Monday, 2015-09-21

*** lawrancejing has joined #openstack-ceilometer00:28
*** lawrancejing has quit IRC00:36
*** lvdongbing has joined #openstack-ceilometer00:39
*** ankita_wagh has joined #openstack-ceilometer01:02
*** summer has joined #openstack-ceilometer01:04
*** jwcroppe has joined #openstack-ceilometer01:09
*** nijaba has joined #openstack-ceilometer01:10
*** nijaba_ has quit IRC01:11
*** ddieterl_ has joined #openstack-ceilometer01:12
*** ddieterly has quit IRC01:15
*** ddieterl_ has quit IRC01:34
*** jwcroppe has quit IRC01:56
*** ldb_ has joined #openstack-ceilometer02:05
*** lvdongbing has quit IRC02:06
*** lvdongbing has joined #openstack-ceilometer02:29
*** ldb_ has quit IRC02:29
*** ankita_wagh has quit IRC02:38
*** ddieterly has joined #openstack-ceilometer02:46
*** ildikov has joined #openstack-ceilometer03:28
*** ankita_wagh has joined #openstack-ceilometer03:42
*** ddieterly has quit IRC03:52
*** ankita_wagh has quit IRC04:04
*** fawadkhaliq has joined #openstack-ceilometer04:20
*** ddieterly has joined #openstack-ceilometer04:53
*** ddieterly has quit IRC04:58
*** deepthi has joined #openstack-ceilometer05:04
*** deepthi has quit IRC05:29
*** deepthi has joined #openstack-ceilometer05:30
*** ddieterly has joined #openstack-ceilometer05:54
*** ankita_wagh has joined #openstack-ceilometer05:56
*** ddieterly has quit IRC05:59
*** nadya has joined #openstack-ceilometer06:09
*** lsmola has joined #openstack-ceilometer06:13
*** fawadkhaliq has quit IRC06:14
*** fawadkhaliq has joined #openstack-ceilometer06:15
*** fawadkhaliq has quit IRC06:19
*** shardy has joined #openstack-ceilometer06:47
*** ildikov has quit IRC06:54
*** ddieterly has joined #openstack-ceilometer06:55
*** ddieterly has quit IRC07:00
*** gzhai has joined #openstack-ceilometer07:05
*** fawadkhaliq has joined #openstack-ceilometer07:08
*** gzhai has quit IRC07:14
*** gzhai has joined #openstack-ceilometer07:16
*** shardy_ has joined #openstack-ceilometer07:21
*** shardy has quit IRC07:22
*** shardy_ has quit IRC07:27
*** shardy has joined #openstack-ceilometer07:27
*** ankita_wagh has quit IRC07:30
*** ildikov has joined #openstack-ceilometer07:40
*** Ala has quit IRC07:41
*** lawrancejing has joined #openstack-ceilometer07:42
*** Ala has joined #openstack-ceilometer07:43
*** yprokule has joined #openstack-ceilometer07:46
*** safchain has joined #openstack-ceilometer07:53
*** ddieterly has joined #openstack-ceilometer07:56
*** ddieterly has quit IRC08:00
*** ala_ has joined #openstack-ceilometer08:02
*** Ala has quit IRC08:04
*** shardy_ has joined #openstack-ceilometer08:15
*** shardy has quit IRC08:16
*** yassine_ has joined #openstack-ceilometer08:17
*** shardy_ has quit IRC08:20
*** shardy has joined #openstack-ceilometer08:21
openstackgerritMehdi Abaakouk (sileht) proposed openstack/python-gnocchiclient: Fix help typo
*** fawadkhaliq has quit IRC08:26
*** gzhai has quit IRC08:28
openstackgerritJulien Danjou proposed openstack/gnocchi: devstack: install gnocchiclient
*** nadya has quit IRC08:43
*** alexpilotti has joined #openstack-ceilometer08:45
openstackgerritMerged openstack/python-gnocchiclient: Fix help typo
*** boris-42 has quit IRC08:53
*** smekel has quit IRC08:53
*** patrickeast has quit IRC08:53
*** rex_lee_ has quit IRC08:53
*** ddieterly has joined #openstack-ceilometer08:57
*** shardy_ has joined #openstack-ceilometer08:57
*** shardy has quit IRC08:58
*** ddieterly has quit IRC09:01
*** shardy_ has quit IRC09:02
*** shardy has joined #openstack-ceilometer09:03
*** shardy has quit IRC09:05
*** shardy has joined #openstack-ceilometer09:08
*** lawrancejing has quit IRC09:12
*** eglynn has joined #openstack-ceilometer09:18
*** lawrancejing has joined #openstack-ceilometer09:20
*** boris-42 has joined #openstack-ceilometer09:24
*** IBerezovskiy has joined #openstack-ceilometer09:40
*** eglynn has quit IRC09:40
*** eglynn has joined #openstack-ceilometer09:40
*** rex_lee_ has joined #openstack-ceilometer09:42
*** lawrancejing has quit IRC09:43
*** lvdongbing has quit IRC09:48
*** smekel has joined #openstack-ceilometer09:50
*** lawrancejing has joined #openstack-ceilometer09:51
*** patrickeast has joined #openstack-ceilometer09:53
*** ddieterly has joined #openstack-ceilometer09:58
*** yassine_ has quit IRC10:01
*** ddieterly has quit IRC10:02
*** fawadkhaliq has joined #openstack-ceilometer10:02
*** eglynn_ has joined #openstack-ceilometer10:02
*** eglynn has quit IRC10:04
*** nadya has joined #openstack-ceilometer10:05
*** ala_ has quit IRC10:11
*** summer has quit IRC10:12
*** lawrancejing has quit IRC10:33
openstackgerritMehdi Abaakouk (sileht) proposed openstack/python-gnocchiclient: Replace keystoneclient by keystoneauth1
openstackgerritMehdi Abaakouk (sileht) proposed openstack/python-gnocchiclient: Remove now useless build_url method
*** cmyster has joined #openstack-ceilometer10:51
cmystereglynn_: heya, there ?10:51
eglynn_cmyster: 'sip?10:53
eglynn_sup even10:53
cmysterall well, need to ask you something re a bug10:53
cmysterif you have the time10:53
*** ddieterly has joined #openstack-ceilometer10:58
*** ddieterly has quit IRC11:03
openstackgerritMehdi Abaakouk (sileht) proposed openstack/python-gnocchiclient: Remove now useless build_url method
openstackgerritMehdi Abaakouk (sileht) proposed openstack/python-gnocchiclient: Replace keystoneclient by keystoneauth1
openstackgerritMehdi Abaakouk (sileht) proposed openstack/python-gnocchiclient: Don't register an auth plugin for help
*** yassine_ has joined #openstack-ceilometer11:35
*** gordc has joined #openstack-ceilometer11:41
gordchey folks, any other items to target against rc1? we have 3 remaining:
openstackgerritAvi Weit proposed openstack/ceilometer: Change to cpu_util metric for kvm overcommited
silehtgordc, nothing for me11:56
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: Factorize field definition of declarative code
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: Factorize yaml loading of declarative stuffs
*** ddieterly has joined #openstack-ceilometer11:59
gordcsileht: cool cool12:02
*** ddieterly has quit IRC12:03
*** thorst has joined #openstack-ceilometer12:04
*** ildikov has quit IRC12:12
*** lawrancejing has joined #openstack-ceilometer12:12
gordcsileht: if you can provide feedback on remaining 3 rc1 items, that'd be great.12:12
*** ildikov has joined #openstack-ceilometer12:25
*** _gryf has quit IRC12:26
*** edmondsw has joined #openstack-ceilometer12:31
*** ddieterly has joined #openstack-ceilometer12:43
gordcany additional comments to operators survey?
openstackgerritAvi Weit proposed openstack/ceilometer: Change to cpu_util metric for kvm overcommited
*** Ala has joined #openstack-ceilometer12:50
*** fxpester has joined #openstack-ceilometer12:51
*** deepthi has quit IRC12:53
*** shardy_ has joined #openstack-ceilometer12:54
*** shardy has quit IRC12:55
*** lawrancejing has quit IRC12:56
*** shardy_ has quit IRC12:59
*** shardy has joined #openstack-ceilometer13:00
*** ildikov has quit IRC13:03
*** ildikov has joined #openstack-ceilometer13:04
*** jaypipes has joined #openstack-ceilometer13:05
*** ddieterly has quit IRC13:05
jd__sileht: aren't we supposed to have a gate against a real gnocchi on gnocchiclient?13:07
*** exploreshaifali has joined #openstack-ceilometer13:07
silehtjd__, we have13:07
silehtjd__, that the normal pyXX jobs that run everything13:08
jd__sileht: oh, awesome.13:08
*** renatoarmani has joined #openstack-ceilometer13:08
silehtjd__, no slow dsvm job :p13:08
jd__yeah that's pretty cool indeed13:08
*** julim has joined #openstack-ceilometer13:09
*** renatoarmani has quit IRC13:09
*** fawadkhaliq has quit IRC13:10
*** pradk has joined #openstack-ceilometer13:19
*** fawadkhaliq has joined #openstack-ceilometer13:19
*** jaypipes has quit IRC13:22
*** avi has joined #openstack-ceilometer13:23
gordcildikov: any idea how to test this?
*** avi is now known as Guest4360013:23
gordcit's probably the main item we need in for rc1 since it's a regression from kilo13:24
ildikovgordc: hmm, I guess if we Mock out the half world and then simulate the rest then it could be tested13:25
gordcwhich half?13:26
gordchm... let me try building a functional test13:28
ildikovI meant the half where we get the data from13:28
ildikovdo we have tests for transformers?13:28
*** dan-t has joined #openstack-ceilometer13:29
gordcildikov: yeah... it's tested... i'll try playing with the current test13:30
ildikovhmm, ok, cool, I just couldn't find them in the new structure for the first very quick look :)13:31
*** ddieterly has joined #openstack-ceilometer13:32
*** Guest43600 is now known as aviw13:33
*** ddieterly has quit IRC13:33
*** ddieterly has joined #openstack-ceilometer13:33
*** ddieterly has quit IRC13:36
ildikovgordc: thanks, it's definitely not the file where I would think they are13:37
gordcildikov: yeah... i'm not sure why the file is not under ceilometer/tests/unit... it's in some weird middlezone13:38
ildikovgordc: yeap, it looks like some helper file13:38
*** renatoarmani has joined #openstack-ceilometer13:38
ildikovI checked the unit/functional folders and then got confused13:39
ildikovwe can fix it later though13:39
openstackgerritMehdi Abaakouk (sileht) proposed openstack/python-gnocchiclient: Don't always load client
gordcildikov: agreed13:39
jd__sileht: I like cliff idea but it really seems slow13:41
jd__sileht: I think it's an issue with entry points, no?13:41
*** jaypipes has joined #openstack-ceilometer13:42
silehtjd__, we can remove entrypoints lookup13:42
* sileht will propose a change for that13:43
openstackgerritMerged openstack/ceilometer: retain existing listeners on refresh
jd__sileht: would be cool yeah, I don't think we need to use it and it's just too slow13:43
*** rbak has joined #openstack-ceilometer13:43
jd__sileht: what's the CLI option to change the auth option?13:43
jd__I don't see it in --help13:43
jd__--os-auth-type ?13:43
silehtjd__, yes13:44
jd__would be cool to have that as "auth-type"13:44
jd__not a big deal13:44
silehtjd__, that comes from keystoneauth113:45
openstackgerritMerged openstack/ceilometer: tox: generate config file on test run
jd__sileht: I guessed so13:45
*** r-daneel has joined #openstack-ceilometer13:46
*** rbak has quit IRC13:47
*** fawadkhaliq has quit IRC13:51
*** rbak has joined #openstack-ceilometer13:51
openstackgerritgordon chung proposed openstack/ceilometer: remove batch processing requirement from arithmetic transformer
jd__facepalm devstack keystone13:54
openstackgerritMerged openstack/python-gnocchiclient: Replace keystoneclient by keystoneauth1
openstackgerritMerged openstack/python-gnocchiclient: Remove now useless build_url method
jd__can you guess what's wrong in :]13:57
*** david-ly_ is now known as david-lyle14:01
gordcjd__: i leave the complicated stuff for you to figure out14:04
openstackgerritMehdi Abaakouk (sileht) proposed openstack/python-gnocchiclient: Don't always load client
silehtjd__, I have removed the loading entrypong from GnocchiShell, but the keystoneauth1 lib use stevedore too14:12
openstackgerritGuangyu Suo proposed openstack/ceilometer: Fix reconnecting to libvirt
*** lawrancejing has joined #openstack-ceilometer14:17
gordcsileht: i'm seeing a lot of errors in central agent logs (when i don't have neutron enabled). don't know if you have an idea how to fix it14:17
openstackLaunchpad bug 1498019 in Ceilometer "EndpointNotFound errors thrown in polling agent" [High,Triaged]14:17
gordcif not,  i can start debugging.14:17
silehtgordc, I take a look14:18
jd__sileht: is it still faster?14:21
jd__gordc: :/14:22
gordcjd__: oh... that fixes ^^?14:22
*** david-lyle has quit IRC14:22
openstackgerritMehdi Abaakouk (sileht) proposed openstack/python-gnocchiclient: Remove entry_points
silehtjd__, you can try, but I don't think so
jd__gordc: no just what I pointed earlier14:23
gordcjd__: ah i see.14:24
openstackgerritMonty Taylor proposed openstack/aodh: Change ignore-errors to ignore_errors
openstackgerritMonty Taylor proposed openstack/ceilometer: Change ignore-errors to ignore_errors
openstackgerritMonty Taylor proposed openstack/ceilometermiddleware: Change ignore-errors to ignore_errors
jd__sileht: it's faster actually: ~0.6s rather than ~0.7s14:25
jd__sileht: so it's a pretty good start14:25
jd__-15 % it's a good improvement to me :)14:26
*** david-lyle has joined #openstack-ceilometer14:26
silehtgordc, the issue seems not related to my change, I have I guess just unhide it14:30
*** aviw has quit IRC14:30
gordcsileht: i see... do you have time to look at it? i can continue debugging it if you have more urgent items.14:31
silehtgordc, I can continue14:31
openstackgerritOpenStack Proposal Bot proposed openstack/ceilometer: Updated from global requirements
gordcsileht: kk thanks very much!14:32
gordcpradk: just an fyi, i updated the operator questionnaire a bit.
pradkgordc, cool14:32
gordcmaybe we can give the list to nijaba tomorrow or EOD today?14:33
nijabagordc: would be perfect14:33
gordcnijaba: cool cool. red hat email fine?14:34
*** sergio_nubeliu has joined #openstack-ceilometer14:36
openstackgerritMerged openstack/python-gnocchiclient: Don't always load client
*** lawrancejing has quit IRC14:38
*** ddieterly has joined #openstack-ceilometer14:55
*** devlaps has joined #openstack-ceilometer15:02
ildikovgordc: quick question15:06
ildikovis the expected value in the test changed because of the evaluation will not happen until we do not have the second memory.usage value?15:07
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: Fix variable typos
silehtgordc, finally this is my fault:
openstackgerritMehdi Abaakouk (sileht) proposed openstack/ceilometer: Don't create neutron client àt loadtime
gordcildikov: it changed because previously it was grouped by polling task. so if for some reason you polled the a meter for a resource twice in a polling task, it would only take the latest15:09
gordcildikov: now it processes value in order, unrelated to polling task. once it finds a match, it will calculate tranform data... so in this case it takes the first and not latest15:10
gordcthe 2nd value would be used for the next match15:10
ildikovyeap, this is how I meant the question, thanks for clarification15:11
ildikovwas this described somewhere?15:11
gordcsileht: ah, good catch.15:11
gordcildikov: yep. let me find you spec.15:11
gordci basically had to change the orignal functionality since transformers have nothing to do with polling task15:12
gordci basically changed from proposed option 1 to proposed option 215:13
ildikovI meant the documentation this time, I mean admin and developer's guide15:13
ildikovbut I haven't found mentioning the polling task luckily so it seems to be good from that perspective15:14
gordcildikov: oh, i don't think it's documented anywhere.15:15
gordci'm surprised it was linked to polling task before.15:15
ildikovthe transformer itself is documented, but it only says the interval should be the same for the meters and that's all15:16
ildikovso it looks how it's supposed to, I just wasn't sure15:16
gordcildikov: right... logically it should be same polling interval.15:18
*** devlaps has quit IRC15:19
sergio_nubeliuHi guys, we are planning to start using gnocchi api for live metrics and statisctics in medium and large environments, is it production ready?15:22
ildikovgordc: has any of the new version tests failed for you without the code change?15:22
gordcildikov: new version tests?15:23
silehtjd__, what do you think of the state of the gnocchi client ?15:24
ildikovgordc: I meant the modifications in the tests15:24
gordcsergio_nubeliu: i'm not sure we have conclusive evidence yet. we are encouraging users to start trying it so we can fix issues that arise. the more testers/feedback we get, the more stable we can make it.15:24
jd__sileht: I stopped testing because I found a bug in Keystone -_-15:25
jd__sileht: but according to --help it looks pretty good…15:25
silehtjd__, I guess we could release it soon15:25
*** yprokule has quit IRC15:25
silehtat least in alpha/beta/whatever for testing the release process15:26
jd__sileht: agreed15:26
silehtdo we start we 1.0.0rc1 ?15:26
jd__sileht: let me play it a bit with it this week and we'll do a 1.0.0 if everything is fine by the end of the week :)15:26
jd__sileht: rc is useful if people test it… nobody is going to test :p15:26
ildikovgordc: I need to drop a bit, be back online in about 20 mins15:26
*** yprokule has joined #openstack-ceilometer15:27
gordcildikov: i'm not sure exactly what you mean. i updated to the tests to perform how the pipeline performs in notification agent... all seems fine from existingn tests15:27
silehtjd__, ok cool15:27
gordcildikov: ack15:27
*** pradk has quit IRC15:27
ildikovgordc: I ran the modified test suite with the old code, I supposed that should've failed15:27
*** devlaps has joined #openstack-ceilometer15:28
*** pradk has joined #openstack-ceilometer15:29
gordcildikov: it should probably fail yes.15:29
dhellmanngordc: I'm going to cut stable branches today, and want to double check that you're ok with me doing that for the latest releases of python-ceilometerclient and ceilometermiddleware15:31
*** ildikov has quit IRC15:32
gordcdhellmann: we should be good to use the existing releases... the only items i see are requirements changes...15:33
dhellmanngordc: sounds good, thanks15:33
gordcdhellmann: do we need to sync them up for a final release? they seem related to pbr15:33
dhellmanngordc: good question. I hope not, at this point.15:34
gordclooks like it's dropping an upper bound... i'm not sure what the other projects have currently -- it's relatively recent change.15:35
*** exploreshaifali has quit IRC15:41
openstackgerritMerged openstack/python-gnocchiclient: Remove entry_points
*** yassine_ has quit IRC15:50
sergio_nubeliugordc: is it possible to use mysql as the indexer backend? why do you recommend postgre? where can I find documentation to configure the collectors to push data to gnocchi?15:51
*** vkmc has joined #openstack-ceilometer15:55
gordcsergio_nubeliu: you can use mysql, i think it's just recommended to use postgres because we have a slight bias towards postgres being better (ie. apple vs microsoft)15:57
gordcsileht: jd__: you have a quick link to how to configure gnocchi? ^15:57
*** nadya has quit IRC15:57
silehtgordc, a documentation or something for a hacker ?15:58
gordcsileht: lol type documentation.16:00
silehtgordc, nothing more than:
vkmchi all! there is someone interested in joining Ceilometer for next Outreachy internships round, and I'm afraid I don't know anyone willing to mentor for Ceilometer this round16:03
gordcsileht: ok.16:03
*** alejandrito has joined #openstack-ceilometer16:03
vkmcI send a call for mentors last week, there is more information about this internship there
gordcsergio_nubeliu: if that helps ^^16:04
vkmcif someone is interested on mentoring or know someone that might be interested, please ping me16:04
gordcsergio_nubeliu: you'll also need something similar to in your ceilometer.conf file (and dispatcher=gnocchi16:05
gordcvkmc: thanks for the heads up. we discussed it last meeting actually but don't have any commitments yet16:05
gordcvkmc: i assume we need to come up with a topic of interest for the person to work on?16:06
gordcor do they have their own idea already?16:06
*** ildikov has joined #openstack-ceilometer16:06
vkmcgordc, thanks for the update, didn't know you discussed this in last weekly meeting :)16:06
gordcvkmc: just a quick discussion regarding your thread.16:06
vkmcgordc, yeah, the mentor would need to propose a task idea for the internship... aligned with the project plans16:07
*** nadya has joined #openstack-ceilometer16:07
vkmccool, thanks16:07
sergio_nubeliugordc: that will help!16:07
*** safchain has quit IRC16:08
alejandritogordc, sergio_nubeliu and this kind of "production ready" integration ... its possible on "Kilo" ?16:09
gordcvkmc: we have until october 27?16:10
gordcalejandrito: we didn't port this integration to kilo. although it probably would be 'relatively' easy.16:11
vkmcgordc, october 27 is the deadline for applicants to submit their applications, so I'd say some days before that :)16:13
gordcvkmc: ok. i'll try polling for ideas. i'm sure we can find someone to mentor if it's just a weekly checkpoint16:13
vkmcgordc, thanks a lot16:14
vkmcgordc, I'll try to get this applicant ready with the basics to contribute to Ceilometer in the meantime16:14
jd__gordc: Gnocchi default config just works in theory, you just have to set the indexer_url and that's it – or use devstack if you can it's faster :)16:14
gordcvkmc: do you know where the candidates are generally based in? is it usually india/china?16:14
vkmcgordc, we have people from everywhere, but this one in particular seems to be from India16:15
vkmcI'll ask just in case16:15
gordcvkmc: ack.16:16
alejandritogordc, what if i WANT to run gnocchi integration on kilo ? what part of the ceilometer code should i "port" ? can you please give me a hint ?16:16
vkmcgordc, confirmed, they are from India16:19
gordcvkmc: great thanks.16:19
*** liusheng has quit IRC16:19
gordcalejandrito: let me find the appropriate files.16:20
*** khushbu_ has joined #openstack-ceilometer16:20
*** liusheng has joined #openstack-ceilometer16:20
alejandritogordc, sergio_nubeliu  thanks so much ... waiting for the instructions16:23
*** Ala has quit IRC16:25
gordcalejandrito: i haven't tried this but this should be the bare minimum:
*** shardy_ has joined #openstack-ceilometer16:26
gordcwhoops ->
*** shardy has quit IRC16:27
gordcyou'll need to make additional changes to configure alarms with gnocchi.16:27
openstackgerritWen Zhi Yu proposed openstack/ceilometer: Remove duplicated pipeline support meter check
*** shardy_ has quit IRC16:31
*** yprokule has quit IRC16:31
gordcpradk: do you have the bugzilla link for
openstackLaunchpad bug 1484579 in Ceilometer "Incomplete documentation/RPM Packaging Error" [Undecided,New] - Assigned to Pradeep Kilambi (pkilambi)16:32
pradkgordc, yea one sec16:32
*** shardy has joined #openstack-ceilometer16:32
*** yprokule has joined #openstack-ceilometer16:32
pradkgordc, bug 1263839 in openstack-ceilometer "openstack-ceilometer should requires python-oslo-policy in kilo" [Unspecified,New] - Assigned to pkilambi16:33
gordcpradk: cool stuff. i'm going to close launchpad bug16:34
sergio_nubeliugordc, what about using Aodh for alarming with kilo? is it possible?16:34
*** khushbu_ has quit IRC16:35
gordcsergio_nubeliu: that should be possible.16:38
gordcildikov: ^ any progress on the ceilometer+aodh doc?16:38
ildikovgordc: fighting with the config ref :S16:39
ildikovgordc: but I will start the install guide too soon, the content will be added to the Liberty version of the docs16:40
*** ankita_wagh has joined #openstack-ceilometer16:40
gordcildikov: great :)16:41
ildikov... but I can add a comment there that it's supposed to work with Kilo also16:41
gordcildikov: yeah, i think that would be a good idea.16:42
*** cdent has joined #openstack-ceilometer16:46
ildikovgordc: ok16:50
*** cdent has quit IRC16:53
*** tongli has joined #openstack-ceilometer16:55
*** _cjones_ has joined #openstack-ceilometer16:56
*** nadya has quit IRC16:58
*** cdent has joined #openstack-ceilometer17:00
sergio_nubeliugordc, ildikov thanks so much for your help. we'll try all on kilo!17:03
ildikovsergio_nubeliu: if you experience any issues or anything I should add to the docs, please let me know :)17:05
*** devlaps has quit IRC17:12
*** nadya has joined #openstack-ceilometer17:13
*** yprokule has quit IRC17:15
*** nadya has quit IRC17:26
*** IBerezovskiy has quit IRC17:37
gordccdent: just an fyi, pradk made a list of questions for operators in case you are interested.17:38
gordcwe'll probably start building form tomorrow17:38
*** nadya has joined #openstack-ceilometer17:43
*** harlowja has joined #openstack-ceilometer17:46
cdenthey gordc wasn't paying attention.17:54
cdentanything else besides that I should be aware of? did I see that you fixed the problem with arithmetic transforms? We must have some pretty crap tests...17:55
gordccdent: yeah.. it's here:
gordcthe test is configured to perform as it would if done in polling agent17:59
gordci'm not going to lie. i didn't make the test any better17:59
cdentwhy do I always feel like every time you say "I'm not going to lie" (which you've done more and more lately), you're lying?18:00
gordchaha. well if i'm lying here, it's actually a positive.18:01
cdentI'm feeling pretty out of sync, but won't have a chance to sync up until tomorrow morning18:03
cdentat which point I will try to catch up on pending reviews and what not18:03
cdentbut for now I'm having some serious hunger issues18:04
*** cdent has quit IRC18:04
ildikovregarding those tests my only concern that the modified tests still pass with the old code18:05
gordc... who said he could leave... how dare he leave...18:05
gordcildikov: it doesn't pass if i run just the tests... are you sure?18:05
gordcare you running unit or functional?18:06
gordcildikov: ... you sure you didn't pull in entire patch?18:09
ildikovok, I will double check the env now18:09
*** fawadkhaliq has joined #openstack-ceilometer18:11
ildikovgordc: my bad, apologies, you were right, I haven't saved one file18:12
gordcildikov: :)18:13
ildikovgordc: BTW, based on what these tests are run under the unit test suite?18:15
*** fawadkhaliq has quit IRC18:15
ildikovor they are inculded in functional too?18:16
*** nadya has quit IRC18:17
gordcildikov: no. the pipeline_base file should really be only under unit18:18
ildikovI was wonder more that how it works now18:18
ildikov... I also wonder whether we could split it somehow into multiple files, it's quite a large one18:20
gordcildikov: probably.18:21
gordcthe base file is because we use to have the two pipeline definitions18:22
gordcnow we just have the one... so really only test_decoupled_pipeline uses it18:22
gordcwe can simplify it... i just didn't.18:22
*** fawadkhaliq has joined #openstack-ceilometer18:25
alejandritogordc, ildikov guys, one quick question ... projects like, for example, cloudkitty, that consumes METRICS from ceilometer api, in case of a ceilometer / gnocchi integration ... METRICS, get consumed ONLY from gnocchi ? or we still call ceilometer api and ceilometer calls for gnocchi api for a reply ?18:30
gordcalejandrito: gnocchi represents a fundamental change in how we model data so it's not actually possible to map gnocchi metrics to ceilometer meters.18:35
gordci'm interest in what the current integration looks like between ceilometer+cloudkitty.18:36
gordci don't want to break users of ceilometer so it'd be nice to understand what you are trying to achieve to ensure we have something equivalent in gnocchi18:36
alejandritogordc, cloudkitty ( billing ) uses ceilometer's statistics resource ( and the metadata asociated ) to bill do the math and apply his own rules / formulas to bill for resources18:38
gordchmm... so we still have metadata but it's a bit more segregated from the metric value.18:40
*** fawadkhaliq has quit IRC18:41
gordcalejandrito: do you have example code of how you consume ceilometer data?18:41
alejandritogordc, sergio_nubeliu will look for an example on how it does it18:45
openstackgerritMerged openstack/aodh: Change ignore-errors to ignore_errors
ildikovalejandrito: we will surely not have pure one to one mapping, but we will have the old API kept alive long enough to give time to switch between the two18:47
ildikovalejandrito: although as how gordc says, if you can give us input that we can try to adapt to that would help for sure to make the switch as smooth as possible18:48
*** renatoarmani has quit IRC18:49
alejandritoildikov, keeping both apis alive ? that means that gnocchi will expose ceilometer's api call ? ps: regarding an example call : here's one of the mains py's
EmilienMgordc: in which section do you configure os_endpoint_type in ceilometer?18:54
ildikovalejandrito: as the data models are different, we cannot evolve the current Ceilometer API to fit so in this way yes18:55
EmilienMin service_credentials ?18:55
ildikovalejandrito: when we will have the next version of the API that will be closer to what we have now Gnocchi rather than the current Ceilometer API18:55
ildikovgordc: correct me if I'm wrong regarding ^^18:56
EmilienMgordc: nevermind, I found it, it's in ceilometer/service.py18:57
EmilienMin service_credentials section of ceilometer.conf file18:57
alejandritoildikov, gordc is that ALREADY available on gnocchi's current version ? gnocchis & ceilometers api exposed ?18:57
ildikovalejandrito: hmm, it's getting late I think I haven't interpreted your question correctly18:59
gordcalejandrito: so what ildikov was saying is that the existing api exists and you can continue to consume data as such. but we won't be making any noticable functionality to the current metering api... it's mainly in maintenance mode now18:59
gordcall new innovations in api/storage will be targetted at gnocchi.19:00
gordcEmilienM: sure.19:00
ildikovalejandrito: we have the two separate APIs and Gnocchi will be the one evolved for future usage19:00
gordcEmilienM: i charge you 5mins worth of consulting fee19:01
gordcyou can send cheque to toronto19:01
ildikovalejandrito: it should be possible though to adapt it to the actual needs if necessary19:02
EmilienMgordc: hopefuly you're cheap19:02
gordcEmilienM: 10,000$/hr... so it's cheap for only 5min19:03
gordcbut for you... 1000/hr. because we friends19:03
alejandritoildikov, perfect, i'll try to make myself clearer ... sorry if im not. so, we can use ceilometer API in maintenance mode but the data being stored on gnocchi ? or we need to maintain gnocchi and MONGODB for ceilometer backend ?19:05
EmilienMgordc: I thought PTLs worked for free, damn19:05
gordcEmilienM: oh yeah? lucky i'm ptl of puppet and not ceilometer19:06
gordcalejandrito: you can use ceilometer api in maintenance mode but it'll be stored in existing db (either mongo, sql, hbase)19:07
gordcyou can't use existing api but gnocchi storage. you need to use gnocchi api when storing in gnocchi.19:07
alejandritogordc, PERFECTLY clear, thanks ildikov also !19:09
*** nadya has joined #openstack-ceilometer19:10
openstackgerritMerged openstack/ceilometermiddleware: Change ignore-errors to ignore_errors
alejandritogordc, ildikov remember my nikname ;D ... we are starting a huge project with chargeback and showback  ... and we are deciding if we are gonna use ceilometer + mongo OR ceilometer + gnocchi  ON KILO ! ... you helped us a lot to decide, so we are starting to test ceilometer + gnocchi on kilo ... so we expect to contribute , at least with feedback or docs ( maybe code too )19:15
ildikovalejandrito: sure, sounds very cool19:16
ildikovalejandrito: feedback is always pretty much appreciated and docco also ;)19:17
ildikovalejandrito: of course code too, but that's still the easiest to get from these three things19:18
alejandritoildikov, sure !19:18
*** ankita_wagh has quit IRC19:21
*** ankita_wagh has joined #openstack-ceilometer19:21
ildikovalejandrito: tnx :)19:22
*** ankita_wagh has quit IRC19:26
openstackgerritOpenStack Proposal Bot proposed openstack/ceilometermiddleware: Updated from global requirements
*** sergio_nubeliu has quit IRC20:08
*** sergio_nubeliu has joined #openstack-ceilometer20:09
*** tongli has quit IRC20:09
alejandritoim getting this "tox.ConfigError: ConfigError: substitution key '27,34' not found" when running "tox -e genconfig" on stable/1.220:14
alejandritoanyone has gone through this ?20:14
gordcwhat's stable/1.2?20:22
*** sergio_nubeliu has quit IRC20:27
*** sergio_nubeliu has joined #openstack-ceilometer20:27
* nijaba just updated the questionaire a bit more
alejandritogordc, -b stable/1.220:36
alejandritogordc, on stable branch the same happens20:37
gordcnijaba: nice! we're just waiting on input from cdent and i think we should be good to go20:37
gordcalejandrito: ah, gnocchi. let me give it a try20:37
pradkgordc, we went through the questionaire in our team mtg, so cdent has already reviewed it fyi20:39
gordcpradk: word? aight. yeah i asked him earlier and he said he was busy with other stuff... maybe he thought it was something else.20:40
pradkgordc, was just giving you fyi.. he probably dint review the latest changes though. i'm fine with waiting20:41
alejandritogordc, i've tried on gnocchi master also and fails also after "python install" and then "tox", should i run "pip install -r requirements.txt" BEFORE toxing the config files ?20:42
gordci guess we can start building now. i'm sure he can give feedback tomorrow morning and jd or someone can check.20:42
gordcalejandrito: hmmm. you probably shouldn't shouldn't run pip install -r requirements unless you don't care that it applies it across system.20:43
gordcwhat version of tox are you using?20:43
gordci'd probably run tox -r -egenconfig20:43
alejandritogordc, python-tox 1.6.0 ubuntu 14.0420:44
gordcalejandrito: hmm. i have tox 2.1.120:44
alejandritogordc, i'll try it20:44
*** ankita_wagh has joined #openstack-ceilometer20:49
* EmilienM 20:49
gordcEmilienM: ... your irc script is broken?20:51
* gordc 20:54
* gordc 20:54
gordc /you doesn't do antyhing20:54
*** pradk has quit IRC20:55
EmilienMgordc: btw I found a new shortcut.20:55
gordcshortcut to what?20:55
EmilienMsend /me everywhere20:56
EmilienMnext time I'll try to send my IRC password eeverywhere20:56
gordcyeah good choice20:56
* gordc waits for EmilienM to get banned20:56
jd__gordc: alejandrito: you need tox >= 1.820:57
jd__gordc: alejandrito: there's a minversion in tox.ini but maybe it's ignored by old tox20:57
alejandritogordc, delelete python-tox and "pip install -U tox" to 2.1.1 and worked, thx jd__ also !20:57
gordcjd__: cool stuff20:57
gordcalejandrito: cool cool20:58
*** nadya has quit IRC21:03
*** julim has quit IRC21:06
alejandritogordc, when running "tox -e genconfig" im having this stack when installing -r requirements.txt ->
gordcalejandrito: odd21:12
gordcwhat version of urllib3 do you have 'sudo pip show urllib3'21:12
*** r-daneel has quit IRC21:14
alejandritogordc,  Version: 1.7.121:14
openstackgerritMerged openstack/ceilometer: Change ignore-errors to ignore_errors
gordcalejandrito: try bumping that up to >1.8.321:15
alejandritogordc, i saw  > Error: pg_config executable not found. :O, i'll try yours too21:16
gordcseems like there's a restriction between >=1.8.3, <1.1121:16
*** ankita_w_ has joined #openstack-ceilometer21:23
*** ankita___ has joined #openstack-ceilometer21:24
*** ankita_wagh has quit IRC21:25
*** ankita_wagh has joined #openstack-ceilometer21:26
*** ankita___ has quit IRC21:26
*** ankita_w_ has quit IRC21:27
*** promulo has joined #openstack-ceilometer21:28
*** thorst has quit IRC21:29
*** alexpilotti has quit IRC21:32
*** promulo has left #openstack-ceilometer21:33
*** promulo has joined #openstack-ceilometer21:35
*** promulo has left #openstack-ceilometer21:37
*** eglynn_ has quit IRC21:50
*** nadya has joined #openstack-ceilometer22:03
*** jaypipes has quit IRC22:05
*** alejandrito has quit IRC22:05
*** nadya has quit IRC22:08
*** dan-t has quit IRC22:08
*** gordc has quit IRC22:19
*** ddieterly has quit IRC22:29
*** alexpilotti has joined #openstack-ceilometer22:31
*** rbak has quit IRC22:54
*** edmondsw has quit IRC23:13
*** alexpilotti has quit IRC23:16
openstackgerritMerged openstack/ceilometer: Fix variable typos
*** ankita_wagh has quit IRC23:47
*** thorst has joined #openstack-ceilometer23:57

Generated by 2.14.0 by Marius Gedminas - find it at!