14:00:50 #startmeeting Rally 14:00:51 Meeting started Mon Oct 26 14:00:50 2015 UTC and is due to finish in 60 minutes. The chair is ikhudoshyn. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:52 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:54 The meeting name has been set to 'rally' 14:01:03 o/ 14:01:30 hi, everybody. boris-42 asked me to lead the meeting this time 14:01:44 hi 14:03:36 let's wait couple more minutes.. 14:04:26 DST was switched off recently for the huge part of the team so we could expect people missing the meeting today 14:05:37 o/ 14:05:40 Many other participants may opt for (preparation to) summit 14:06:01 Ok lets start 14:06:18 Our agenda does not look busy tday 14:06:22 ikhudoshyn: do we need a meeting at this weekend? 14:06:38 andreykurilin: weekend? 14:06:49 *week 14:06:50 :) 14:06:51 I thought it's Monday 14:07:10 boris-42 asked me to conduct it )) 14:07:17 we're gonna be short 14:07:18 okey:) 14:07:26 #topic [ikhudoshyn] Distributed load generation - https://review.openstack.org/#/c/236979/ - https://review.openstack.org/#/c/238547/ - https://review.openstack.org/#/c/239406/ 14:07:39 I could be the 1st 14:08:09 so, I work on (some parts of) distributed load generation functionality AKA DistributedRunner 14:08:26 by the links you could find a draft for the spec 14:08:51 o/ 14:08:54 as well as a couple of patches which contain some independent parts 14:09:10 do rally droped bps ? 14:09:20 not use bp anymore ? 14:09:31 those include algorythm for merging result streams from remote nodes, running tasks 14:09:47 and infrastructure changes for SLA validation 14:10:04 should I dive in details now? 14:10:12 hello guys, how many guys are in Tokyo now? 14:10:37 kun_huang: hi. Not much I believe 14:11:12 ikhudoshyn: there is a rally related topic in neutron 14:11:13 boris-42 is not there AFAIK, so are most of those who are here 14:11:28 ikhudoshyn: yet, he should be there 14:11:40 kun_huang: agree 14:11:53 ikhudoshyn, if there is a bp, wouldn't it be better to add the bp links on each patch , then we can find all patches related to the same topic from the bp. 14:12:50 chenli: ikhudoshyn I think we should talk about those details in #openstack-rally 14:13:00 it's meeting time :) 14:13:00 chenli: well there is no BP. Rally is mostly spec-driven )) 14:13:29 kun_huang: agree 14:13:37 ikhudoshyn, yep, that's my question, because bp and spec do not confilict at all. 14:13:41 kun_huang, ok 14:14:27 ikhudoshyn: I nearly loss two weeks from rally 14:14:35 watching your https://review.openstack.org/#/c/236979/3/doc/source/images/Rally_Distributed_Runner.png,unified now :) 14:15:39 kun_huang: I believe that means that you're to reach me if any questions arise 14:15:47 so we could move on 14:16:10 any objections? 14:16:39 guess, not 14:16:49 #topic Open discussion 14:17:06 ikhudoshyn: just read a little from your spec 14:17:32 ikhudoshyn: so we could run many runer+runer+... 14:18:07 kun_huang: yes, on separate nodes 14:18:48 and one DistributedRunner on 'master' node 14:19:00 that will rule them all 14:19:44 Is that means rally can run 2 different scenarios at the same time ? 14:19:56 chenli: nope 14:20:12 rally could run 2 different scenarios a long time before 14:20:20 chenli: that is to be able to generate much more load than now 14:20:46 kun_huang, o, how ? Didn't see any examples.. 14:21:22 now, rally would handle the case like users A (call APIs serially) and users B (call APIs concurrenyly) 14:21:45 chenli: https://github.com/openstack/rally/blob/master/rally-jobs/rally.yaml 14:23:07 kun_huang, I still don't understand. When we add more than 1 scenarios in one yaml, all secenarios will be start at the same time ? 14:23:27 kun_huang, then, how I control them to run one by one 14:24:03 chenli: the secenarios run one by one 14:24:26 chenli: no, when you add more that 1 scenario to one yaml/json file, all scenarios will be launched serially 14:24:35 kun_huang, then, rally can't run 2 scenarios at the same time, 14:24:58 andreykurilin, kun_huang right ? 14:25:12 yes 14:29:11 it seems there are no more talks :) 14:29:20 yep )) 14:29:23 :) 14:29:39 looks like we're done 14:30:01 see you 14:30:03 let's end meeting, hope those guys have a nice week 14:30:05 :) 14:30:10 #endmeeting