance Team 15:31:48 Meeting started Tue Feb 13 15:31:48 2018 UTC and is due to finish in 60 minutes. The chair is DinaBelova. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:31:50 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:31:53 The meeting name has been set to 'performance_team' 15:31:53 hi DinaBelova 15:31:56 hello tovin07 ilyashakhat :) 15:32:04 hi ilyashakhat DinaBelova 15:32:07 good to see you folks :) 15:32:13 okay, so let's get started 15:32:20 with action items first 15:32:23 #topic Action Items 15:32:29 o/ 15:32:30 :) 15:32:44 msimonin yaay! 15:32:45 good to see you sir :) 15:32:47 hi msimonin 15:32:49 okay, going to action items 15:32:53 hi all, long time 15:32:55 last time we have only one 15:33:20 for me to review https://review.openstack.org/#/c/529816/ 15:33:27 frankly speaking it looks good to me 15:33:37 so I posted my +2 there 15:33:40 thank DinaBelova 15:33:41 tovin07 ^^ 15:33:55 ilyashakhat msimonin please take a look as well 15:34:00 yes 15:34:29 ilyashakhat, msimonin you should take a look at it too 15:34:44 #info ilyashakhat msimonin review https://review.openstack.org/#/c/529816/ 15:34:45 ok I'll do 15:34:47 I took a look too, my concern is that the code for error traces looks very similar to the code that processes ordinary traces 15:35:32 so why error traces are not stored into different dbs/indexes, but handled special way? 15:35:50 tovin07 ^^ 15:36:15 I mean list_error_traces is similar to list_traces but reads from different source 15:36:16 yes 15:36:31 it look similar to normal traces 15:36:33 however 15:36:35 currently 15:37:14 we do not have the ability to handle all the traces, therefore, I choose to store just base_id to other db/indexes 15:38:29 I see 15:38:53 okay, let's jump to the current discussion section, and proceed there 15:38:54 #topic Current status 15:39:03 tovin07 please go ahead 15:39:32 basically if we really store only ID and timestamp then the solution looks fine 15:39:43 it is easier than scan the database to get the error traces/spans with etype 15:39:57 yeah, I agree 15:40:17 especially if there is no index on etype 15:40:31 yes 15:40:47 did you verify on all DBs? 15:41:00 I tried on Redis only 15:41:06 yes, ES, redis, and Mongo 15:41:15 but not LogInsight 15:41:42 because I don’t have one 15:42:21 I think we need to care about open-source solutions only :) 15:42:32 agree :D 15:42:40 however, hope that Vipin can help 15:42:44 ilyashakhat I agree :) 15:43:55 tovin07 let's ping him in the commit 15:44:02 yup 15:44:05 I will 15:44:05 :D 15:44:33 okay, ilyashakhat msimonin do you have anything to share? 15:44:44 tovin07 anything else sir? 15:44:44 I've submitted a talk 15:45:11 msimonin cool :) and the topic? 15:45:12 a cited the performance wg in the abstract :) 15:45:18 msimonin: is the link ready? 15:45:35 https://www.openstack.org/summit/vancouver-2018/call-for-presentations/preview/21007 15:45:48 oups sorry 15:46:01 is the link ok ? 15:46:33 oh, it’s not public for now 15:46:37 it shows your submission 15:46:46 OpenStack internal messaging at the edge : in depth evaluation 15:46:48 title "OpenStack internal messaging at the edge : in depth evaluation" 15:46:53 yes 15:47:02 interesting :) 15:47:08 maybe it's available only after login 15:47:11 I do not see it :D As well as tovin07 :D 15:47:16 yeah, probably 15:47:23 We are currently working on evaluating different buses w 15:47:25 you have to log in DinaBelova 15:47:31 yeah, thanks 15:47:39 msimonin go ahead, sorry 15:47:51 with ten thousands of RPCs clients and servers 15:48:22 Someone arrived recently in our team to work specifically o 15:48:25 n this 15:48:39 that's it :) 15:48:46 msimonin sounds really good 15:49:04 challenging task, will be really good to observe results 15:49:15 msimonin please keep us updated 15:49:21 sure ! 15:49:28 thank you sir 15:49:37 ilyashakhat anything from your side? :) 15:49:39 I publish a patch to back port OSProfiler bug fix for OSProfiler issue in Nova 15:49:45 #link https://review.openstack.org/#/c/541645/ 15:50:10 and sent an email to mailing list to get FFE 15:50:12 #link http://lists.openstack.org/pipermail/openstack-dev/2018-February/127253.html 15:50:17 and waiting 15:50:23 tovin07 okay, wating for ffe 15:50:42 DinaBelova, minor thing requires your +W https://review.openstack.org/#/c/541193/ 15:50:43 tovin07 thanks 15:50:59 omg 15:51:02 yes, you're right 15:51:06 done 15:51:17 somebody keeps performance-docs repo alive :) 15:51:32 ilyashakhat true :D 15:51:48 I don't have anything else 15:51:58 ilyashakhat thank you sir 15:52:01 #topic Open Discussion 15:52:17 okay, so open discussion 15:52:44 msimonin tovin07 ilyashakhat - do we have anything to dig into? 15:53:08 this time in Vietnam is holiday season :D 15:53:17 not from my side - I'm quite limited in time now 15:53:28 tovin07 so we're keeping you late in night :D during holiday season :D 15:53:29 do you have New Year? 15:53:34 I feel bad about this 15:53:39 but I still can online 15:53:41 no problem 15:53:56 ilyashakhat: yes, Lunar New Year 15:54:30 Happy New Year :) 15:54:30 DinaBelova: no problem 15:54:40 thank you ilyashakhat 15:54:58 okay, so it looks like we can wrap up :) tovin07 please have a nice Tet! 15:55:07 thank you guys for joining 15:55:11 Thank you 15:55:12 same here tovin07 ! 15:55:16 #endmeeting