21:00:19 <jd__> #startmeeting ceilometer
21:00:19 <Slower> stevebaker: wow, 1 second late! :)
21:00:20 <openstack> Meeting started Wed Mar 12 21:00:19 2014 UTC and is due to finish in 60 minutes.  The chair is jd__. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:00:21 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:00:24 <openstack> The meeting name has been set to 'ceilometer'
21:00:26 <jd__> hi
21:00:28 <dragondm> o/
21:00:30 <gibi> o/
21:00:30 <jd__> #link https://wiki.openstack.org/wiki/Meetings/Ceilometer
21:00:33 <eglynn-afk> o/
21:00:33 <ildikov_> o/
21:00:35 <terriyu> o/
21:00:38 <stevebaker> Slower: :)
21:00:39 <_nadya_> o/
21:00:42 <nsaje> o/
21:00:45 <gordc> o/
21:02:34 <jd__> that should be a quick meeting
21:02:40 <jd__> #topic Milestone status icehouse-rc1
21:02:52 <eglynn> https://bugs.launchpad.net/ceilometer/+bug/1288372 is proving a tad awkward
21:02:56 <jd__> so we're pretty much on track, I don't know if VMware stuff are all merged I didn't check
21:03:03 <jd__> no it should be mostly about bugs
21:03:17 <eglynn> current approach on LP 1288372 requires upping the min mongo version to 2.4 to get access to the new $setOnInsert operator
21:03:20 <jd__> eglynn: I feel your pain
21:03:28 <eglynn> LOL :)
21:03:35 <jd__> eglynn: we already require 2.2 right?
21:03:42 <eglynn> yep
21:03:56 <jd__> if we have 2.4 on at least one CI node, I'm good with that
21:04:01 <jd__> (which is not sure)
21:04:08 <jd__> (I think py27 has 2.2 and py26 has 2.3)
21:04:15 <gordc> jd__: i believe vmware stuff is merged... just cleaning up docs now... and probably need to apply fixes to the minor comments.
21:04:17 <eglynn> the pain will be also be felt by the RHEL RPM sausage makers ;)
21:04:21 <jd__> gordc: ack
21:04:32 <eglynn> (only 2.2 packaged on RHEL right now)
21:04:33 <jd__> eglynn: is there any other option?
21:05:02 <eglynn> jd__: I think not, but in any case I'll have a patch out for review tmrw morning, we can trash it out on gerrit ...
21:05:38 <gordc> i put a patch to spawn multiple workers: https://review.openstack.org/#/c/79962 ... local testing shows improve me.
21:05:48 <gordc> s/improve me/improvement
21:06:04 <gordc> only issue is it brings up deadlocking issues again.
21:06:18 <jd__> gordc: uh?
21:07:10 <gordc> jd__: yeah... it's going to need a rewrite of how we store records... or we keep collector as single thread for now.
21:07:14 <jd__> gordc: another improvement you can make is to set the greenpool size to 1 to avoid all races condition
21:07:25 <jd__> gordc: that is part of my plan too
21:07:33 <_nadya_> we have dependency-problem regarding rpm with happybase too. as far as I remember only 0.6 version is available and it doesn't work. need to check 0.8
21:07:48 <eglynn> gordc: does this multi-worker change impact on the API service, or just the agents?
21:07:54 <jd__> gordc: anyhow that patch looks a very good first step, thanks
21:08:15 <gordc> jd__: i'll look into that. the problem is the updates we make to (useless) secondary tables (source, project, resource, etc...)
21:08:29 <eglynn> gordc: ... /me just wondering about the old behaviour of the out-of-the-box API service being locked up during long-running API requests
21:08:36 <gordc> eglynn: i only turned on multiple workers for collector and notification services.
21:08:43 <eglynn> gordc: fair enough
21:08:58 <gordc> eglynn: i figure it makes no sense for agents and api...
21:09:15 <jd__> indeed
21:09:31 <jd__> anything else?
21:09:40 <eglynn> gordc: yeah I guess there's an alternative way of making the api service able to handle concurrent requests
21:09:51 <jd__> eglynn: real people uses a WSGI server :)
21:10:01 <eglynn> jd__: yeap
21:10:15 <gordc> jd__: not from me. i'll keep playing with it to see if we can avoid deadlocks.
21:10:26 <jd__> cool
21:10:36 <jd__> #topic Tempest integration
21:10:43 <jd__> _nadya_: any update?
21:10:51 <_nadya_> hi all. We were working a lot on notification problem and as a result is #link https://bugs.launchpad.net/ceilometer/+bug/1291054 . I hope that the fix will allow us to test ceilometer in tempest. As I wrote in email [Collector's performance], all tempest tests creates a rather high load on ceilometer services. So the plan is to test the fix with changed polling interval. Because I plan to create pollster-tests by the end of week. An
21:10:51 <_nadya_> d if test is not successful we will discuss what do do on next meeting :)
21:11:54 <jd__> sounds like a plan :)
21:12:32 <_nadya_> afaik, ci-nodes has 8 cpus a least. so with Gordon's patch it should work 8x faster
21:12:41 <jd__> yep
21:12:47 <gordc> _nadya_: lol. i like the optimism.
21:13:01 <jd__> we'll add rainbows later gordc :)
21:13:18 <gordc> perfect!
21:13:21 <_nadya_> :)
21:13:22 <jd__> #topic Release python-ceilometerclient?
21:13:26 <eglynn> still patches in the queue for the new icehouse features
21:13:31 * eglynn needs to find his reviewing mojo ;)
21:13:38 <jd__> hehe
21:13:41 <ildikov_> our patch sets are all ready for review
21:13:44 <eglynn> will aim to release next week once the backlog has landed
21:13:46 <jd__> cool, we still have time to release
21:13:55 <eglynn> ildikov_: noted :)
21:14:09 <ildikov_> eglynn: thanks :)
21:14:10 <nsaje> ours too :)
21:14:24 <jd__> #topic Open discussion
21:14:44 <eglynn> terriyu brought up the OPW last week
21:14:51 <eglynn> so we've had some movement on that ... spun out an idea
21:14:58 <eglynn> https://wiki.openstack.org/wiki/OutreachProgramForWomen/Ideas#Ceilometer_-_Period-spanning_statistics
21:15:05 <eglynn> and interest has been expressed in it ...
21:15:14 <terriyu> yes, a friend of mine from grad school is working with eglynn on it
21:15:39 <eglynn> yep, so if you have the opportunity to help anamalagon with her pre-application work, please do so
21:15:40 <terriyu> her nickname is anamalagon and she's finishing up her physics PhD
21:16:03 <eglynn> (it would be good for Ana to have landed at least one ceilo patch before the OPW deadline of Mar 19th)
21:16:25 <terriyu> so far, she seems to be doing very well, so if we help her as much as possible, she has a great shot at getting into OPW and becoming a Ceilometer intern this summer
21:16:59 <gordc> cool, i'll keep an eye out for her work.
21:17:17 <ildikov_> if she succeeds with sphinx, then nothing can stop her :)
21:17:29 <gordc> lol
21:17:44 <terriyu> eglynn: do you have a link to the patch she's working on right now?
21:17:46 <ildikov_> I can offer some help too, if she needs, even with sphinx
21:18:09 <eglynn> terriyu: https://bugs.launchpad.net/ceilometer/+bug/1290924
21:18:53 <_nadya_> please remind me the date of rc1
21:18:53 <terriyu> everyone, this is the bug that anamalgon is working on ^^
21:19:23 <_nadya_> *sorry for interrupting
21:19:28 <eglynn> _nadya_: https://wiki.openstack.org/wiki/Icehouse_Release_Schedule
21:19:42 <eglynn> _nadya_: ... so w/c Mat 27th I guess
21:19:46 <eglynn> *Mar
21:20:04 <_nadya_> eglynn: ok, thanks
21:20:39 <jd__> yeah though we can release rc1 as soon as we're ready
21:20:57 <eglynn> jd__: a-ha ok, cool
21:21:17 <eglynn> ... but ttx still pulls the trigger on it, or?
21:21:22 <gordc> jd__: have we branched already? just wondering when we are able to approve non-icehouse stuff.
21:21:53 <eglynn> gordc: usually coincides with RC1 being cut IIRC
21:22:09 <jd__> yep
21:22:20 <gordc> eglynn: cool cool. i was waiting for jd__ to start tossing the usually red x's around for ff
21:22:36 <jd__> I'm lazy this time, I don't think I'll do that
21:22:37 <eglynn> ... his favourite time of the cycle ;)
21:22:57 <jd__> I envision a world where every core reviewer is smart enough to not press the button
21:22:57 <gordc> it'd be mine too :)
21:23:02 <jd__> :D
21:23:45 <gordc> jd__: i need to disable my approve radio button... it's calling to be pressed.
21:25:32 <eglynn> jd__: how long to leave the core nominations for _nadya_ & ildikov_ open?
21:25:39 <jd__> eglynn: 5 days IIRC
21:25:42 <eglynn> ... tradition is what, 5 working days?
21:25:45 <eglynn> cool
21:26:15 <_nadya_> ildikov_: ;)
21:26:31 <ildikov_> _nadya_: ;)
21:26:49 <eglynn> ... 7 of 8 votes in so far, it's looking like a shoo-in :)
21:26:59 <ildikov_> _nadya_: here comes the women power ;)
21:27:13 <_nadya_> ildikov_: hehe
21:27:24 <eglynn> ... /me hides ;)
21:27:49 <ildikov_> eglynn: LOL :)
21:28:36 <jd__> eglynn: ah so basically we already have a majority?
21:29:02 <eglynn> jd__: yep every core save nijaba has +1'd already
21:29:54 <jd__> pff I totally missed the count indeed
21:30:08 <jd__> well, no need to wait I'll add you guys tomorrow and will send a mail ton confirm
21:30:22 <eglynn> jd__: nice!
21:30:48 <eglynn> ildikov_, _nadya_: welcome to the inner sanctum! :)
21:30:50 <gordc> congrats to both of you.
21:30:52 <jd__> anything else or should I close that meeting?
21:31:02 <nsaje> _nadya_ ildikov_ congratulations :)
21:31:20 * terriyu claps for _nadya_ and ildikov_
21:31:23 <_nadya_> thanks :)
21:31:29 <ildikov_> I assume it is tha part, when I can thank to my parents and all the cores for helping all the way through :)
21:31:39 <ildikov_> s/tha/the/
21:32:12 <gordc> :)
21:32:15 <terriyu> _nadya_: speech?
21:33:17 <_nadya_> oh, I'm not so prepared :) I promise to be a responsible core :)
21:33:34 <jd__> :)
21:33:36 <terriyu> :)
21:34:54 <eglynn> _nadya_: nah, we need more irresponsibility ;) ... swash-buckling and risk-taking
21:35:28 <_nadya_> eglynn: really:)? I will think about it
21:35:40 <eglynn> _nadya_: ... I jest ;)
21:35:56 <ildikov_> eglynn: as I mentioned girl power is here, we will not disappoint you ;)
21:36:11 <eglynn> ildikov_: yeah I'm counting on that :)
21:36:45 <ildikov_> eglynn: awsome ;)
21:37:46 <jd__> wrapping up gentlemen and gentlwomen
21:37:56 <jd__> see you next week and all the time on #openstack-ceilometer
21:38:04 <eglynn> good night all!
21:38:09 <gordc> laters
21:38:09 <jd__> happy hacking!
21:38:12 <jd__> #endmeeting