15:00:21 #startmeeting ceilometer 15:00:22 Meeting started Thu Nov 28 15:00:21 2013 UTC and is due to finish in 60 minutes. The chair is jd__. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:23 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:25 The meeting name has been set to 'ceilometer' 15:00:29 o/ 15:00:36 o/ 15:00:40 o/ 15:00:45 happy turkey day 15:00:49 o/ 15:01:02 * eglynn prefers goose ;) 15:01:13 o/ 15:01:15 :) 15:01:18 o/ 15:01:41 duck for me ;) 15:02:14 #topic icehouse-1 status 15:02:25 #link https://launchpad.net/ceilometer/+milestone/icehouse-1 15:02:41 so we're still rather right on track 15:02:56 except for one low prio blueprint that I am not sure I know what it is about 15:03:15 just need to review llu work 15:03:38 for people not aware, Icehouse-1 release is in one week 15:03:42 good to hear that :) 15:03:45 so perhaps bump https://blueprints.launchpad.net/ceilometer/+spec/zabbix-agent-adoption to i-2? 15:04:22 eglynn: yes, just did :) 15:04:35 cool :) 15:04:43 I'll try to send patches on counter renaming, if any, soon 15:04:52 but it's not a big deal if it slips to i2 15:05:08 #topic progress on writing up blueprints for icehouse 15:05:44 has everybody written everything? 15:05:49 #link https://blueprints.launchpad.net/ceilometer/icehouse 15:06:03 #link https://blueprints.launchpad.net/ceilometer/icehouse 15:06:06 arf :-) 15:06:35 nadya_: this one should be targeted also? https://blueprints.launchpad.net/ceilometer/+spec/aggregation-and-rolling-up 15:06:55 (or too premature as yet?) 15:06:59 eglynn: I'll set the status to "Not started", is that ok for you? 15:07:12 eglynn, I think yes. We may try to start it at least 15:07:15 jd__: can you target https://blueprints.launchpad.net/ceilometer/+spec/monitoring-physical-devices to icehouse-2 ? 15:07:19 jd__: yep 15:07:30 llu-laptop: sure, done 15:07:45 jd__: thx 15:08:30 I'll set priorities on the ones without after the meeting 15:08:36 anything else gentlemen? 15:09:24 #topic Telemetry: rename? 15:09:38 so the Ceilometer project has now an official name, being "OpenStack Telemetry" 15:10:00 cool, that governance patch landed? 15:10:05 (thanks to our Official Namer of Things) 15:10:08 eglynn: it did 15:10:10 :) 15:10:13 #link https://git.openstack.org/cgit/openstack/governance/commit/?id=d66f47d89616d2832af3ead686d1469fcdcd3236 15:10:29 I saw official email about it :) 15:10:34 I've updated the Launchpad page 15:10:46 I saw the mail too today :) 15:10:47 the question is: should we rename everything, including this channel? 15:11:03 do we have some official answer template when being asked why naming Telemetry? :) 15:11:10 not this channel, but Ceilometer's one, sorry :) 15:11:27 llu-laptop: "because why not?" 15:11:41 jd__ probably for longer term it would worth it to rename 15:11:48 llu-laptop: "ask the Irish guy"? 15:12:02 I think we should probably rename wiki pages and the irc channel 15:12:03 jd__: so we keep metering channel active with topic saying to go to openstack-telemetry? 15:12:05 I tend to agree for a rename of things too 15:12:14 gordc: something like that 15:12:16 llu-laptop: because it was the only name we could think of that didn't seem to exclude a big chunk of our expanded project mandate 15:12:20 works for me 15:12:28 the other teams use their code name for their irc channel, so maybe instead of openstack-telemetry that should be openstack-ceilometer? 15:12:35 llu-laptop: you see, "ask the Irish guy" was the right one :) 15:12:41 LOL :) 15:12:53 jd__, eglynn: :) 15:12:58 dhellmann: i'm good with that too. 15:13:05 dhellmann: agreed 15:13:08 dhellmann: agreed 15:13:28 dhellmann: agreed 15:13:42 should we rename https://wiki.openstack.org/wiki/Meetings/MeteringAgenda? 15:13:48 #action jd__ rename #openstack-metering to #openstack-ceilometer 15:13:55 llu-laptop: yes 15:14:09 s/MeteringAgenda/Telemetry/ 15:14:10 +1 for the IRC rename, better to follow the general patterm 15:15:00 cool, cool 15:15:12 a lot of grunt work :-) 15:15:50 feel free to change or notify me if you encounter any occurence that needs to be changed 15:15:56 #topic Release python-ceilometerclient? 15:16:09 already did earlier today 15:16:17 here's the tag desc: http://fpaste.org/57485/38564721/raw/ 15:16:18 yay thanks 15:16:27 #topic Open discussion 15:16:39 WRT to i-1 targeted bugs 15:16:47 just to re-iterate the old mantra about backporting to stable 15:16:53 * eglynn is like a broken record ... 15:16:59 you know the blast radius of your own bug fix the best 15:17:14 so please take a moment after it lands on master to consider whether a suitable backport candidate 15:17:28 #link https://wiki.openstack.org/wiki/StableBranch#Appropriate_Fixes 15:17:34 critera there ^^^ 15:17:49 and if so, tag the bug with 'havana-backport-potential' in LP 15:18:01 or even better, cherry-pick and propose the backport yourself ;) 15:18:15 BTW experience suggests the density of backport-suitable patches is highest *early* in the dev cycle 15:18:18 i.e. now-ish 15:18:25 (for obvious reasons ...) 15:18:45 BTW here's what we've got so far for 1st stable/havana cut ... https://launchpad.net/ceilometer/+milestone/2013.2.1 15:19:03 * eglynn puts away his soap-box ;) 15:19:45 i'd recommend this as well: https://review.openstack.org/#/c/52638/ 15:19:53 kind of just sitting there. 15:19:56 :) 15:20:02 guys, let's discuss https://review.openstack.org/#/c/57883 15:20:13 gordc: that's testing only, right? 15:20:26 gordc: (so not directly user-visible ...) 15:20:40 eglynn: yeah... is that not an appropriate fix? (i didn't read the wiki link.lol) 15:21:02 jd__, we investigated api-clients today and see that apiclient from oslo is used only in cinder 15:21:02 gordc: well, usually user visibility is a requirement 15:21:18 gordc: ... but if it helps other patches gate on stable 15:21:30 gordc: ... then would be still OK IMO 15:21:48 nadya_: the api client in oslo is a work in progress 15:21:53 jd__, other clients use 'requests' lib for proxy support 15:22:09 nadya_: that's another plan, what shall we do dhellmann? 15:22:17 eglynn: i think it was to allow tests to be run outside +000 15:22:32 i don't think we can run stable/havana tests in north america. 15:23:05 gordc: that's a good enough motivation methinks 15:23:14 gordc: I'll review it after meeting 15:23:17 jd__: I'm not certain of the best path forward here 15:23:30 eglynn: cool cool. thanks. 15:23:31 there is some other work going on to try to create a common client library 15:23:38 it's not clear which version is going to win :-) 15:23:54 so we should probably do the most expedient thing to fix the client for now, and then keep an eye on that other work 15:23:59 "sit tight and buy popcorn" 15:24:13 dhellmann: fair enough 15:24:23 what's our conclusion about the pagination support in ceilometer? Should we continue with our current patches in gerrit or what? 15:24:25 eglynn, gordc : I just posted the question on the changeset 15:24:40 jd__, dhellmann maybe we may have proxy support as we suggested :)? and remove it after client in oslo is done? 15:25:14 dhellmann: ok. i'll answer there. 15:25:47 nadya_: that seems reasonable to me 15:26:01 hello 15:26:08 I want to create plugin for metering statistics of physical switches via OpenDayLight. 15:26:15 nadya_: yep 15:26:27 nadya_, jd__ : this looks interesting: https://pypi.python.org/pypi/slumber/0.6.0 15:26:43 yfujioka: how are those metric exposed? 15:26:43 dhellmann: indeed 15:26:45 Should plugin works in ceilometer-agent-central? 15:26:58 yfujioka: as AMQP notifications or via an API? 15:27:22 yfujioka: the central agent is for polling APIs 15:27:42 yfujioka: we've a new notifications-agent for consuming AMQP notifications 15:27:48 eglynn:I will via OpenDayLight collector API 15:28:06 btw I don't think I'll be available for the meeting next week as I will likely be eating frogs and snails with dhellmann, if anyone wants to chair we can keep it, otherwise I'll cancel it 15:28:09 yfujioka: is that a public REST API? 15:28:14 yes 15:28:45 jd__: we can decide next week if we need it depending on agenda items, I can chair it if we do ... 15:28:56 yfujioka: the central agent is the place so 15:29:06 eglynn: https://jenkins.opendaylight.org/controller/job/controller-merge/lastSuccessfulBuild/artifact/opendaylight/northbound/statistics/target/site/wsdocs/resource_StatisticsNorthbound.html 15:29:07 yfujioka: that seems to fit in central agent to me 15:29:13 eglynn: works for me! I'll leave it up to you then, thanks! :) 15:29:17 np! 15:31:12 sandywalsh, I saw your email about performance. I'm interested in this too. Now I'm getting familiar with multi-node deployment. will come to your etherpads soon 15:31:39 jd__: what's our conclusion about the pagination support in ceilometer? 15:31:41 thanks, I see > eglynn, llu-laptop 15:32:31 by the way, I'm almost installed ceilometer to the big lab. Will start to load it tomorrow. Hope I will get interesting results 15:32:38 I think I must create BP for the plugin. 15:33:12 llu-laptop: not sure, I think we should go for whatever makes sense 15:33:23 I have found template of BP. https://wiki.openstack.org/wiki/Ceilometer/BlueprintTemplate 15:33:30 Is this in available? 15:34:25 also PA: I stopped sending meeting agendas on openstack-dev@l.os.o as it seems unnecessary and the list is already really busy 15:35:05 sensible move methinks, that ML is like a firehose already ... 15:35:20 jd__: how about we go on with the current pagination patch? I don't see any harm. 15:35:32 llu-laptop: I didn't review them 15:35:57 I think Fengqian needs to rebase it 15:36:35 I mean we should continue to review the pagination patch. 15:36:42 agreed 15:37:09 yeap, since there's a bunch of storage driver code landed already 15:37:15 (related to pagination) 15:37:24 IIUC 15:37:43 eglynn: you're right. part of them are already there 15:39:50 I think we're done 15:39:57 thanks guys! 15:40:03 and join #openstack-ceilometer 15:40:04 yep, thanks all! 15:40:05 laters 15:40:10 #endmeeting