Monday, 2015-04-13

*** Longgeek has joined #openstack-ceilometer01:04
*** megh has joined #openstack-ceilometer01:07
*** liusheng has quit IRC01:24
*** liusheng has joined #openstack-ceilometer01:28
*** llu-laptop has joined #openstack-ceilometer01:44
*** megh has quit IRC01:54
*** Ephur has joined #openstack-ceilometer01:55
*** llu-laptop has left #openstack-ceilometer01:58
*** Kennan2 has joined #openstack-ceilometer02:02
*** Kennan has quit IRC02:03
*** Kennan has joined #openstack-ceilometer02:06
*** Kennan2 has quit IRC02:06
*** Longgeek has quit IRC02:38
*** Longgeek has joined #openstack-ceilometer02:40
openstackgerritliusheng proposed openstack/ceilometer: Fix scheduling notifications handler: InstanceScheduled
openstackgerritliusheng proposed openstack/ceilometer: Fix scheduling notifications handler: InstanceScheduled
*** meghb has joined #openstack-ceilometer03:12
*** meghb has quit IRC03:19
*** ankita_wagh has joined #openstack-ceilometer03:45
*** gpocentek has joined #openstack-ceilometer03:48
*** gpocentek has quit IRC03:48
*** gpocentek has joined #openstack-ceilometer03:48
openstackgerritgordon chung proposed openstack/ceilometer: broadcast data to relevant queues only
*** Ephur has quit IRC03:59
*** yatin has joined #openstack-ceilometer04:29
*** Kennan has quit IRC04:56
*** Kennan has joined #openstack-ceilometer04:57
*** megh has joined #openstack-ceilometer04:59
*** vijay_ has joined #openstack-ceilometer05:01
*** vijay__ has joined #openstack-ceilometer05:01
*** vijay_ has quit IRC05:01
*** vijay__ has quit IRC05:01
*** vijay_ has joined #openstack-ceilometer05:01
*** vijay__ has joined #openstack-ceilometer05:01
*** vijay_ has quit IRC05:01
*** vijay__ has quit IRC05:01
*** megh has quit IRC05:06
*** liusheng has quit IRC05:06
*** liusheng has joined #openstack-ceilometer05:06
*** 77CAAMTCY has joined #openstack-ceilometer05:26
*** 7GHABDTYB has joined #openstack-ceilometer05:26
*** ildikov has quit IRC05:53
*** afazekas has joined #openstack-ceilometer06:08
*** Longgeek has quit IRC06:14
*** Longgeek has joined #openstack-ceilometer06:14
*** Ala has joined #openstack-ceilometer06:21
*** rwsu has joined #openstack-ceilometer06:34
*** Kennan has quit IRC06:46
*** Kennan has joined #openstack-ceilometer06:46
*** _nadya_ has joined #openstack-ceilometer06:57
*** safchain has joined #openstack-ceilometer07:00
*** kaliya_ has quit IRC07:06
*** kaliya_ has joined #openstack-ceilometer07:06
*** ildikov has joined #openstack-ceilometer07:14
*** ankita_wagh has quit IRC07:14
*** ankita_wagh has joined #openstack-ceilometer07:15
*** 77CAAMTCY has quit IRC07:16
*** 7GHABDTYB has quit IRC07:16
*** ankita_wagh has quit IRC07:19
*** eglynn has quit IRC07:20
*** _nadya_ has quit IRC07:22
*** eglynn has joined #openstack-ceilometer07:52
*** eglynn has quit IRC07:57
openstackgerritMehdi Abaakouk proposed openstack/gnocchi: Documentation typo
openstackgerritMehdi Abaakouk proposed openstack/gnocchi: Correctly enable keystone middleware
*** vijay_ has joined #openstack-ceilometer08:02
*** vijay__ has joined #openstack-ceilometer08:02
*** eglynn has joined #openstack-ceilometer08:02
*** deepthi has joined #openstack-ceilometer08:05
*** _nadya_ has joined #openstack-ceilometer08:11
*** eglynn has quit IRC08:15
*** lsmola_ has joined #openstack-ceilometer08:16
*** _nadya_ has quit IRC08:32
*** Ala has quit IRC08:34
*** eglynn has joined #openstack-ceilometer08:36
*** Ala has joined #openstack-ceilometer08:39
*** IBerezovskiy has joined #openstack-ceilometer08:43
*** haigang has joined #openstack-ceilometer08:46
openstackgerritliusheng proposed openstack/ceilometer: Fix scheduling notifications handler: InstanceScheduled
openstackgerritMerged openstack/gnocchi: Documentation typo
*** ildikov has quit IRC09:04
*** spredzy has left #openstack-ceilometer09:05
*** haigang has quit IRC09:13
*** haigang has joined #openstack-ceilometer09:15
*** andymaier has joined #openstack-ceilometer09:17
*** cdent has joined #openstack-ceilometer09:17
*** _nadya_ has joined #openstack-ceilometer09:25
*** haigang has quit IRC09:27
*** vijay__ has quit IRC09:28
*** vijay_ has quit IRC09:28
*** haigang has joined #openstack-ceilometer09:28
*** kaliya_ has quit IRC09:39
*** theanalyst has quit IRC09:42
*** theanalyst has joined #openstack-ceilometer09:44
*** haigang has quit IRC09:54
*** haigang has joined #openstack-ceilometer09:55
openstackgerritJulien Danjou proposed openstack/gnocchi: doc: fixes a few typos and update some data
*** haigang has quit IRC10:00
*** cdent has quit IRC10:06
*** cdent has joined #openstack-ceilometer10:18
openstackgerritMerged openstack/gnocchi: Correctly enable keystone middleware
*** Ala has quit IRC10:21
*** Ala has joined #openstack-ceilometer10:27
openstackgerritJulien Danjou proposed openstack/gnocchi: doc: fixes a few typos and update some data
*** deepthi has quit IRC11:16
*** yatin has quit IRC11:17
*** yatin has joined #openstack-ceilometer11:21
*** _nadya_ has quit IRC11:26
cdentjd__: you mind if I quote you in my gabbi presentation: "I'd prefer we go full gabbi where possible. It's less verbose and less error prone than using the generic test framework."11:27
*** yatin has quit IRC11:37
jd__cdent: now sure go ahead :)11:40
*** liusheng has quit IRC11:44
openstackgerritIlya Tyaptin proposed openstack/ceilometer-specs: Spec for using a aggregation pipeline in MongoDB
*** _nadya_ has joined #openstack-ceilometer11:53
*** Ala has quit IRC12:10
*** Ala has joined #openstack-ceilometer12:12
openstackgerritRyota MIBU proposed openstack/ceilometer-specs: Add spec for notification alarm evaluator
*** EmilienM|afk is now known as EmilienM12:19
*** Ephur has joined #openstack-ceilometer12:33
*** Ephur has quit IRC12:37
*** Ephur has joined #openstack-ceilometer12:39
*** _nadya_ has quit IRC12:41
*** jroll has quit IRC12:50
*** jroll has joined #openstack-ceilometer12:50
*** gordc has joined #openstack-ceilometer12:51
*** _nadya_ has joined #openstack-ceilometer12:57
*** openstackgerrit has quit IRC13:00
*** openstackgerrit has joined #openstack-ceilometer13:03
openstackgerritChris Dent proposed openstack/gnocchi: Remove an xfail test that is now handled by another test
openstackgerritMehdi Abaakouk proposed openstack/gnocchi: WIP: Fix keystone authentification
*** ozialien has joined #openstack-ceilometer13:16
*** jruano has joined #openstack-ceilometer13:19
openstackgerritMehdi Abaakouk proposed openstack/gnocchi: WIP: Fix keystone authentification
*** prad has joined #openstack-ceilometer13:28
openstackgerritMehdi Abaakouk proposed openstack/gnocchi: WIP: Fix keystone authentification
openstackgerritCiaran O Tuathail proposed openstack/python-ceilometerclient: Updates examples to reference new sample format
*** thomasem has joined #openstack-ceilometer13:53
*** thomasem has quit IRC13:54
*** thomasem has joined #openstack-ceilometer13:55
openstackgerritMerged openstack/gnocchi: Remove an xfail test that is now handled by another test
*** jruano has quit IRC14:04
*** kodoku has joined #openstack-ceilometer14:05
kodokuHi, Is it possible to enable ssl on ceilometer's API without apache or nginx ? with ceilometer.conf14:06
*** jruano has joined #openstack-ceilometer14:23
*** rbak has joined #openstack-ceilometer14:35
*** Longgeek has quit IRC14:45
*** _nadya_ has quit IRC14:45
*** jruano has quit IRC14:46
*** Longgeek has joined #openstack-ceilometer14:46
*** rwsu has quit IRC14:54
*** rwsu has joined #openstack-ceilometer14:54
gordckodoku: nope. you'll need to patch the api i believe.14:56
*** jruano has joined #openstack-ceilometer15:01
kodokugordc ok thx ;)15:09
kodokuso this is the only project core of openstack which has not a ssl conf.15:10
cdentkodoku: ceilometer's api runs a lot more reliably under mod_wsgi (or some other wsgi container) than it does running itself15:11
cdent(if that's any consolation)15:11
cdenteventlet gets flakey under load15:11
gordc^ i should probably have given that as a detailed explanation.lol15:12
kodokuyes thx ;)15:12
kodokuSo I'll add an local apache for ceilometer15:13
cdentIf you've got other api services on the same node, may as well put them with mod_wsgi to, as they'll show the same issues under eventlet15:15
cdent(those where it is possible)15:15
*** afazekas has quit IRC15:18
*** zqfan has joined #openstack-ceilometer15:23
*** pauldix has joined #openstack-ceilometer15:23
*** Ala has quit IRC15:30
gordcwill post again at meeting but for those with session proposals ready:
*** rbak has quit IRC15:33
*** kodoku has quit IRC15:38
*** rbak has joined #openstack-ceilometer15:42
*** claudiub has joined #openstack-ceilometer15:42
*** pauldix has quit IRC15:45
*** _cjones_ has joined #openstack-ceilometer15:47
cdentgordc: does "burn it all down" require a full spec?15:52
gordccdent: yes. i'd like to know what sort of accelerant you'll be using...15:54
gordci suggest you avoid using google because you will be added to all kinds of watch lists15:54
pradi would like to review that spec15:55
cdentspec as performance art15:55
*** afazekas has joined #openstack-ceilometer16:10
*** Longgeek has quit IRC16:18
*** ozialien has quit IRC16:22
*** ankita_wagh has joined #openstack-ceilometer16:23
*** lsmola_ has quit IRC16:25
*** afazekas has quit IRC16:30
*** safchain has quit IRC16:32
*** sputnik13 has joined #openstack-ceilometer16:40
*** eglynn has quit IRC16:41
*** jruano has quit IRC16:52
*** afazekas has joined #openstack-ceilometer16:52
*** dikonoor has joined #openstack-ceilometer16:57
dikonoorgordc: This is about LP bug
openstackLaunchpad bug 1364708 in Ceilometer "notifications.error should be consumed or not be sent" [Medium,Fix released] - Assigned to gordon chung (chungg)16:58
dikonoorThis is about ceilometer consuming error notifications.16:59
dikonoorgordc: I went through the fix and if I am not wrong , the fix is applicable only in the case the store_events is set to true16:59
dikonoorgordc: Also , the changes made consume only notifications.error and not for notifications.warn17:00
*** sbfox has joined #openstack-ceilometer17:00
dikonoorzqfan : The bug was reported by you. Could you confirm if my understanding above is right/wrong?17:03
openstackLaunchpad bug 1364708 in Ceilometer "notifications.error should be consumed or not be sent" [Medium,Fix released] - Assigned to gordon chung (chungg)17:03
gordcdikonoor: are there notifications.warn events?17:04
gordcand yes, it only applies to events right now. i don't know if there's any meters you can actually generate from the error notifications.17:05
*** andymaier has quit IRC17:06
dikonoorgordc : I am not completely sure if there are notifications.warn..but someone reported it few days back (that the notifications.error and .warn ) and lying around in the queue not being consumed17:07
*** sputnik13 has quit IRC17:08
*** afazekas has quit IRC17:09
gordcdikonoor: hmm... if there are warn notifications, we should capture them. i just didn't add them because i couldn't find any17:10
gordcnot that i looked particularly hrad.17:10
*** sbfox has quit IRC17:12
*** ozialien has joined #openstack-ceilometer17:12
dikonoorgordc: On a setup with store_events set to false , there were 7 errors and 300 warns found lying idle in the queue.17:12
dikonoorgordc : Once I find the warnings, I will open a LP bug17:12
gordcdikonoor: i guess it doesn't hurt to target warn priority level as well.17:13
dikonoorgordc : yeah ..I am not sure if there are more ..i e anything other than error, warn and info17:13
gordcdikonoor: regarding the store_events part. i don't think that is a bug. we should only consume what we will use so if you aren't making events, then ceilometer shouldn't be pulling in the notifications.17:15
*** sbfox has joined #openstack-ceilometer17:17
dikonoorgordc: I agree with that part.  But I am not sure if all the event notifications generated from different services are consumed /captured by ceilometer only as events. I guess at least in Kilo they are consumed as both meter (for backward compatibility perhaps) and event. Isn't that so17:19
dikonoorgordc : and thus the changes made as part of this fix will not benefit those who have still not moved on to start using the ceilometer events17:20
dikonoorgordc : and probably that was intentional17:20
gordcdikonoor: sort of. regardless of events, meters are derived from specific, known notifications.17:21
gordceven if you add error and warn to the meter endpoints, it won't create any new meters as the event_types of these notifications are unknown.17:22
gordc(it may create notifications if for some reason the event_types are named the same)17:22
gordcbut for the most part, all it will do is act as a 'cleaner' and just pull messages from the error/warn queues and toss them out.17:23
dikonoorgordc : ok..I wasn't aware that .error (and probably .warn) do not have any associated event_types (other than unknown)17:24
gordcdikonoor: well they have event_types... they're just unknown from Ceilometer pov.17:25
gordcone way to think of it is, for events, we capture everything to represent the entire system.17:25
gordcfor meters, we know we can grab certain measurements from specific notifications. so we listen only to those.17:26
dikonoorgordc : ok..makes sense.. That answers my queries..Thanks very much for your time !17:27
gordcthat's not to say there aren't measurements in the error/warn notifications... we just need to be specific about those notifications.17:27
gordcdikonoor: i agree about the warn priority level though. we might want to add that as you can clearly see it in your environment.17:28
*** jruano has joined #openstack-ceilometer17:29
dikonoorgordc : yeah..Let me try to generate it again.Will keep you posted on that17:30
gordcdikonoor: cool cool.17:30
*** ozialien has quit IRC17:31
*** afazekas has joined #openstack-ceilometer17:34
*** dikonoor has quit IRC17:36
*** dikonoor has joined #openstack-ceilometer17:36
*** harlowja_away is now known as harlowja17:37
*** yatin has joined #openstack-ceilometer17:49
*** yatin has quit IRC17:50
*** megh has joined #openstack-ceilometer17:50
*** r-daneel has joined #openstack-ceilometer17:51
*** eglynn has joined #openstack-ceilometer17:52
*** ozialien has joined #openstack-ceilometer17:52
*** eglynn has quit IRC17:57
*** eglynn has joined #openstack-ceilometer17:57
*** afazekas has quit IRC17:57
*** afazekas has joined #openstack-ceilometer18:01
*** dikonoor has quit IRC18:05
*** afazekas has quit IRC18:05
*** edmondsw has joined #openstack-ceilometer18:09
*** renatoarmani has joined #openstack-ceilometer18:17
*** zqfan has quit IRC18:19
*** sputnik13 has joined #openstack-ceilometer18:19
*** cdent has quit IRC18:20
*** drjones has joined #openstack-ceilometer18:25
*** _cjones_ has quit IRC18:25
*** ozialien has quit IRC18:33
*** msly has joined #openstack-ceilometer18:47
*** fnaval has joined #openstack-ceilometer18:51
*** eglynn has quit IRC18:55
meghgordc: can you please help out with review
meghgordc: I've submitted FFE and addressed the review comments19:05
*** jmank has joined #openstack-ceilometer19:13
*** promulo_ has joined #openstack-ceilometer19:28
*** promulo__ has quit IRC19:28
*** _cjones_ has joined #openstack-ceilometer19:30
*** drjones has quit IRC19:31
*** _cjones_ has quit IRC19:46
*** _cjones_ has joined #openstack-ceilometer19:46
*** cdent has joined #openstack-ceilometer19:48
*** renatoarmani has quit IRC20:00
*** cdent has quit IRC20:03
*** jruano has quit IRC20:03
*** jbrendel has joined #openstack-ceilometer20:04
*** promulo__ has joined #openstack-ceilometer20:04
*** jbrendel has left #openstack-ceilometer20:04
*** promulo_ has quit IRC20:05
*** jbrendel has joined #openstack-ceilometer20:06
*** jruano has joined #openstack-ceilometer20:09
gordcmegh: hi, we tagged our rc1 build at the end of last week already. i'll bring it up with Eoghan but this might not make it into kilo given timing.20:12
*** sbfox has quit IRC20:24
*** srsakhamuri has joined #openstack-ceilometer20:25
*** sbfox has joined #openstack-ceilometer20:30
srsakhamurigordc: do you know if gnocchi API uses the models that are similar to current models used in v2 api? Or they are entirely different?20:31
meghgordc: ok thanks20:34
*** noye has joined #openstack-ceilometer20:43
*** rjaiswal has joined #openstack-ceilometer20:43
*** sbfox has quit IRC20:46
edmondswgordc, is still failing in the gate... how are things coming on fixing that?20:47
edmondswthis is the one we talked about last week... Requirement.parse('python-ceilometerclient<=1.0.12,>=1.0.6')20:47
edmondswdhellmann had indicated that we needed to have icehouse and juno pull from something other than master20:48
edmondswneed to get this fixed so we can merge and then spin ceilometerclient 1.0.1420:49
gordcsrsakhamuri: they're different although i'd say the resources aspect is similar. (with a few additional attributes)20:53
gordcsrsakhamuri: trying to see if there's any good defintion on metrics part as that's probably the key part20:54
gordcsrsakhamuri: o20:56
gordci'd say eoghan's deck is probably a good comparison:
gordcedmondsw: let me try creating branch again. neither eoghan or i had permission last week. (i need to check what correct command is again)20:57
srsakhamurigordc: is there a equivalent in gnocchi? Thanks I will take a look20:57
*** jbrendel has quit IRC20:57
*** rjaiswal has quit IRC20:58
gordcthis might be the equivalent:
*** rjaiswal has joined #openstack-ceilometer20:58
srsakhamuriThanks again gordc, that is helpful.21:00
noyehow do i add a new meter ?21:01
*** ozialien has joined #openstack-ceilometer21:04
gordcnoye: depends... where is the meter coming from? are you polling from openstack service apis, building it from notificatoins, or other?21:08
*** cdent has joined #openstack-ceilometer21:09
noye@gordc what i need is this meter storage.containers.objects.outgoing.bytes21:10
noyewhere, currently there’s only storage.objects.outging.bytes21:11
noyeideally, i’d like to be able to do a command like this: ceilometer statistics —meter etc…..21:12
noyeso i guess what i’m asking is there a commandline function for adding meters, or do have to create the code or what?21:14
noyeI’m happy to read any documentation, i’m just not finding any examples or info on how to do it21:15
gordchmm. can that be calculated via api? how we calculate it currently:
*** jbrendel has joined #openstack-ceilometer21:18
gordcif not, you could post samples via api (without code change)21:18
noyeok, let me see where i get with that code21:20
noyethanks @gordc21:20
*** amalagon has joined #openstack-ceilometer21:20
gordcnoye: np. this might help:
dhellmanngordc: just confirming, the stable/icehouse and stable/juno branches of python-ceilometerclient should start with -- is that right?21:23
gordcdhellmann: looking at patch21:23
dhellmannthat should correspond to 1.0.1221:23
cdentcontainer is in resource metadata noye so it ought to be possible to make a statistics query that matches on the bytes for just a particular container21:23
gordcdhellmann: yep. that makes sense to me.21:24
cdentgordc you're gonna be a busy fellow for a few months, I'm already feeling bad for you and you've barely started21:24
gordc... /me secretly practicing my deferring skills21:25
*** amalagon has quit IRC21:26
dhellmanngordc: branches are created; you may want to fast-approve and to make other updates more convenient21:27
noye@cdent good, but how, lol21:27
*** jbrendel has quit IRC21:27
cdentnoye hmmm, yes, excellent question, I've never been quite clear on such things, lemme try a little experiment21:28
gordcdhellmann: awesome. will do that. thanks for the help!21:30
meghgordc: Re: If not rc1, then can the change go in before kilo rc2 ? It's not a big change21:31
noye@cdent much appreciated21:31
openstackgerritPradeep Kilambi proposed openstack/gnocchi: Use Archive Policy Rule in create metric api
gordcmegh: typically, we reserve rc2+ builds for critical bugs. rc1 was the time to get less critical stuff in.21:38
gordcmegh: i'll speak with eglynn since he controls the releases for Kilo but too be honest, this might be difficult (not saying impossible)21:39
meghgordc: ok thanks for taking this up and please let me know21:40
cdentnoye: I haven't got any swift meters available to test but a command line query like so21:44
cdentceilometer statistics -m disk.write.requests.rate -q
cdentmight get you started so you could try21:44
noye@cdent yeah, i’ve done something simular21:44
noyeno dice21:45
cdentceilometer statistics -m storage.objects.outgoing.bytes -q metadata.container=<container>21:45
cdentwhat did you try?21:45
* gordc sneaks out... may sign on later...21:45
*** gordc has quit IRC21:45
noye -q metadata.container21:45
noyehaven’t tried that21:45
*** jruano has quit IRC21:48
cdentSadly I find the query syntax and the entire process of exploring the ceilo data to be more than a little bit confusing :(21:48
cdentnoye: i'll be back in a few minutes, gotta checkon something21:49
*** MVenesio has joined #openstack-ceilometer21:51
cdentI'm back now noye21:52
cdentany luck with -q metadata.container?21:52
cdenthave you confirmed that you have samples for the storage.objects.outgoing.bytes meter?21:54
openstackgerritPradeep Kilambi proposed openstack/gnocchi: Add support for ArchivePolicy Rule Endpoints
openstackgerritPradeep Kilambi proposed openstack/gnocchi: Use Archive Policy Rule in create metric api
cdentThe reason I ask noye is because in my environment it looks like I should but I don't21:55
noyeceilometer statistics -m storage.objects.outgoing.bytes meter21:55
noyegives good data21:55
noyebut i only want data for container, bytes in/out respectfully21:56
cdentcan you do: ceilometer sample-list -q meter=storage.objects.outgoing.bytes -l 121:57
*** ankita_w_ has joined #openstack-ceilometer21:57
cdentand then take the id from the one sample you should get back and do a21:58
cdentceilometer sample-show <the id>21:58
cdentand paste the output somewhere please?21:58
noyesure, just working between two machines, give me minute21:58
*** ozialien has quit IRC21:59
cdentno problem, I'll be around for a while21:59
cdentin the meantime I'll try and get my setup to produce some proper samples21:59
*** ankita_wagh has quit IRC22:00
noyeroot@os1:~# ceilometer sample-list -q meter=storage.objects.outgoing.bytes -l 122:01
noyeusage: ceilometer sample-list [-q <QUERY>] -m <NAME> [-l <NUMBER>]22:01
noyeceilometer sample-list: error: argument -m/--meter is required22:01
cdentwhat version are you in?22:03
cdentnoye you can do:22:05
cdentceilometer sample-list -m storage.objects.outgoing.bytes -l 122:05
noyeroot@os1:~# ceilometer --version22:05
cdentSorry I mean more generally icehouse, juno, etc?22:06
cdentsince you're on something older that what I'm on, instead of sample-show do:22:08
cdentceilometer resource-show 4b5fa6a2761e40c988cb89af4dd21fc722:08
cdent(and paste that)22:09
*** rjaiswal has quit IRC22:10
cdent(what we're doing it trying to figure out what metadata is available to query)22:13
cdent(and how it is structured)22:13
*** megh has quit IRC22:28
*** MVenesio has quit IRC22:30
*** Ephur has quit IRC22:43
*** megh has joined #openstack-ceilometer22:49
*** rbak has quit IRC23:00
*** cdent has quit IRC23:03
*** SergeyLukjanov has quit IRC23:39
*** SergeyLukjanov has joined #openstack-ceilometer23:40
*** r-daneel has quit IRC23:43

Generated by 2.14.0 by Marius Gedminas - find it at!