15:01:01 <gordc> #startmeeting telemetry
15:01:01 <openstack> Meeting started Thu Dec 10 15:01:01 2015 UTC and is due to finish in 60 minutes.  The chair is gordc. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:01:02 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:01:05 <openstack> The meeting name has been set to 'telemetry'
15:01:06 <sileht> o/
15:01:29 <cdent> o/
15:02:10 <gordc> hmm... maybe we can just give up right away \o/
15:02:25 <pradk> o/
15:02:29 <ildikov> o/
15:02:56 <gordc> let's just assume this is good enough
15:03:09 <ildikov> gordc: :)
15:03:14 <gordc> #topic recurring: roadmap items (new/old/blockers) https://wiki.openstack.org/wiki/Telemetry/RoadMap
15:03:28 <gordc> i have no updates.
15:03:36 <gordc> just review the specs currently up
15:03:51 <gordc> cdent: re: wsme spec, i assume you won't actually be implementing that
15:03:59 <gordc> given you're moving to nova
15:04:18 <cdent> I reckon that's a fair assessment
15:04:29 <cdent> because I'm a horrible person
15:04:53 <gordc> cdent: kk. well i guess we can approve it if we get a resource somewhere.
15:05:23 <cdent> I think the grenade resource.sh one is more important
15:05:34 <gordc> sileht: i took keystoneauth stuff off roadmap since you're already working on it (and merging it)
15:05:44 <gordc> cdent: yeah... ...
15:05:46 <cdent> as that allows testing upgrades and multi-node rolling upgrades
15:06:01 <sileht> gordc, that's was required to use gnocchiclient :p
15:06:11 <gordc> cdent: what's the state of multi-node testing?
15:06:52 <cdent> it's operational but not fully stable yet
15:07:06 <cdent> but is a very high priority for infra/qa/nova folk
15:07:45 <gordc> cdent: kk
15:08:03 <gordc> cdent: yeah i was probably targetting the multi-node testing for n*
15:08:19 <gordc> but mainly because my mind goes numb testing
15:08:59 <cdent> that explains a lot about my brain
15:09:14 <gordc> that's the drugs.
15:09:33 <gordc> anyone else for mitaka bp?
15:09:34 * cdent channels eric stolz from pulp fiction
15:10:19 <gordc> anyone have extra resources they can bring in?
15:10:28 <gordc> or should we just keep giving everything to sileht ?
15:10:37 <cdent> sileht++
15:10:45 <gordc> awesome!
15:11:07 <gordc> #topic aodh
15:11:28 <gordc> r-mibu: i saw your patches for tempest
15:11:31 <sileht> no problem, but someone can give me a potion to not sleep ?
15:11:44 <r-mibu> yep, tempest plugin is ready for review
15:11:50 <gordc> sileht: those are fake. i think they make you go crazy.
15:12:09 <gordc> r-mibu: awesome stuff. are we tracking anything else in aodh?
15:12:24 <gordc> i might take a look at client later but only if we can't find resources
15:12:24 <sileht> gordc, gnocchilient and keystoneauth1 too
15:12:32 <gordc> sileht: ah right
15:12:41 <sileht> same work as I have done in ceilometer
15:12:44 <gordc> sileht: did you manage to get it working and happy with heat?
15:13:06 <sileht> gordc, I have fixes the policy issue https://review.openstack.org/#/c/254078/
15:13:15 <sileht> gordc, but now heat fail into an other error
15:13:25 <gordc> :(
15:13:48 <gordc> i wonder why out integration gate with heat works but heats functional test with aodh don't?
15:14:14 <sileht> me too :(
15:14:17 <gordc> s/out/our/
15:14:38 <r-mibu> could you tell the job name for aodh?
15:14:52 <sileht> gordc, I will continue to investigate on heat issue
15:14:53 <gordc> r-mibu: in heat?
15:15:00 <r-mibu> yes, in heat
15:15:31 <gordc> r-mibu: it's something like CeilometerTest
15:15:40 <gordc> i know it does't say Aodh
15:16:07 <gordc> r-mibu: i think it's this https://github.com/openstack/heat/blob/864181e522efefd96f367e3174a2a0e74123532e/heat_integrationtests/scenario/test_ceilometer_alarm.py
15:16:22 <r-mibu> thanks!
15:17:06 <gordc> let's change to ceilometer i guess
15:17:14 <gordc> #topic ceilometer
15:17:54 <gordc> i think we have everything covered here?
15:18:07 <gordc> except for polling
15:18:42 <gordc> anyone have issues to bring up with ceilometer?
15:19:35 <gordc> let's assume no.
15:19:38 <gordc> #topic gnocchi
15:19:58 <gordc> jd__: sileht:?
15:20:27 <sileht> nothing fancy for me, just some bugfixe/doc
15:20:47 <gordc> jd__:  is the next item you're working on splitting timeseries files into smaller chunks?
15:22:04 <gordc> i assume yes. and that he is working on solving world peace
15:22:24 <gordc> sileht: i think it'd be good if we start periodically running performance tests
15:22:45 <gordc> just to verify it... not sure if you have a test plan for that.
15:22:57 <gordc> i know we have a few benchmark methods.
15:23:14 <sileht> We have a benchmark tools in the clients
15:23:34 <sileht> it have some limitation if your hardware is too big, but work most of the time
15:23:52 <gordc> sileht: cool cool. maybe once we have servers setup we can start doing that.
15:24:05 <gordc> llu: any updates on the intel environment?
15:24:35 <sileht> gordc, https://julien.danjou.info/blog/2015/gnocchi-benchmarks
15:25:27 <gordc> sileht: yep. i think we should do that regularly (maybe after we merge a (few) feature)?
15:26:33 <sileht> gordc, I don't think still have the hardware where we did this but yeah that would be cool to recheck regulary
15:26:52 <llu> gordc: sorry, not much. Stucked in the training for more time than I expected
15:27:09 <llu> the machines in intel environment are all ready
15:27:16 <gordc> sileht: i might be able to get hardware next year. i have hardware but i'm waiting on lab to be built.
15:27:30 <sileht> cool
15:27:33 <gordc> llu: kk. so maybe we'll start next year?
15:27:34 <llu> but the process of IT approving to give us access right out side US is lengthy
15:27:54 * cdent sighs
15:28:18 <gordc> cdent: embrace the enterprise.
15:28:20 <llu> the initial draft test plan is at https://docs.google.com/document/d/17zx7tcwhcW5gwa1gkmX3-GnPjjnUPbGFHec9x7u1gO8/edit
15:28:33 <llu> welcome to review and give comment/suggestions
15:28:55 <gordc> llu: cool stuff. will review maybe tomorrow
15:29:24 <llu> just FYI
15:29:49 <llu> we have 2 more people from OSIC may help about this
15:30:12 <gordc> osic?
15:30:35 <llu> openstack innovation center, intel/rackspace
15:30:58 <llu> stevelle from Racksapce(ansible core) is adding gnocchi support in ansible
15:31:29 <gordc> llu: oh there's a fancy name for intel/rax cluster. :)
15:31:34 <llu> and sumant murke from Intel will working on gnocchi/ceilometer integration things
15:31:49 <gordc> llu: that's awesome to hear.
15:32:07 <llu> I believe sumant just published his first patch to gnocchi
15:32:15 <gordc> look forward to the feedback so we can draft a good doc on how to configure
15:33:06 <gordc> llu: that's really good to hear.
15:33:21 <gordc> any other items?
15:33:33 <cdent> related to gnocchi: people hear about influxdb's rebranding and new suite of tools?
15:34:40 <cdent> influxdata™
15:34:57 <cdent> and some monitoring toolset and graphing stuff to go before and after it
15:35:12 <gordc> does it work?
15:35:29 <cdent> based on the past, I'd guess no
15:35:42 <gordc> seems like they're going the elastic route. building the full stack.
15:36:01 <cdent> they call it tick
15:36:15 <cdent> and appear to be really running for the finance markets
15:36:19 <cdent> new data storage model
15:36:21 <cdent> too
15:36:32 <gordc> but finance can afford splunk...
15:36:38 <gordc> *shrugs*
15:36:58 <cdent> just pointing it out, not recommending it or anything
15:37:04 <gordc> i think jd__ tried to do some influx work on gnocchi driver recently but ran into more issues
15:37:14 <jd__> cdent: didn't hear, but yeah whatever :)
15:37:22 <gordc> jd__: sell gnocchi to finance
15:37:32 <jd__> haha
15:37:36 <cdent> jd__: exactly
15:37:49 <gordc> i'll email morgan stanley... ceo@morganstanley.com? that work?
15:38:23 <gordc> jd__: i assume you still have issues with influx driver?
15:38:49 <jd__> it seems to work but it does not pass the test suite anymore (3 tests fail)
15:38:54 <jd__> it should be pretty easy to fix them I guess
15:39:08 <jd__> but when I tried, I got frustrated by InfluxDB giving me weird result to simple query
15:39:26 <jd__> and having the first examples in the doc being syntaxicaly invalid just made me giving up
15:39:36 <gordc> jd__: kk. well if anyone is interested in that driver, those patches are open to look at it
15:39:41 <jd__> exactly
15:40:10 <gordc> cool let's change
15:40:14 <gordc> #topic open discussion
15:40:52 <gordc> oh. i guess we should talk about whether anyone will show up next week
15:40:55 <gordc> do we want a meeting?
15:41:14 <cdent> next week is pretty normal, it is the two weeks following that are not?
15:41:35 <gordc> cdent: yeah, i'm around but not sure if anyone has holiday plans
15:42:25 <gordc> right now it's still on. if i'm the only one there, y'all owe me drinks.
15:42:42 <jd__> i'm on
15:42:51 <ildikov> me too :)
15:42:57 * cdent adds drinks to his donut shopping list
15:43:20 <gordc> cool cool. any last items?
15:44:05 <gordc> let's call it
15:44:07 <gordc> thanks folks.
15:44:12 <gordc> #endmeeting