15:00:16 <ildikov> #startmeeting Ceilometer
15:00:20 <openstack> Meeting started Thu Jul 23 15:00:16 2015 UTC and is due to finish in 60 minutes.  The chair is ildikov. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:21 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:24 <openstack> The meeting name has been set to 'ceilometer'
15:01:43 <prad> o/
15:02:07 <prad> is gordc out today?
15:02:23 <ildikov> prad: yes
15:02:44 <fabiog> hi
15:02:51 <ildikov> prad: he asked a few guys to chair the meeting, I volunteered finally :)
15:04:14 <prad> cool
15:05:55 <ildikov> I asked around on the channel, I will directly ping a few folks, if we will not have more attendees then we just cancel this one
15:07:12 <cdent> i'm happy to skip it if nobody shows up
15:07:42 <ildikov> cdent: we have prad, fabiog and you and me so far
15:07:54 <janonymous> o/
15:08:11 <ityaptin> hi!
15:08:52 <jasonamyers> o/
15:08:58 <ildikov> cdent: I'm not against skipping the today's one, if no one has anything to bring up
15:09:06 <jd__> o/
15:09:26 <jasonamyers> I only wanted to share that I pushed up the first part of my bp, and testing will be up next week
15:10:16 <ildikov> ok, let's run through the topics and then we can close early
15:10:31 <ildikov> #topic recurring: roadmap items (new/old/blockers)
15:10:31 <ildikov> #link https://wiki.openstack.org/wiki/Ceilometer/RoadMa
15:11:04 <ildikov> we have now support in the ceilometerclient for Aodh thanks to liusheng
15:11:54 <ildikov> is there any other update from anyone regarding to our roadmap?
15:13:09 <ildikov> if not, then let's move on
15:13:36 <ildikov> #topic recurring: ceilometerclient release?
15:13:52 <ildikov> anything else from anyone regarding to the client, then what I mentioned already?
15:14:26 <cdent> nothing from me
15:14:49 <ildikov> I guess we can wait a bit with releasing as we keep the alarming code in our repo for Liberty, so it's no rush
15:15:20 <janonymous> I have a small ques relating to py3 compat of ceilometer.. is it fullt compatible
15:15:24 <janonymous> *fully
15:15:45 <ildikov> janonymous: we will get to the open discussion in a minute
15:15:55 <janonymous> ohkay..
15:15:55 <ildikov> janonymous: let's get back to this question there
15:16:11 <ildikov> #topic recurring: Gnocchi status
15:16:24 <ildikov> cdent: jd__: anything to mention regarding Gnocchi?
15:16:50 <cdent> lots of progress in the last week or two on getting it running properly in devstack setups
15:16:54 <cdent> found some bugs
15:16:55 <cdent> fix them
15:16:57 <cdent> found some more
15:17:00 <cdent> fixed them too
15:17:13 <cdent> open question on the state of influxdb work
15:17:15 <ityaptin> Did anybody try influxdb driver?
15:17:44 <fabiog> what version of InfluxDB is using?
15:17:52 <ityaptin> Now it doesn't support search_value and crossmetric measures
15:18:13 <ityaptin> inlfuxdb 0.9.1
15:18:27 <fabiog> ityaptin: thanks
15:18:37 <prad> fyi i packaged up python-influxdb and submitted to fedora
15:18:50 <jd__> sorry got side tracked but what cdent said
15:19:15 <ildikov> ityaptin: these two things are not supported by our driver or it's on Influx side?
15:19:43 <ildikov> prad: cool, thanks for the info
15:20:16 <ityaptin> in our driver, I haven't researched possibility to support it yet, but think that is not so hard.
15:20:54 <ildikov> ityaptin: ok, cool
15:21:18 <ildikov> ityaptin: is there anything else you consider as missing items or any issues identified so far?
15:21:45 <ityaptin> Also, now we don't use metricd for InfluxDb. We only make aggregate queries through raw data. Maybe we can use Continuous Queries for it.
15:23:59 <ityaptin> ildikov: We have some issues with GroupBy queries in InfluxDB and don't have information about HA InfluxDB. Unfortunately at last two weeks I had no time to InfluxDB research or benchmarks.
15:24:35 <ildikov> ityaptin: I see
15:25:00 <ildikov> ityaptin: How far is the driver from a viable state?
15:25:38 <ildikov> ityaptin: I mean when in functionality and performance it can be considered as a usabale one
15:26:11 <ildikov> ityaptin: also does this issue require some eyes from InfluxDB folks?
15:26:57 <ityaptin> In a first approximation basic functionality looks good, but we need to test it  for argumentative answer.
15:27:30 <ildikov> ityaptin: cool, sounds good
15:27:35 <ityaptin> yep, I think we can suggest them our thoughts
15:29:27 <ildikov> ityaptin: great, let's aim for preparing a list of issues/questions and then we can reach out to them for solution
15:29:38 <ityaptin> ildikov: ok, good
15:30:09 <ildikov> ityaptin: anything else on this topic?
15:31:21 <ityaptin> Nothing from my side
15:31:30 <ildikov> ityaptin: cool, tnx
15:31:37 <ildikov> anything else on Gnocchi?
15:32:36 <ildikov> cdent: ityaptin: jd__: thanks for the update
15:32:51 <ildikov> let's move on then
15:32:53 <ildikov> #topic Open discussion
15:33:09 <jd__> is the influxdb driver mergeable as it is?
15:33:14 <jd__> I didn't test it and I remember I said I would
15:33:20 * jd__ slaps jd
15:33:36 <jd__> though it's failling devstack right now
15:33:51 <ityaptin> jd__: Gate tests are still failing :(
15:34:12 <jd__> ityaptin: and nobody was able to solve that?
15:34:47 <ityaptin> unfortunately, last two weeks I didn't try :(
15:35:05 <rjaiswal> Request some eyes on https://blueprints.launchpad.net/ceilometer/+spec/dynamic-event-pipeline and https://bugs.launchpad.net/ceilometer/+bug/1477321
15:35:07 <openstack> Launchpad bug 1477321 in Ceilometer "Enabling refresh_pipeline_cfg leads to restart of event listener even when sample_pipeline is refreshed" [Undecided,New] - Assigned to Rohit Jaiswal (rohit-jaiswal-3)
15:35:07 <uvirtbot> Launchpad bug 1477321 in ceilometer "Enabling refresh_pipeline_cfg leads to restart of event listener even when sample_pipeline is refreshed" [Undecided,New]
15:35:08 <uvirtbot> Launchpad bug 1477321 in ceilometer "Enabling refresh_pipeline_cfg leads to restart of event listener even when sample_pipeline is refreshed" [Undecided,New] https://launchpad.net/bugs/1477321
15:36:44 <janonymous> Yeah, open-discussions :   I have a small ques relating to py3 compat of ceilometer.. is it fullt compatible ?
15:36:55 <cdent> janonymous: not yet
15:36:56 <ityaptin> rjaiswal: I will. And some question - Do anyone plans to make reloadable event/meter definition?
15:37:01 <cdent> getting closer though
15:37:20 <janonymous> Hahaa.. okay i'll wait :) ping me then..
15:38:29 <ityaptin> Folks, how often we want to make ceilometerclient release?
15:38:34 <ildikov> janonymous: you can join and help out in that activity :)
15:38:35 <rjaiswal> Ityapt
15:39:18 <ildikov> rjaiswal: the event_pipeline bp would use all the same mechanisms that you introduced already?
15:39:34 <rjaiswal> ityaptin: sample pipeline is reloadable, no is for making event pipeline reloadable
15:39:41 <rjaiswal> ildikov: yes
15:40:21 <rjaiswal> Don't have a spec detailed out, but it should use the same mechanisms
15:40:21 <ildikov> rjaiswal: then I guess we will not necessarily need a spec for it or do you hae that too?
15:40:39 <rjaiswal> that's a question I had
15:40:49 <rjaiswal> do we need a spec or a bug raised?
15:41:20 <janonymous> ildikov: you are referring to py3.. ryt..
15:42:03 <ildikov> rjaiswal: personally I'm fine with a bug that this dynamic reload is not supported for event_pipeline
15:42:12 <ildikov> janonymous: yeap
15:43:00 <janonymous> ildikov:  Cool, wanted to know it's status.. we'll discuss after your discussions..
15:43:09 <ildikov> jasonamyers: did you want to talk more about your BP?
15:43:22 <rjaiswal> ildikov: thanks
15:43:49 <janonymous> ildikov:  my BP ?
15:44:54 <ildikov> janonymous: no, unless you don't have one you would like to talk about :)
15:45:01 <ildikov> rjaiswal: np
15:45:09 <jasonamyers> ildikov: really just wanted to share that progress is being made
15:45:29 <jasonamyers> I have tests left to write, and some performance testing to attempt
15:45:43 <jasonamyers> working on this next week
15:45:51 <ildikov> jasonamyers: cool, sounds good
15:46:28 <janonymous> Ohh i got confused : i thought you misspelled me as jasonamyers :P
15:46:54 <ildikov> rjaiswal: as for the bug, is there anything to discuss or it was more a heads up?
15:47:22 <ildikov> janonymous: no, this time I pinged, who I planned :)
15:47:23 <rjaiswal> ildikov: was hoping it be triaged
15:47:34 <ityaptin> rjaiswal, ildikov: Sorry for misunderstanding, I told only about event_definitions.yaml. Was anybody willing to make events_definition.yaml reloading?
15:48:27 <ildikov> ityaptin: hmm, I'm not sure how much it will be changed on the fly
15:49:43 <ildikov> ityaptin: on the other hand maybe it would make sense to have a separate file for custom events to avoid that someone accidentally messing up the ones that are already there for OpenStack services
15:49:55 <ildikov> rjaiswal: ok, thanks for clarifying
15:50:28 <ityaptin> ildikov: good idea! :)
15:51:05 <ildikov> ityaptin: :)
15:52:03 <ildikov> ok, anything else from anyone?
15:52:23 * cdent nopes
15:53:08 <ildikov> ok, then let's close, have a nice day/evening for everyone
15:53:18 * cdent waves
15:53:21 <ityaptin> ildikov: thanks!
15:53:25 <ildikov> #endmeeting