17:03:15 #startmeeting rally 17:03:15 Meeting started Tue Jul 15 17:03:15 2014 UTC and is due to finish in 60 minutes. The chair is boris-42. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:03:16 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:03:18 The meeting name has been set to 'rally' 17:03:24 coolsvap hi 17:03:30 rediskin hi 17:03:33 hello 17:03:33 olkonami hi 17:03:41 hi 17:05:34 so let's start=) 17:05:45 #topic OSprofiler Integration 17:06:09 1) So HMAC keys shouldn't be used, we should use PKI approach 17:06:20 This will be covered by DuncanT- 17:06:29 k4n0 hi 17:06:35 boris-42, Hi, sorry for being late 17:07:03 Hopefully I'll have a patch tomorrow 17:07:18 2) we should be able to support list of secret keys (that allows us to do rolling upgrade of services, with downtime) 17:07:22 DuncanT- thanks 17:07:37 that will be covered by me (I am going to make it today/tomorrow) 17:07:59 Patch that allows Ceilometer to collect profiling data from Keystone is almost merged 17:08:14 https://review.openstack.org/#/c/103420/ 17:08:29 So after this KPI stuff will be changed 17:08:43 I will update one more time patches in Glance / Kyestone / Cinder 17:09:13 so I hope they will be merged in Juno 17:09:21 hi. sorry 17:09:25 oanufriev hi there 17:09:38 So let's move to next topic 17:09:44 if no questions? 17:10:13 #topic refactoring periodic runner (new name RPS) 17:10:21 oanufriev could you share details about your work 17:10:24 oanufriev to others 17:10:53 whell. there is the patch on review https://review.openstack.org/#/c/102363/ 17:11:07 oanufriev does it work?) 17:11:38 yes. i have results on 20000 times 100 rps dummy runner 17:12:28 without swapping and lugs 17:12:33 oanufriev what is the name of task? 17:12:36 http://logs.openstack.org/63/102363/14/check/check-rally-dsvm-rally/603de6c/rally-plot/results.html.gz 17:12:38 oanufriev ^ 17:13:02 oh I found it=) 17:13:14 oanufriev hmmm 17:13:27 oanufriev why there is one or couple of iterstation that works for 20 seconds?? 17:15:05 oanufriev still here?) 17:15:58 okay moving to next topic 17:16:07 we will discuss this in rally chat 17:16:20 #topic Functional testing 17:17:01 check-rally-dsvm-cli now works 17:17:08 http://logs.openstack.org/99/102899/22/check/check-rally-dsvm-cli/acc5502/ 17:17:18 and you can take a look at results.html.gz 17:17:22 http://logs.openstack.org/99/102899/22/check/check-rally-dsvm-cli/acc5502/results.html.gz 17:17:28 that show that everything passed 17:17:29 woot 17:17:30 =) 17:17:35 nice :) 17:18:20 k4n0 so when you finish your work on --josn 17:18:28 k4n0 we will be able to test whole CLI 17:18:35 k4n0 in gates on every patch 17:18:40 boris-42, yes, will be done early morning tomorrow 17:18:44 boris-42, or even tonite 17:18:54 #topic unit tests 17:19:01 sorry again 17:19:04 coolsvap could you share that?) 17:19:13 coolsvap e.g. what is the progress and so on 17:19:20 boris-42, sure 17:20:01 I have been working on it for some days now, got 8 patches of which 6 are merged 17:20:15 the coverage has increased by 1-2% in that 17:21:05 also reviewing patches and would recommend everyone to add unit tests in the patch itself 17:22:40 boris-42, thats all from my side 17:23:03 coolsvap ok great and nice work 17:23:17 coolsvap btw you can review not only unit tests of patches 17:23:29 coolsvap but whole patch, that will help a lot core team 17:23:45 #topic Incubation Request 17:23:58 Okay guys I think it's time to put Incubation Request 17:24:17 so during these days I am going to fill all documents that are required by this 17:24:30 And as well we should have official PTL 17:24:36 so we should make PTL election 17:24:50 so don't miss email in mailing list 17:24:52 =) 17:24:57 boris-42, do we need elections :) 17:24:58 k4n0 coolsvap harlowja marcoemorais ^ 17:25:07 k4n0 yep we need=) 17:25:12 hmmm 17:25:18 there is no formal PTL in rally lol=) 17:25:31 dear leader boris-42 17:25:52 *not to be confused with deal leader kim-jong-il 17:25:53 harlowja lol 17:25:54 harlowja, "dear leader" haha 17:26:28 boris-42, are we missing anything critical for incubations? 17:26:43 k4n0 I don't think so 17:26:52 k4n0 we have community , a lot of cores 17:27:08 k4n0 we have functional gates and as well not so bad coverage by unit test =) 17:27:21 is dear leader boris-42 sick? 17:27:21 boris-42, sounds good 17:27:30 harlowja lol=) 17:27:50 k4n0 so I'll fill incubation template and we will see if something is missing=) 17:27:57 boris-42, sure 17:28:32 boris-42, :) 17:28:54 #topic free discussion 17:29:10 so guys do you have any topics that you wanna speak about 17:29:11 ?) 17:29:35 my firefox dies on http://logs.openstack.org/63/102363/14/check/check-rally-dsvm-rally/603de6c/rally-plot/results.html.gz 17:29:54 when i try to check Dummy.dummy #1 17:30:06 boris-42 likes to kill browsers 17:30:15 harlowja serial morder =) 17:30:23 def 17:30:35 oanufriev heh open it in chrome 17:30:42 oanufriev at that scale it works only in chrome 17:30:48 oanufriev or just open text results 17:31:04 oanufriev why there is one or couple of iterstation that works for 20 seconds?? 20 seconds == 20000ms. and we have 20000 iterations.... may be some misstake presents& 17:31:07 ? 17:31:11 oanufriev http://logs.openstack.org/63/102363/14/check/check-rally-dsvm-rally/603de6c/rally-plot/results.json.gz 17:31:23 oanufriev nope 17:31:35 oanufriev if you take a look at graph you'll see 17:33:03 oanufriev oh sorry 17:33:05 oanufriev my bad 17:33:15 oanufriev it's 0.1 sec 17:33:20 oanufriev instead of 0.01 17:34:05 Btw guys I forgot to say 17:34:21 #topic Scalability of Rally 17:34:30 There are couple of changes related to scalability of rally 17:34:35 one of them is already merged 17:35:00 boris-42, the workers patch? 17:35:03 https://review.openstack.org/#/c/104518/ 17:35:16 ^ nope this one that allows async to deal with results 17:35:33 next one will be about chunking data (and storing separated chunks) 17:35:50 e.g. chunk shouldn't have more then 100k iterations 17:35:59 or some configurable option 17:36:23 and there is patch 17:36:24 https://review.openstack.org/#/c/106031/ 17:36:28 on review ^ 17:36:36 that reduces amount of points on graphs 17:36:51 that allows us to avoid killing everybody browsers=) 17:36:55 if there is too many points=) 17:37:20 any questions about scalability? 17:38:22 #topic free discussion 17:38:29 Okay guys anything that you would like to discuss ? 17:38:36 boris-42, nothing from my side 17:38:43 coolsvap harlowja k4n0 oanufriev olkonami ? 17:38:45 rediskin ? 17:38:50 ummm 17:38:58 boris-42 when can that music thing be ready? 17:39:07 * harlowja needs to listen to ma tunes 17:39:12 openstack tunes 17:39:14 harlowja for http://boris-42.github.io/profiler/ ?) 17:39:16 ya 17:39:31 harlowja I need to find some lib for playing music lol) 17:39:50 http://kolber.github.io/audiojs/ 17:39:56 ^ seems like ready=) 17:40:06 nothing from my side 17:40:18 boris-42 niceee 17:40:26 boris-42 so u need to integrate that into osprofiler 17:40:26 lol 17:40:38 harlowja yep we can have for example 100 tons 17:40:41 tones* 17:40:57 ya, make 10000ms things use the slow tone, and such 17:41:04 yep yep 17:41:09 boris-42 for those that are visually impaired that will help them 17:41:12 useability !! 17:41:13 lol 17:41:16 =) 17:41:33 don't be discrimatatory 17:41:37 ha 17:41:45 *discriminatory 17:42:07 okay I want go and do something=) 17:42:12 so let's finish this meeting 17:42:21 if somebody has any questions ping in rally 17:42:24 #endmeeting