16:00:14 #startmeeting Performance Team 16:00:15 Meeting started Tue Jan 26 16:00:14 2016 UTC and is due to finish in 60 minutes. The chair is DinaBelova. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:16 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:19 The meeting name has been set to 'performance_team' 16:00:26 hello everyone :) 16:00:27 o/ 16:00:51 I wonder if people are around :) 16:00:57 o 16:00:59 o/ 16:01:10 hehe, cool, I see some human beings :) 16:01:11 it's 12:00 am for me 16:01:37 kun_huang meh :( 16:01:38 DinaBelova, how d'you know? 16:01:50 ikhudoshyn haha :D 16:01:55 if they are humans? 16:02:04 kun_huang - let's try to keep it quick 16:02:24 #topic Action Items 16:02:50 kun_huang - were you able to ping DragonFlow team? 16:03:11 about their test plan sharing? 16:03:23 #link http://eavesdrop.openstack.org/meetings/performance_team/2016/performance_team.2016-01-19-16.00.html 16:03:32 not yet, super busy last week 16:03:45 ok, so I'll left this action item on you for next one as well 16:03:50 let me make a call tomorrow 16:03:53 #action kun_huang ask DragonFly team to share the test plan 16:03:58 kun_huang that will be very cool 16:04:08 DragonFly or DragonFlow? 16:04:13 * regXboi thinks #undo is in order 16:04:52 regXboi - DragonFlow, last time my auto-correction made the meeting messy 16:04:55 :) 16:05:08 ok, if you want - chair me and I'll correct the minutes 16:05:14 as we go 16:05:16 hi all :) 16:05:20 hi 16:05:55 regXboi - oh how can I do that? 16:06:00 it'll be perfect :) 16:06:06 just undo? 16:06:31 here... #chair regXboi and I'll show you 16:06:37 #chair regXboi 16:06:38 :) 16:06:38 Current chairs: DinaBelova regXboi 16:06:41 #undo 16:06:42 Removing item from minutes: 16:06:44 :D 16:06:54 ok, I knew that :D but forgot as usually 16:06:59 #action kun_huang to ask DragonFlow team to share their performance/scale test plan 16:07:05 thank you sir :) 16:07:07 thanks:) 16:07:10 and you are welcome :) 16:07:40 kun_huang - and I suppose we should move one more item for you on next week as well - sharing Huawei lab doc to performance-docs 16:07:56 #action kun_huang publish info about Huawei lab to performance-docs 16:08:18 as for my action items - I had very busy week and did not share the Intel-Mirantis lab info as well :( 16:08:20 I know this, trying in this weekend 16:08:26 so leaving it as is on myself 16:08:42 #action DinaBelova share details about Mirantis-Intel lab on performance-docs 16:08:49 kun_huang - sure, take your time 16:09:20 as for the left action item on me - so yeah, I retested Nova osprofiler change, and the issue was not in what we thought 16:09:51 the issue was fixed successfully and after one more osprofiler release I'll be able to finalize the Nova change 16:09:56 DinaBelova: when we get to OpenDiscussion, remind me - I have a few updates 16:10:04 regXboi ack! 16:10:08 #topic Test plans statuses 16:10:23 so about test plans - DB test plan was merged to the performance-docs 16:10:43 #link http://docs.openstack.org/developer/performance-docs/test_plans/db/plan.html 16:11:00 thanks Jeff again :) 16:11:28 gokrokve - please let him know I'm very glad with his first OpenStack contribution :) it was cool :) 16:11:43 ok. will do 16:12:03 also ilyashakhat has published one more test plan regarding tenant Networking testing 16:12:05 #link https://review.openstack.org/#/c/270204/ 16:12:15 kun_huang - please find the time to review it 16:12:31 I'll look at it as well 16:12:38 regXboi - thank you sir! 16:13:10 in fact we were not able (yet) to start running these test plans against Intel-Mirantis lab, due to the long preparation cycle 16:13:18 will keep you on track folks 16:13:33 I suppose that's pretty all about test plans for this week 16:13:40 #topic Huawei lab status 16:14:00 kun_huang - is the Feb 15th the current estimation for lab to be fully ready? 16:15:00 kun_huang - and did you have a chance to check if at least some nodes are ready already to be used by listomin? 16:15:48 DinaBelova: the ready day maybe delay 16:16:00 kun_huang - any estimations? 16:16:09 2.6~2.13 is chinese new year 16:16:21 should be ready at 3.1 16:16:45 kun_huang - hehe, first I thought what's the strange version strings :D 16:16:53 kun_huang - ok, I got it, very cool 16:17:10 :) chinese style typiing :) 16:17:14 typing 16:17:19 kun_huang - please check if some nodes are ready now (about 3-4 of them?) and if so ping linuxgeek 16:17:23 oh sorry 16:17:26 listomin ***** 16:17:44 ok, very cool 16:17:52 yep I would ping him :) 16:18:04 it looks like we can smoothly move to the OSprofiler news 16:18:07 #topic OSProfiler weekly update 16:18:42 so to proceed we need to cut new OSprofiler release, that needs to contain several changes 16:18:47 lemme list them 16:18:57 #link https://review.openstack.org/270898 16:19:08 #link https://review.openstack.org/269860 16:19:16 #link https://review.openstack.org/268974 16:19:36 after they will be merged, I'll be able to update Nova and Keystone change to their final view 16:19:56 also I'm currently working on Neutron change on my dev env 16:19:59 regXboi ^^ 16:20:05 ack 16:20:16 regXboi - probably I'll need your help in some days, I'll let you know if so 16:20:25 and surely I'll add you to the reviewers 16:20:27 :) 16:20:39 again ack - I'm usually in this channel 16:21:15 DinaBelova: we need to hurry up with Nova's change due to feature freeze 16:21:25 andreykurilin - I know that :( 16:21:31 but I can do nothing without new release :( 16:21:39 OSprofiler release I mean 16:21:48 DinaBelova: is there any output.html for osprofiler? our QA team is trying it 16:21:55 boris-42 is having this task as #1 osprofiler priority 16:22:03 or we can ask johnthetubaguy about feature freeze exception for osprofiler:) 16:22:16 kun_huang - do you mean some example html report? 16:22:22 yep 16:22:31 andreykurilin - let's see if we'll need it after today 16:22:35 andreykurilin: yeah, I was thinking about that one, will try bring that up at the midcycle 16:22:42 its technically needed 16:22:49 johnthetubaguy - thank you sir! 16:22:52 thanks 16:23:01 kun_huang - one moment sir 16:23:04 I have some 16:23:45 #link http://dinabelova.github.io/trace_nova_compute.html 16:23:59 kun_huang ^^ - this is the example of Just Nova tracing 16:24:23 nice! 16:24:26 with recent version of Nova change, that covers all eventlet-related areas as well 16:24:40 so it was nova boot request in fact 16:25:29 note: this particular env suffered from sporadic issues with the hypervisor, so this particular trace is for VM frozen in BUILD state :D 16:25:44 I did not manage to win the fight with the env :D 16:26:00 but still this trace is more or less full 16:26:02 yet 16:26:09 you haven't won the fight ... yet 16:26:18 regXboi :D true :D 16:26:38 ok, I believe it's pretty all about OSprofiler status 16:26:46 #topic Open Discussion 16:26:52 DinaBelova: andreykurilin btw it is possible to backport osprofiler to old release? 16:26:52 regXboi - the floor is your sit 16:26:55 sir* 16:27:16 kun_huang - that will require modified commits for OpenStack services 16:27:16 here at IBM, we've stood up a 125 node cloud to test out OVN for networking 16:27:33 kun_huang - it's possible, but will require the development effort 16:27:33 I'm trying to see if I can winkle out test plans for the repository 16:27:37 kun_huang: no, since osprofiler is a "feature" 16:28:04 regXboi - very interesting! 16:28:14 regXboi I believe that's a good stuff to be published 16:28:41 kun_huang, andreykurilin - so I meant it's possible to do it on the env if that's really needed :) 16:28:58 but not in the official OpenStack - due to what andreykurilin has mentioned 16:29:13 we do not backport features, only bugs 16:29:23 yep, community may not allow a feature backport 16:29:40 but we could make development effort ourselves 16:29:44 kun_huang: *may not* -> *won't* 16:29:50 we can post a backports to stable branches and put W-1, just to simplify usage 16:30:24 folks - I need to run to a call - DinaBelova, you can find me in channel for the neutron stuff 16:30:34 regXboi - ack 16:30:50 kun_huang - so yeah, I believe that's possible 16:30:58 we need to evaluate the efforts 16:31:39 ok, some news from me as well: I've added performance team related proposal for https://wiki.openstack.org/wiki/Internship_ideas 16:31:50 for Outreachy OpenStackProgram 16:32:14 * DinaBelova needs to fix the Outreach period there 16:32:48 so if this will be chosen, we'll have additional help :) 16:32:52 with our initiatives 16:33:11 that's all from my side in fact 16:33:16 anything else? 16:33:57 nothing from me 16:34:01 ok, thank you guys for quick and productive meeting :) 16:34:13 kun_huang - please finish your working day :) 16:34:23 #endmeeting