13:00:13 #startmeeting senlin 13:00:14 Meeting started Tue Apr 5 13:00:13 2016 UTC and is due to finish in 60 minutes. The chair is Qiming. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:00:15 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:00:17 The meeting name has been set to 'senlin' 13:00:37 evening/morning 13:00:45 hi 13:00:45 hi 13:00:55 #link https://wiki.openstack.org/wiki/Meetings/SenlinAgenda 13:00:55 o/ 13:01:03 pls check if you have items to add 13:01:23 Hi 13:01:49 #topic Rally testing intro 13:02:11 Yanyanhu, could you share with us something? 13:02:11 my part? 13:02:15 sure 13:02:24 who is rally expert in the team? 13:02:30 in last two weeks, I worked on Senlin support in Rally 13:02:39 haiwei, I guess :) 13:02:56 not an expert, Yanyanhu 13:03:09 we try to leverage Rally's scenario test ability to support stress/performance test of senlin api/engine services 13:03:19 haiwei, met neither :) 13:03:33 so the first step is adding senlin support into Rally 13:03:46 saw your patch 13:03:56 two patches have been proposed here https://review.openstack.org/298109 13:04:02 and here https://review.openstack.org/301522 13:04:03 rally is not using tempest model? 13:04:20 not very sure about it, but I guess so 13:04:26 Will it overlap with existing tempest scenarios? 13:04:41 user can define a test job using decription file 13:04:53 elynn, I guess there could be overlap 13:05:53 Okay, then two codes need to be maintained... 13:05:54 but IMHO, Rally is very suitable for performance test 13:05:59 I'm wondering what's the advantage of using both rally and tempest 13:06:33 Rally site seems to imply that Rally runs tempest test under the covers. 13:06:36 is it legitimate to create 1000 VMs ? 13:06:43 elynn, actually, for senlin support of Rally, all codes will be submitted to Rally project 13:07:04 Qiming, that's possible 13:08:10 I'm a little bit concerned with patch #301522, because we are (will be) duplicating quite some code 13:08:21 cschulz, nice. If so, we can consider to integrate the senlin rally test and tempest support 13:08:39 Qiming, that's true 13:08:43 I think so. 13:09:38 Qiming, but I'm not sure how to let rally support senlin based on our current work about tempest? 13:09:44 OK, then the question would be how to integrate these two things. 13:09:55 if test cases are written in the tempest plugin model 13:10:03 should they live in senlin code base instead? 13:10:06 I guess that means we need to contribute some support about senlin to tempest project? 13:10:24 Qiming, not sure how rally can use them if so 13:10:40 it is just stevedore plugin 13:10:45 will make investigation 13:10:49 Yanyanhu, with tempest plugin , we don't need to touch tempest project. 13:24:43 am I still alive? 13:24:43 Qiming, yes 13:24:43 Qiming: Yes. 13:24:43 hello, world! 13:24:43 again 13:24:50 okay, ping you tomorrow morning 10am bejing time? 13:24:56 ok, I will draw a graph and mail to you 13:24:58 we can have a quick call on this 13:25:00 thanks 13:25:11 I am fine with it 13:25:17 my whole afternoon is about ...s..h...i..t meeting 13:25:28 me too. But I'm only free on Friday 13:25:46 have visa review on Thursday 13:25:54 oh, xinhui sent me a text this evening 13:26:01 Qiming, our patent review is on wednesday? 13:26:23 she wants me to talk to vmware chief architect about senlin 13:26:29 will do that on Friday morning 13:26:30 so your budget is ok, Yanyanhu? 13:26:44 haiwei, yes, but I'm a little worried about the visa 13:26:47 will update you if we have some outcomes from the discussion 13:26:59 I am also waiting for the visa 13:27:02 #topic newton work items 13:27:12 #link https://etherpad.openstack.org/p/senlin-newton-workitems 13:27:42 if the application go into administrative processing, it costs at least 3 weeks... 13:27:50 I pinged china mobile, hopefully we can schedule a meeting on scalability testing and experience sharing 13:28:00 Qiming, cool 13:28:03 that's very helpful 13:28:21 elynn, tempest infra is ready? 13:28:29 yes 13:28:38 removing line 8 13:29:03 Initial patch is ready for review. 13:29:18 Just feel free to +-1 13:29:24 https://review.openstack.org/#/c/295647/ 13:29:24 did I +2 it? 13:29:34 Not yet. 13:29:53 ah, will check again 13:30:03 I think Mark's suggestion is a good one 13:30:18 The rest of it would be move the functional tests to it and add negative tests. 13:30:23 was waiting for your response 13:30:57 Qiming: Yes, but I'm haven't find a way to generate a tempest config file... 13:31:10 that is easy 13:31:27 just a separate conf file for oslo.config generator 13:31:41 then you specify where the output will be placed 13:31:55 check the tools/ subdir 13:32:05 there is the conf file for senlin as a whole 13:32:15 If we generate it under senlin repo, can I get the options from tempest? 13:32:21 for tempest, you may need just one or two modules I believe 13:32:56 you can try that, not 100% sure 13:33:04 well, I will have to try. 13:33:17 what options? 13:33:26 Will seek your help if you have time :) 13:33:26 anyway we will have dependency on tempest because we are using its plugin/lib 13:33:49 so we can certainly reference its module when generating conf file 13:34:07 #action Elynn to sync with Qiming on tempest conf file generation 13:34:39 But I don't see other project add tempest to their requirements.txt. 13:34:55 since xinhui is not online today, we can skip the health management part 13:35:07 it should be in test-requirements.txt 13:35:26 we do have 'from tempest import blablah' right? 13:35:29 Oh, I see... 13:35:44 Will update my patch :) 13:35:58 xinhui told me that she is looking into the fencing support 13:36:11 hopefully she can make some progress this coming week 13:36:19 elynn, Qiming, can we have a f2f talk tomorrow to sync our understanding on tempest and rally? 13:36:48 +1 from me 13:36:48 or a call 13:36:53 Morning or afternoon are both ok for me Yanyanhu. 13:37:02 maybe starbucks time again? 13:37:11 But noon is not... 13:37:16 bingo 13:37:20 oh... 13:37:34 ok, let me try to arrange a phone call tomorrow morning 13:37:46 just allocated 10am-11am to haiwei 13:37:48 maybe at 10:30 am? 13:37:51 Have a training at 12~13 oclock. 13:38:06 my whole afternoon was occupied as well 13:38:15 then 9am? 13:38:29 a little early for me... 13:38:42 maybe 11:00 to 11:20? 13:38:44 9 is ok for me. 13:38:47 just after the call with haiwei 13:38:53 I don't think it will take much time. 13:38:54 that's fine 13:38:54 since I will also join that call as well 13:39:05 Okay 13:39:07 just a quick talk will be ok 13:39:12 I suggest we spend more time on studying rally 13:39:24 Qiming, sure, just try to sync our understanding 13:39:28 before the meeting to make the meetings more efficient 13:39:31 but anyway 13:39:39 11:00am okay for me 13:39:54 ok 13:40:11 no documentation progress update from me 13:40:32 API microversioning is already in 13:41:13 cschulz, any thing to share from you? 13:41:25 Yes, 13:41:28 semi-autoscaling and zaqar support 13:41:49 I've gotten through the IBM administration stuff to contribute. So can do that now. 13:42:01 \o/ 13:42:11 Been working with Zaqar some. to become familiar. 13:42:13 That's great! 13:42:36 I'd like to have a talk with you about contribution. Can we have a call? 13:42:49 yes 13:42:52 Choose a time I'll be available. 13:43:15 sure, feel free to schedule a call :) 13:43:34 OK. Your morning OK? 13:43:41 not tomorrow 13:43:53 you can check my calendar availability 13:43:53 OK Thurs. maybe. 13:44:28 Thusday morning good for me 13:44:50 #topic open discussions 13:45:42 any thing, from anyone? 13:45:57 we have 15 minutes to complain about visa now, ;) 13:46:03 ... 13:46:04 hi, guys, I want to quit earlier today, just got a cold 13:46:13 ...Hope I can pass... 13:46:15 haiwei, have a good rest 13:46:20 elynn, me too 13:46:23 sure, haiwei, good night 13:46:25 i am worrying visa tooo 13:46:27 it's late for you 13:46:30 see u 13:46:36 haiwei, it should be ok for U I think 13:46:48 :) 13:46:49 haiwei: have a good night 13:46:50 88 13:47:07 bye 13:47:21 I'm little worried because my college 13:47:41 anyway, good luck for me 13:47:43 just tell them you are from beihang, not BUAA 13:47:46 and for elynn as well :) 13:47:55 good luck for us 13:47:56 yes, I write it in my resume 13:47:57 elynn has the same problem? 13:48:03 but they know it, you know 13:48:12 he doesn't :) 13:48:22 No, my university is not that good :P 13:48:23 Pehang 13:48:38 sigh... 13:49:55 seriously, anything else you want to talk about? 13:50:04 nope from me 13:50:21 Is newton starts? 13:50:22 Nothing here 13:50:23 we can call it a day if no 13:50:38 yes, master is already about newton 13:50:50 this week is the last one before mitaka release 13:51:03 thanks gusy 13:51:06 guys 13:51:07 ok, got it 13:51:18 #endmeeting