Tuesday, 2014-05-27

*** kun_huang has joined #openstack-ceilometer00:06
*** kun_huang has quit IRC00:09
*** nati_ueno has joined #openstack-ceilometer00:12
*** nati_ueno has quit IRC00:17
*** packet has quit IRC00:18
*** nati_uen_ has joined #openstack-ceilometer00:22
*** matsuhashi has joined #openstack-ceilometer00:26
*** nati_ue__ has joined #openstack-ceilometer00:27
*** nati_uen_ has quit IRC00:30
*** _cjones_ has quit IRC00:31
openstackgerritAlexei Kornienko proposed a change to openstack/ceilometer: Refactor tests to remove direct access to test DBManagers  https://review.openstack.org/9561300:31
*** nati_ue__ has quit IRC00:32
*** matsuhashi has quit IRC00:33
*** matsuhashi has joined #openstack-ceilometer00:34
*** nati_uen_ has joined #openstack-ceilometer00:35
*** nati_uen_ has quit IRC00:39
*** igor_ has joined #openstack-ceilometer00:52
*** nati_uen_ has joined #openstack-ceilometer00:52
*** nati_ue__ has joined #openstack-ceilometer00:54
*** igor_ has quit IRC00:57
*** nati_uen_ has quit IRC00:57
*** nati_ueno has joined #openstack-ceilometer00:57
*** nati_ue__ has quit IRC00:59
*** nati_ueno has quit IRC00:59
*** nati_ueno has joined #openstack-ceilometer01:00
*** nati_ueno has quit IRC01:04
*** nati_ueno has joined #openstack-ceilometer01:09
*** matsuhashi has quit IRC01:33
*** nati_ueno has quit IRC01:33
*** matsuhashi has joined #openstack-ceilometer01:34
*** zqfan has joined #openstack-ceilometer01:42
*** nosnos has joined #openstack-ceilometer01:44
*** Ruetobas has quit IRC01:50
*** Ruetobas has joined #openstack-ceilometer01:51
*** igor_ has joined #openstack-ceilometer01:53
*** Ruetobas has quit IRC01:56
*** Ruetobas has joined #openstack-ceilometer01:56
*** igor_ has quit IRC01:57
*** nati_ueno has joined #openstack-ceilometer02:00
*** nati_ueno has quit IRC02:10
*** _cjones_ has joined #openstack-ceilometer02:12
*** _cjones_ has quit IRC02:16
*** nati_ueno has joined #openstack-ceilometer02:19
*** nati_ueno has quit IRC02:25
*** nati_ueno has joined #openstack-ceilometer02:33
*** igor_ has joined #openstack-ceilometer02:54
*** zqfan has quit IRC02:58
*** igor_ has quit IRC02:58
*** zqfan has joined #openstack-ceilometer03:04
*** _cjones_ has joined #openstack-ceilometer03:13
*** _cjones_ has quit IRC03:17
*** nosnos has quit IRC03:25
*** raymondr has joined #openstack-ceilometer03:31
*** igor_ has joined #openstack-ceilometer03:55
*** igor_ has quit IRC03:59
openstackgerritZhiQiang Fan proposed a change to openstack/python-ceilometerclient: Avoid empty entity field in uri path  https://review.openstack.org/9130104:01
*** psharma has joined #openstack-ceilometer04:12
*** nati_ueno has quit IRC04:31
openstackgerritZhiQiang Fan proposed a change to openstack/python-ceilometerclient: Avoid unnecessary stderr message when run test  https://review.openstack.org/9564004:32
*** nati_ueno has joined #openstack-ceilometer04:32
*** nosnos has joined #openstack-ceilometer04:46
*** raymondr has quit IRC04:49
*** igor_ has joined #openstack-ceilometer04:55
*** ildikov has quit IRC04:56
*** igor_ has quit IRC05:00
*** _cjones_ has joined #openstack-ceilometer05:14
*** _cjones_ has quit IRC05:19
*** psharma has quit IRC05:27
*** psharma has joined #openstack-ceilometer05:29
*** ildikov has joined #openstack-ceilometer05:40
*** _nadya_ has joined #openstack-ceilometer05:46
*** igor_ has joined #openstack-ceilometer05:56
*** igor_ has quit IRC06:00
openstackgerritOpenStack Proposal Bot proposed a change to openstack/ceilometer: Imported Translations from Transifex  https://review.openstack.org/8850606:01
*** _nadya_ has quit IRC06:08
*** _nadya_ has joined #openstack-ceilometer06:11
*** inc0 has joined #openstack-ceilometer06:18
*** gpocentek has quit IRC06:31
*** gpocentek has joined #openstack-ceilometer06:34
*** liusheng has quit IRC06:43
*** liusheng has joined #openstack-ceilometer06:43
*** igor_ has joined #openstack-ceilometer06:57
*** igor_ has quit IRC07:01
*** boris-42 has quit IRC07:03
*** boris-42 has joined #openstack-ceilometer07:11
*** _cjones_ has joined #openstack-ceilometer07:15
*** nati_ueno has quit IRC07:18
*** nati_ueno has joined #openstack-ceilometer07:19
*** _cjones_ has quit IRC07:20
*** nati_ueno has quit IRC07:23
openstackgerritChristian Berendt proposed a change to openstack/ceilometer: use LOG instead of logger as name for the Logger object  https://review.openstack.org/9566607:39
*** nati_ueno has joined #openstack-ceilometer07:50
*** nati_ueno has quit IRC07:54
*** safchain has joined #openstack-ceilometer07:54
*** nacim has joined #openstack-ceilometer08:02
*** cdent has joined #openstack-ceilometer08:09
*** pnascimento has joined #openstack-ceilometer08:11
*** Ju has joined #openstack-ceilometer08:19
*** eglynn has joined #openstack-ceilometer08:22
*** igor_ has joined #openstack-ceilometer08:41
openstackgerritChristian Berendt proposed a change to openstack/ceilometer: use logger instead of LOG as name for the Logger object attribute  https://review.openstack.org/9566608:43
*** nosnos has quit IRC08:48
inc0hello, I'll risk statement that we have probobly one of biggest ceilometer installation, which uses sqlalchemy. I've just stumbled upon table with 200m + records...and thats just one of 2 datacenters...08:48
*** nosnos has joined #openstack-ceilometer08:48
*** nati_ueno has joined #openstack-ceilometer08:50
openstackgerritIhar Hrachyshka proposed a change to openstack/ceilometer: Synced jsonutils from oslo-incubator  https://review.openstack.org/9569008:51
*** nosnos has quit IRC08:53
*** nati_ueno has quit IRC08:55
eglynninc0: how long is that running for?09:20
eglynninc0: is the polling cadence set to the out-of-the-box default (600s) or faster?09:20
eglynninc0: what order of magnitude of resources (instances, volumes, images, etc.) exist at any given time in your cloud?09:21
eglynninc0: did you have TTL enabled at any point?09:21
inc0eglynn, about month, polling is default, we have almost 2000vms running currently and growing, and no TTL so fat09:22
inc0since we are going to use this data for billing, I'd like to keep them09:22
eglynninc0: yeah, so doing the math ...09:25
eglynninc0: we have 20+ compute meters09:25
eglynninc0: ... and 30 days * 24 hrs * 6 polling cycles per hour = 4320 polling cycles09:26
*** idegtiarov has joined #openstack-ceilometer09:26
eglynninc0: ... so 4320 cycles * 2000 VMs * ~20 meters == ~172,800,000 samples09:26
inc0yeah, sounds about right09:27
eglynninc0: ... so that would be the correct ballpark09:27
inc0and then there is metadata table09:27
inc0tables *09:27
eglynninc0: ... so the main problem I guess is not so much the sheer number of samples, as the fact that each *individual* sample is bloated with a resource metadata snapshot09:28
eglynninc0: ... this is what we are trying to address with the v3/TSDaaS/gnocchi initiative for Juno09:28
cdent(is there a medium term fix in normalizing that association?)09:28
eglynninc0: ... for background see the summit etherpad https://etherpad.openstack.org/p/ceilometer-tsdaas09:28
openstackgerritA change was merged to openstack/ceilometer: Fix network notifications of neutron bulk creation  https://review.openstack.org/9370209:29
*** admin0 has joined #openstack-ceilometer09:30
eglynncdent: ... there are various ideas around improving the v2 schema so as to perform & scale better09:30
inc0for now I'll play with pipeline.yaml to limit meters I'm gathering, since right now I just need instance:flavor and volume meters09:30
eglynncdent: ... but we're still hamstrung by the necessity to keep these metadata "snapshots" around09:30
eglynninc0: that's a sensible approach09:31
inc0TTL removes records right?09:31
inc0might be worth considering mechanism which will stash them in some other db instead of deleting09:32
eglynninc0: yeap09:32
eglynnjust a straight deletion supported right now, but we had previous discussions about gradually aging the samples out via a sequence of ever more coarse-grained roll-ups09:33
inc0for some use cases (myself included;)) that would be relative quick fix imho, I only work on data from last 2 days or so, but I need them archived just in case (billing system)09:33
inc0yeah, that would be best09:34
eglynninc0: we'd need to work out the details of how the archiving transfer from the "live" DB would work without interfering too much with throughput of fresh writes09:34
*** nosnos has joined #openstack-ceilometer09:35
eglynninc0: maybe some kind of per-day partitioning of the metering DB, where the oldest "partition" falls off the end as each day rolls around09:35
* eglynn is just riffing ...09:36
inc0I thought of partitions also09:36
inc0embedded func might also do the trick, but then we have different sql databases09:36
eglynninc0: if you would like to capture these ideas in a BP, please do so09:37
eglynninc0: (... by proposing a spec based on https://github.com/eglynn/ceilometer-specs/blob/master/specs/template.rst to gerrit, as opposed to filing on LP)09:38
eglynncdent: :)09:38
*** kun_huang has joined #openstack-ceilometer09:38
eglynnpnascimento: good morning sir!09:38
eglynnpnascimento: I answered your unit testing questions on the ML09:39
eglynnpnascimento: http://lists.openstack.org/pipermail/openstack-dev/2014-May/035913.html09:39
eglynnpnascimento: ... but prolly more efficient to bring such detailed questions to the channel here, as opposed to the ML09:39
pnascimentoeglynn: good morning (no need for the 'sir' part :P)09:40
pnascimentoeglynn: I have seen, I was reading it now ;)09:41
eglynnpnascimento: cool enough09:41
DinaBelovajd__, sileht - can you advise the config file for the gnocchi? :)09:42
pnascimentoeglynn: the only reason that I used the ML, it's because I'm not always connected to the Internet (IRC), I "needed" some kind of offline conversation and tried to reach more people through ML... if I have done some kind of misusage of ML, sorry for that, it was not my intention09:44
eglynnpnascimento: it's all good ...09:45
eglynnpnascimento: .... no "misusage of the ML" was implied by my point above09:45
eglynnpnascimento: ... just that you're more likely to get to an quick answer with a short specific question posed on this channel09:46
eglynnpnascimento: ... though of course there's no "SLA" on IRC interactions either09:46
*** nacim has quit IRC09:47
eglynnpnascimento: (... best-effort for ML and IRC)09:47
eglynnpnascimento: makes sense?09:47
eglynnDinaBelova: ... one of the things I asked (OPW intern) anamalagon to do this week is to extend the gnocchi README.rst with a "getting started" guide for newbies09:49
eglynnDinaBelova: (how to setup swift backend; config required; how to launch the gnocchi-api console script etc...)09:50
pnascimentoeglynn: totally! :) as I said, the only reason it was by need of a "asynchronous" conversation, but in the future (e.g. today :P) I will do it as you said it, in a best-effort for ML and IRC :)09:50
eglynnDinaBelova: ... so any info you can throw her way based on your experience of getting up & running, would be useful09:50
*** nati_ueno has joined #openstack-ceilometer09:51
eglynnpnascimento: ... not a biggie either way09:51
eglynnpnascimento: ... you're free to post to the ML anytime you want (with any dev-related questions)09:52
cdenteglynn: jumping back to the metadata topic for a quick clarifying question: the issue is samples and metadata stored 1:1 despite metadata being mostly static, yes?09:53
eglynncdent: yes09:54
cdentso, my comment above was: would it possible to store samples with id-of-metadata-that-hasn't-chagned?09:54
cdentsorry N:109:55
*** nati_ueno has quit IRC09:55
cdent(i guess with gnocchi this is mostly an academic question)09:56
eglynncdent: yep, I understood that09:56
eglynncdent: ... it could possibly be part of a medium term approach09:56
eglynncdent: ... specifically this idea of dumping the incoming data to "raw" table09:57
eglynncdent: ... before asynchronously hiving it off to normalized "analysis" tables09:57
cdentI would have thought sql-a could do that processing without needing the raw table?09:58
*** matsuhashi has quit IRC09:58
cdentor is it too expensive at write time?09:58
*** matsuhashi has joined #openstack-ceilometer10:00
eglynncdent: ... just one of many ideas floating around, but yes the motivation for the "dump to raw first" approach is to minimize the write latency for incoming samples10:01
*** nacim has joined #openstack-ceilometer10:01
eglynncdent: ... since the collector observed as the bottleneck in the gate10:02
eglynncdent: ... fyi your last appears as unprintable chars in my IRC client10:02
cdentbummer, it's a thumbs up10:03
eglynncdent: a-ha ... dumb Konversation client ;)10:03
*** nacim has quit IRC10:03
*** nacim has joined #openstack-ceilometer10:04
*** matsuhashi has quit IRC10:07
*** nosnos has quit IRC10:07
*** matsuhashi has joined #openstack-ceilometer10:07
ildikoveglynn: morning10:09
ildikoveglynn: do you have any idea about how gate jobs are specified and mainly where?10:10
eglynnildikov: wb!10:10
eglynnildikov: ... yeah those jobs are specified in the infra-config repos10:11
eglynnildikov: ... one sec10:11
*** matsuhashi has quit IRC10:12
ildikoveglynn: sorry, I was interrupted, be back in 20 mins10:12
eglynnildikov: k10:12
eglynnildikov: is this what you're looking for (job builder config for the devstack gate ...)10:17
*** zqfan has quit IRC10:21
DinaBelovaeglynn, thanks for the news about "how to start with gnocchi" :)10:21
eglynnDinaBelova: ... cool, so if you do come across any gems of wisedom or any gotchas worth documenting, please pass these on to Ana so she can add to her README10:25
eglynnDinaBelova: ... thanks!10:25
DinaBelovaeglynn, sure)10:25
openstackgerritNejc Saje proposed a change to openstack/python-ceilometerclient: Check if the alarm has time constraints field before displaying  https://review.openstack.org/9571310:27
*** admin0 has quit IRC10:28
*** admin0 has joined #openstack-ceilometer10:28
*** nacim_ has joined #openstack-ceilometer10:31
*** nacim has quit IRC10:33
*** _nadya_ has quit IRC10:44
*** Alexei_987 has left #openstack-ceilometer10:46
*** nati_ueno has joined #openstack-ceilometer10:52
openstackgerritChristian Berendt proposed a change to openstack/ceilometer: replaced unicode() with six.text_type()  https://review.openstack.org/9548810:56
*** nati_ueno has quit IRC10:56
*** pnascimento has quit IRC11:04
ildikoveglynn: hmm, I did not find there the doc realted jobs or maybe I was on vacation ofr a too long time interval for recognizing it :)11:08
eglynnildikov: ... do you mean the docco team's job, or?11:11
eglynnildikov: ... https://git.openstack.org/cgit/openstack-infra/config/tree/modules/openstack_project/files/jenkins_job_builder/config/manuals-jobs.yaml11:11
eglynnildikov: ... and this seems to be the one for the operations guide11:11
eglynnildikov: ... https://git.openstack.org/cgit/openstack-infra/config/tree/modules/openstack_project/files/jenkins_job_builder/config/manuals.yaml11:11
ildikoveglynn: I mean this one: gate-ceilometer-docs11:12
eglynnildikov: ... I'd suspect that that's defined as the generic gate-{name}-docs in the python-jobs11:16
eglynnildikov: ... i.e. https://github.com/openstack-infra/config/blob/master/modules/openstack_project/files/jenkins_job_builder/config/python-jobs.yaml#L24411:17
eglynnildikov: ... so follow on to https://github.com/openstack-infra/config/blob/master/modules/openstack_project/files/jenkins_job_builder/config/python-jobs.yaml#L14811:17
eglynnildikov: ... that what you're looking for?11:17
ildikoveglynn: yes, I guess that's the one11:22
eglynnildikov: coolio11:22
ildikoveglynn; tnx, I try to start working on the docs stuff, so I thought to fix my doc gate job bug first, hopefully it's not that hard11:22
eglynnildikov: great, thank you!11:23
eglynnildikov: BTW do you think the j1 targetting is reasonable for the docs work item?11:23
ildikoveglynn: I started to check the install guide in the manuals, it looks good actually, I will try to figure out where to mention our db related part in the dev docs, but I think after that the manual install section could be removed from our developer doc11:23
ildikoveglynn: I do not have the schedule opened, what is the date of j1?11:24
eglynnildikov: excellent ... so the "manual install" goes, but the "devstack install11:24
eglynn" stays?11:24
eglynnildikov: ... j1 == June 12th11:24
ildikoveglynn: yes, the devstack part is development related, so it should stay11:25
eglynnildikov: agreed :)11:25
eglynnildikov: BTW June 12th is the day that ttx cuts the actual milestone tag, IIRC11:25
ildikoveglynn: hmm, maybe not everything will fit into j1, but I will try my best11:25
ildikoveglynn: I need to have some discussion with the doc team that what is the most efficient way of contributing to the manuals repo11:26
eglynnildikov: cool, thanks11:26
ildikoveglynn: I do not know how to set up the docbook env, but I will ask them11:27
eglynnildikov: ... lemme know if there's anything I can do to help in terms of contrib'ing content (e.g. the RDO quick start guide)11:27
ildikoveglynn: ... my plan is to transform the RDO guide into an admin guide, I think I found the place for it in the manuals repo and check the install part and remove our manual install section11:28
*** 20WAAGQA1 is now known as shadower11:28
ildikoveglynn: ... after this, we can check that what else should be done11:28
eglynnildikov: excellent, thank you!11:28
eglynnildikov: BTW how did the OpenStack CEE day go yesterday?11:29
ildikoveglynn: ... I will update the google doc with the short summary of this plan and mark it as maybe possible to j111:29
eglynnildikov: awesome! thanks!11:29
ildikoveglynn: it was ok, that 40 mins was a bit short for our workshop unfortunately11:30
*** safchain has quit IRC11:31
eglynnildikov: yeah 40 mins seems like a lot, but is very tight once you start talking11:33
eglynn... so /me finds anyway11:33
ildikoveglynn: ... well, I knew that it will not be enough, but we had to try :)11:34
*** safchain has joined #openstack-ceilometer11:36
eglynnidegtiarov: do we need *both* https://blueprints.launchpad.net/ceilometer/+spec/hbase-events-structure11:40
eglynnidegtiarov: ... and also https://blueprints.launchpad.net/ceilometer/+spec/hbase-events-feature ?11:40
eglynnidegtiarov: ... the one seems to duplicate the other11:41
*** Alexei_9871 has joined #openstack-ceilometer11:44
eglynnidegtiarov: ... I'm gonna untarget hbase-events-structure from juno-1 for now11:45
eglynnidegtiarov: ... could you propose a detailed spec for hbase-events-feature to https://review.openstack.org/#/q/project:openstack/ceilometer-specs,n,z ?11:45
idegtiaroveglynn: now we don't need *both* purpose of first bp that you have mentioned is done in frame of second -- the "main" one11:46
idegtiaroveglynn: I am on my way to prepare detailed spec for hbase-events-feature11:48
eglynnidegtiarov: ... thank you sir! if you reckon that the second BP is also needed, then I think a spec would be needed for that too that clearly11:50
*** nati_uen_ has joined #openstack-ceilometer11:55
*** _nadya_ has joined #openstack-ceilometer11:55
*** nati_uen_ has quit IRC11:59
*** _nadya__ has joined #openstack-ceilometer12:00
*** _nadya_ has quit IRC12:00
*** eglynn is now known as eglynn-lunch12:01
idegtiaroveglynn:  ok!12:01
*** IvanBerezovskiy has joined #openstack-ceilometer12:09
*** jdob has joined #openstack-ceilometer12:33
*** thomasem has joined #openstack-ceilometer12:35
*** pnascimento has joined #openstack-ceilometer12:41
*** raymondr has joined #openstack-ceilometer12:50
*** eglynn-lunch has quit IRC12:50
*** rbowen has joined #openstack-ceilometer12:53
*** nati_ueno has joined #openstack-ceilometer12:55
*** heyongli has joined #openstack-ceilometer12:57
*** nati_ueno has quit IRC13:00
*** promulo has quit IRC13:03
*** raymondr has quit IRC13:04
*** julim has joined #openstack-ceilometer13:05
*** thomasem has quit IRC13:07
*** psharma has quit IRC13:12
*** changbl has quit IRC13:13
*** jmckind has joined #openstack-ceilometer13:13
*** thomasem has joined #openstack-ceilometer13:17
*** _nadya__ has quit IRC13:21
*** eglynn has joined #openstack-ceilometer13:22
*** jaypipes has joined #openstack-ceilometer13:22
*** ali-mohsin_ has joined #openstack-ceilometer13:30
*** _nadya_ has joined #openstack-ceilometer13:32
pnascimentoeglynn: On the email about Unit Testing, you said it: "Similarly, most of the test coverage is inherited from the test superclass, ceilometer.tests.agentbase.BaseAgentManagerTestCase."13:35
pnascimentoeglynn: If I understand it right, you are telling me that, if I create a class that inherits from ceilometer.tests.agentbase.BaseAgentManagerTestCase, I inherit some tests, that already asserts common things between agent, like polling samples or pipeline related actions?13:35
eglynnpnascimento: yep, that I did13:35
eglynnpnascimento: one sec13:35
pnascimentoeglynn: of course :)13:36
*** _nadya_ has quit IRC13:37
eglynnpnascimento: consider for example this testcase https://github.com/openstack/ceilometer/blob/master/ceilometer/tests/agentbase.py#L26813:37
DinaBelovafolks, fyi - http://blog.zhaw.ch/icclab/ceilometer-performance-issues/13:38
DinaBelovafunny small moment)13:38
eglynnpnascimento: that's a test method defined on the BaseAgentManagerTestCase class, that any subclasses will inherit13:39
eglynnpnascimento: how ever it won't be exactly the *same* test logic running in each of the subclasses of BaseAgentManagerTestCase13:39
eglynnpnascimento: why?13:39
eglynnpnascimento: ... because the the setUp method on the superclass calls out to the abstract method create_manager()13:40
eglynnpnascimento: ... that abstract method is defined on each subclass to instantiate an agent manager class *specific to the current test*13:41
eglynnpnascimento: ... have you read the code that I'm talking about above? ^^^13:41
eglynnpnascimento: ... does any of that make sense to you?13:41
pnascimentoeglynn: about this "that's a test method defined on the BaseAgentManagerTestCase class, that any subclasses will inherit how ever it won't be exactly the *same* test logic running in each of the subclasses of BaseAgentManagerTestCase", I understand that, it's basic inheritance software pattern :)13:44
eglynnpnascimento: ... eh, ok then13:45
eglynnpnascimento: ... so ...13:45
eglynnpnascimento: ... what exactly are you asking me in that case?13:45
pnascimentoeglynn: if I understand what have you said now, you basically are agreeing with me :)13:45
eglynnpnascimento: OK, I see13:45
pnascimentoeglynn: I admit, with my english level, I may have confused you a little bit :P13:46
pnascimentoeglynn: ... so, in a simple way13:46
eglynnpnascimento: no worries, I misunderstood your initial statement on IRC just now to mean "I didn't fully understand that part of your email"13:47
*** prad_ has joined #openstack-ceilometer13:47
*** ali-mohsin_ has quit IRC13:47
pnascimentoeglynn: ..., I have a simple doubt, will inheritance of BaseAgentManagerTestCase on my TestRunTasks, will execute the tests defined on the parent(BaseAgentManagerTestCase), for example, test_get_samples, even that they are different agent instance because the setUp()?13:49
pnascimentoeglynn: (I tried to simplify :P)13:49
*** jmckind has quit IRC13:49
eglynnpnascimento: yes, that's exactly what I was trying to example above13:51
eglynnpnascimento: ... what is your doubt about it?13:51
*** fzhaw has joined #openstack-ceilometer13:53
eglynnpnascimento: yes, that's exactly what I was trying to *explain* above13:53
*** promulo has joined #openstack-ceilometer13:53
openstackgerritPradeep Kilambi proposed a change to openstack/ceilometer-specs: Spec for metering LBaaS  https://review.openstack.org/9495813:53
*** igor_ has quit IRC13:54
fzhawHi all. A question on memory_usage with libvirt, I saw in recent commits that there is now a way to retrieve this metric for libvirt nodes (whereas before was only max memory allowed), in which version of Ceilometer will it be released? Or is it already there but not on the docs?13:55
pnascimentoeglynn: so, now you have clarified a lot my thoughts on that subject! :) the doubt was exactly, I never used inheritance on Unit Tests and that's I was confused about that, turns out behaves just like "normal" inheritance :P13:55
eglynnpnascimento: yep, they are just classes after all, even if they represent testcases13:56
pnascimentoeglynn: yep :P my bad :P13:56
*** nati_ueno has joined #openstack-ceilometer13:56
*** joshw42 has quit IRC13:57
*** terriyu has joined #openstack-ceilometer13:59
*** nati_ueno has quit IRC14:00
*** vrovachev has joined #openstack-ceilometer14:02
*** raymondr has joined #openstack-ceilometer14:04
*** gordc has joined #openstack-ceilometer14:06
*** heyongli has quit IRC14:07
*** terriyu has quit IRC14:08
*** cdent_ has joined #openstack-ceilometer14:11
openstackgerritArtur Svechnikov proposed a change to openstack/python-ceilometerclient: Add methods to resource classes  https://review.openstack.org/9155414:12
vrovachevgordc: Hey, Nadya asked me to write you that I'm currently testing a performance ceilometer :)14:13
gordcvrovachev: cool. i'd be interested to see results... let me know if you need anything from me.14:14
*** cdent has quit IRC14:14
*** cdent_ is now known as cdent14:14
openstackgerritPradeep Kilambi proposed a change to openstack/ceilometer-specs: Spec for metering FWaaS  https://review.openstack.org/9577914:18
*** ddieterly has joined #openstack-ceilometer14:20
*** _cjones_ has joined #openstack-ceilometer14:21
fzhawReposting just in case someone might know :) :  A question on the memory.usage metric with libvirt, I saw in recent commits that there is now a way to retrieve this metric for libvirt nodes (whereas before was only total memory allowed), in which version of Ceilometer will it be released? Or is it already there but not on the docs?14:24
*** _cjones_ has quit IRC14:25
vrovachevgordc: of course :)14:27
*** sileht has quit IRC14:28
openstackgerritPradeep Kilambi proposed a change to openstack/ceilometer-specs: Spec for metering VPNaaS  https://review.openstack.org/9578114:29
openstackgerritPradeep Kilambi proposed a change to openstack/ceilometer-specs: Spec for metering FWaaS  https://review.openstack.org/9577914:30
*** sileht has joined #openstack-ceilometer14:30
*** terriyu has joined #openstack-ceilometer14:30
gordcfzhaw: this? https://review.openstack.org/#/c/90498/14:35
*** sileht has quit IRC14:37
fzhawgordc: yep, exactly14:38
gordcfzhaw: seems to be wip. so it isn't available yet. might want to check with Rui to see if it's still being worked on...14:39
*** sileht has joined #openstack-ceilometer14:39
fzhawgordc: thanks, will contact then... could be useful as part of a decision engine we are working on here14:40
gordcfzhaw: cool cool, i'd give you irc contact but i'm not sure what their irc handle is.14:41
*** pnascimento has quit IRC14:43
*** pnascimento has joined #openstack-ceilometer14:45
openstackgerritPradeep Kilambi proposed a change to openstack/ceilometer: Metering LoadBalancer as a Service  https://review.openstack.org/9578414:49
*** sileht has quit IRC14:49
*** sileht has joined #openstack-ceilometer14:51
ildikovfzhaw: this memory_usage will not be implemented for a while14:54
ildikovfzhaw: there are too many limitations to calculate actual memory usage of a guest so in most cases it's not actually possible14:55
ildikovfzhaw: so the final decision was to suspend that BP and implementation14:56
*** nati_ueno has joined #openstack-ceilometer14:57
*** zqfan has joined #openstack-ceilometer14:59
*** nati_ueno has quit IRC15:01
*** cmart has joined #openstack-ceilometer15:03
*** kun_huan_ has joined #openstack-ceilometer15:04
*** kun_huang has quit IRC15:05
*** aviau has quit IRC15:09
pnascimentoeglynn: as said on: https://docs.python.org/3/library/unittest.mock.html#unittest.mock.patch it os supposed to add an extra argument when argument 'new' is omitted, on https://github.com/openstack/ceilometer/blob/master/ceilometer/tests/compute/test_manager.py#L53 isn't any extra argument on setUp(...), how so?15:12
pnascimentoeglynn: how this @mock.patch(...) works on Ceilometer context?15:12
eglynnpnascimento: on a call, but briefly in any case ...15:14
eglynnpnascimento: note https://github.com/openstack/ceilometer/blob/master/ceilometer/tests/compute/test_manager.py#L2015:15
eglynnpnascimento: reads "import mock"15:15
eglynnpnascimento: not "import unitest.mock as mock"15:15
*** kun_huan_ has quit IRC15:16
eglynnpnascimento: here's the docco you need to look at IIRC http://www.voidspace.org.uk/python/mock/patch.html15:16
*** kun_huang has joined #openstack-ceilometer15:17
eglynnpnascimento: so it's this line "If new is omitted, then the target is replaced with a MagicMock. If patch is used as a decorator and new is omitted, the created mock is passed in as an extra argument to the decorated function. If patch is used as a context manager the created mock is returned by the context manager."15:17
eglynnpnascimento: ^^^ that's whats confusing you?15:17
pnascimentoeglynn: after reading the link I gave you... I was capable to understand .patch(), but the "missing" argument, that's what confused me, because they say "(...) the created mock is passed in as an extra argument to the decorated function"15:20
pnascimentoeglynn: ... and on .setUp(self) there is no other EXTRA argument :S15:20
eglynnpnascimento: the extra arg is not needed in this case, as the test doesn't IIUC need to make any assertions about that mock15:21
eglynnpnascimento: ... it just needs to exist15:21
pnascimentoeglynn: ok then ;)15:21
eglynnpnascimento: whereas consider this https://github.com/openstack/ceilometer/blob/master/ceilometer/tests/pipeline_base.py#L134215:22
eglynnpnascimento: in that case the test does need to ask questions about the mock15:22
eglynnpnascimento: specifically here: https://github.com/openstack/ceilometer/blob/master/ceilometer/tests/pipeline_base.py#L135115:22
eglynnpnascimento: ... hence the extra arg is defined in that case15:22
eglynnpnascimento: ... forget the strict Java rules about such things, this is the wild west of python ;)15:23
*** `jpg has joined #openstack-ceilometer15:23
eglynnpnascimento: ... gotta get back to my call, ttyl15:24
pnascimentoeglynn: I'm start to think, that is my major enemy :P but, thank you a lot for you help! :)15:24
*** aviau has joined #openstack-ceilometer15:24
*** ddieterly has quit IRC15:25
openstackgerritPradeep Kilambi proposed a change to openstack/ceilometer-specs: Spec for metering FWaaS  https://review.openstack.org/9577915:26
*** kun_huang has quit IRC15:32
*** fnaval has joined #openstack-ceilometer15:39
*** fzhaw has quit IRC15:39
*** _cjones_ has joined #openstack-ceilometer15:54
*** shakamunyi has joined #openstack-ceilometer15:56
openstackgerritPradeep Kilambi proposed a change to openstack/ceilometer-specs: Spec for metering LBaaS  https://review.openstack.org/9495815:57
*** ddieterly has joined #openstack-ceilometer15:59
*** alexpilotti has joined #openstack-ceilometer15:59
*** IvanBerezovskiy has left #openstack-ceilometer15:59
*** Ruetobas has quit IRC16:01
*** Ruetobas has joined #openstack-ceilometer16:03
*** Ruetobas has quit IRC16:08
*** Ruetobas has joined #openstack-ceilometer16:14
*** nati_ueno has joined #openstack-ceilometer16:14
*** idegtiarov has quit IRC16:16
*** pnascimento has quit IRC16:32
*** nacim_ has quit IRC16:42
*** safchain has quit IRC16:42
*** yjiang5 has joined #openstack-ceilometer16:51
openstackgerritMehdi Abaakouk proposed a change to openstack/ceilometer: Don't rely on oslomsg configuration options  https://review.openstack.org/9583016:53
*** nati_ueno has quit IRC16:53
*** Alexei_9871 has quit IRC16:56
*** nati_ueno has joined #openstack-ceilometer16:56
*** admin0 has quit IRC16:57
*** ildikov has quit IRC16:57
*** aviau has quit IRC16:57
*** aviau has joined #openstack-ceilometer16:58
*** _nadya_ has joined #openstack-ceilometer17:00
*** zqfan has quit IRC17:02
*** _nadya_ has quit IRC17:09
*** isviridov is now known as isviridov|away17:09
*** _nadya_ has joined #openstack-ceilometer17:09
*** aviau has quit IRC17:11
*** aviau has joined #openstack-ceilometer17:12
vrovachevgordc: I finished perfomanse testing with applied patch-sets: https://review.openstack.org/#/c/94483/ and https://review.openstack.org/#/c/94682/117:17
vrovachevgordc: Tests in patch-set successful:17:17
vrovachevgordc: Logs in testr: http://paste.openstack.org/show/81735/17:17
*** ali-mohsin has joined #openstack-ceilometer17:18
*** ali-mohsin has quit IRC17:18
*** aviau has quit IRC17:19
DinaBelovavrovachev - it's really cool :)17:19
*** aviau has joined #openstack-ceilometer17:20
*** _nadya_ has quit IRC17:22
*** changbl has joined #openstack-ceilometer17:25
amalagonhi DinaBelova o/ if you have a few minutes can I ask you about gnocchi setup?17:30
DinaBelovaamalagon, well, you may try to))17:30
DinaBelovaI have started gnocchi, but well - I suppose it might not work yet17:30
DinaBelovabut i'll try to help17:31
amalagonDinaBelova: :)) thanks17:31
amalagonDinaBelova: I installed gnocchi and swift in a stand alone install17:31
DinaBelovaokay, cool17:31
amalagonDinaBelova: but I get errors when running ./run_tests.sh17:31
*** nati_ueno has quit IRC17:31
* amalagon finds notes17:31
DinaBelovaI did not try to run this script, one moment17:32
DinaBelovaokay, postgreSQL :) it's failing))17:33
DinaBelova*no* postegresql installed I mean17:33
amalagonyay errors! I found I had to install libpq-dev as well - not sure but maybe related to that17:34
DinaBelovayes, you need to definately17:34
amalagonI also had to install build-essential17:35
DinaBelovabuild-essential libx11-dev libasound2-dev - I've installed these ones17:36
DinaBelovaamalagon, btw - what conf file are you using?17:38
*** alexpilotti has quit IRC17:39
amalagonDinaBelova: er, so I might be misunderstanding conf files - I used setup.cfg when trying to run gnocchi-api but I think you're asking about something else?17:40
DinaBelovaabout gnocchi.conf17:40
DinaBelovaor you did not try to run the service itself yet?17:40
amalagonDinaBelova: ahh, no..lemme try that17:41
*** alexpilotti has joined #openstack-ceilometer17:41
DinaBelovabtw - after postresql installed i have tests run, but not passing due to errors like "MismatchError: '2014-05-20 12:00:45.856219' != u'2014-05-20 10:00:45.856219'"17:41
DinaBelovaI guess it's some time stuff - possibly on my os :-\17:42
*** nati_ueno has joined #openstack-ceilometer17:42
DinaBelovaamalagon, I used this one: http://paste.openstack.org/show/IXsDRLBFt6xmRw3NoqeX/ - almost default one, but just listed to see the opts17:43
*** eglynn has quit IRC17:44
amalagonDinaBelova: thanks! again, I think I don't understand conf files, but did you find gnocchi.conf in the gnocchi directory or did you make it yourself?17:44
DinaBelovamyself, in the /etc/gnocchi/gnocchi.cong file17:45
DinaBelovait's the default dir for OS projects - /etc/<project>/<project>.conf17:45
DinaBelovain these files we usually list the opts you might see in the code like: CONF.register_opts(opts, group)17:46
DinaBelovathey have default values, but we can change this in these files17:46
DinaBelovaand if you installed gnocchi correctly it'll start api service using "gnocchi-api --debug --config-file /etc/gnocchi/gnocchi.conf" or "gnocchi-api --debug"17:47
amalagonDinaBelova: I see! thanks very much for explaining this. I'll try this17:47
amalagonDinaBelova:  also btw I'm not getting the Mismatch error; mine is an Importerror, not sure what to make of that17:47
DinaBelovaI usually use explicit --config-file /etc/gnocchi/gnocchi.conf to be sure where I got these oprtion from17:47
DinaBelovaplease paste your errors to the paste.openstack.org17:48
DinaBelovaI'll try to look on them)17:48
amalagon:)) thanks17:48
cmartDinaBelova: and what about changing the db types? I want to use mongodb, and I changed the /etc/ceilometer/ceilometer.conf (based on this: http://docs.openstack.org/developer/ceilometer/install/manual.html), but after running a ./stack, it doesn´t work17:48
cmartI mean, it's still using mysql as default17:48
DinaBelovawell - the only one way to understand the problem is to debug...17:49
DinaBelovaoh, stop17:49
DinaBelovadid you used ./stack.sh?17:49
DinaBelovait's rewriting all confs, as far as I remember17:49
DinaBelovais your change still in place?17:49
DinaBelovaplease use ./unstack and ./rejoin, or just restart manually in the screen ceilo processes17:50
cmartOK. Thanks!17:50
amalagonDinaBelova: here's the import error ~ http:paste.openstack.org/show/81753/17:58
amalagonDinaBelova: sorry - http://paste.openstack.org/show/81753/17:59
DinaBelovaehm... in which folder are you running the run_tests?18:00
DinaBelovaamalagon ^^18:03
*** alexpilotti has quit IRC18:03
amalagonDinaBelova: in gnocchi18:04
DinaBelovaokay, strange18:04
amalagonI agree!18:04
DinaBelovai simply do not recognize where unittest.loader.ModuleImportFailure.openstack.common.db.sqlalchemy.test_migrations came from18:04
DinaBelovawell, it looks like some two modules concatenation in this error, never mind18:08
*** _nadya_ has joined #openstack-ceilometer18:09
DinaBelovadid you do smth with setup.cfg?18:09
*** Ruetobas has quit IRC18:09
DinaBelovaso thats like - you cloned git repo, ran run_tests and see this18:09
DinaBelovano changes to the setup files?18:10
amalagonyep, cloned git repo, installed dependencies, ran setup.py install, then ran run-tests18:10
amalagonno changes to setup files18:11
amalagonbesides running them I guess18:11
*** nati_ueno has quit IRC18:13
*** nati_ueno has joined #openstack-ceilometer18:14
*** Ruetobas has joined #openstack-ceilometer18:15
*** Ruetobas has quit IRC18:19
DinaBelovaamalagon, well18:22
DinaBelovareally have no idea18:22
DinaBelovapossibly you may reinstall the gnocchi... and try it one more time - although I'm not sure these two places are really connected here18:23
amalagonDinaBelova: ok, I'll try reinstalling gnocchi18:23
amalagonthank you tho for looking into this :))18:23
DinaBelovawell, I hope I would more helpful here18:24
amalagonno, this was helpful! I'll let you know if the reinstall fixes things18:24
*** Ruetobas has joined #openstack-ceilometer18:25
DinaBelovaamalagon, thanks, will be glad to know if it'll help)18:25
*** inc0 has quit IRC18:33
gordccmart: if you set CEILOMETER_BACKEND=mongodb in your localrc file it won't default to mysql.18:55
gordcvrovachev: sorry about that. didnt' see your message. good to see the compute test pass with patches.18:56
cmartgordc: yes! I did that and it worked! (I also try by modifying the lib/ceilometer and it also worked) thansk!18:58
*** nati_ueno has quit IRC18:58
gordccmart: yeah, that works too. i'm a fan of brute force method as well :)18:58
*** nati_ueno has joined #openstack-ceilometer18:59
cmartgordc: hahaah! Sometimes it's the fastest (and the best) way :D18:59
gordccmart: agreed.18:59
cmartgordc: one question regarding mongo and ceilometer. Once my ./stack finishes, I should have a mongo server up and running, right?19:04
cmart..which means that no need to run mongod in a separated console19:04
openstackgerritMartin Geisler proposed a change to openstack/ceilometer: Use correct "coding" file variable to set file encoding  https://review.openstack.org/9586219:06
openstackgerritMartin Geisler proposed a change to openstack/ceilometer: Correct Emacs file variable specification  https://review.openstack.org/9586319:06
gordccmart: you shouldn't need to... i don't explicit start it (using ubuntu)19:08
cmartOK.. thanks!19:08
openstackgerritgordon chung proposed a change to openstack/ceilometer: refactor sql backend to improve write speed  https://review.openstack.org/9448319:15
*** raymondr has quit IRC19:18
*** raymondr_ has joined #openstack-ceilometer19:19
openstackgerritMartin Geisler proposed a change to openstack/pycadf: Use Emacs-friendly file variable to set file encoding  https://review.openstack.org/9586919:27
*** _nadya_ has quit IRC19:31
*** eglynn-afk has joined #openstack-ceilometer19:31
*** nati_ueno has quit IRC19:33
*** melodous has joined #openstack-ceilometer19:34
*** melodous_away has quit IRC19:35
*** terriyu has quit IRC19:38
*** nati_ueno has joined #openstack-ceilometer19:50
cmartgordc, are you around?19:51
gordccmart: done.19:52
openstackgerritMartin Geisler proposed a change to openstack/python-ceilometerclient: Use Emacs-friendly file variable to set file encoding  https://review.openstack.org/9588019:58
*** nati_ueno has quit IRC20:00
*** nati_ueno has joined #openstack-ceilometer20:12
*** dhellman_ has joined #openstack-ceilometer20:15
openstackgerritPradeep Kilambi proposed a change to openstack/ceilometer: Metering LoadBalancer as a Service  https://review.openstack.org/9578420:16
*** r0j4z0 has joined #openstack-ceilometer20:20
*** boris-42 has quit IRC20:21
*** boris-42 has joined #openstack-ceilometer20:21
r0j4z0hi all20:29
*** jdob has quit IRC20:50
*** eglynn-afk is now known as eglynn20:53
*** julim has quit IRC21:01
*** alexpilotti has joined #openstack-ceilometer21:01
*** `jpg has quit IRC21:03
*** nati_ueno has quit IRC21:08
*** chrisf has quit IRC21:10
*** admin0 has joined #openstack-ceilometer21:10
*** admin0 has quit IRC21:11
*** nati_ueno has joined #openstack-ceilometer21:11
*** chrisf has joined #openstack-ceilometer21:11
*** admin0 has joined #openstack-ceilometer21:11
*** cmart has quit IRC21:12
*** eglynn has quit IRC21:15
*** dhellman_ has quit IRC21:19
*** cdent has quit IRC21:20
*** promulo has quit IRC21:20
*** prad_ has quit IRC21:26
*** admin0 has quit IRC21:28
*** admin0 has joined #openstack-ceilometer21:29
-openstackstatus- NOTICE: Zuul is offline due to an operational issue; ETA 2200 UTC.21:31
*** ChanServ changes topic to "Zuul is offline due to an operational issue; ETA 2200 UTC."21:31
*** nati_ueno has quit IRC21:37
*** eglynn has joined #openstack-ceilometer21:43
*** nati_ueno has joined #openstack-ceilometer21:43
eglynnamalagon: does my response here make sense?21:43
eglynnamalagon: https://ask.openstack.org/en/question/30649/running-gnocchi-api-with-swift-all-in-one-install-fails/21:43
eglynnamalagon: TL;DR setup.cfg is not the in format expected, nor does it contain the kind of config settings needed21:44
*** eglynn has quit IRC21:45
*** admin0 has quit IRC21:54
*** shakamunyi has quit IRC22:03
*** raymondr_ has quit IRC22:05
*** nati_ueno has quit IRC22:05
*** ChanServ changes topic to "OpenStack Telemetry (Ceilometer) | https://wiki.openstack.org/wiki/Ceilometer"22:08
-openstackstatus- NOTICE: Zuul is started and processing changes that were in the queue when it was stopped. Changes uploaded or approved since then will need to be re-approved or rechecked.22:08
*** ildikov has joined #openstack-ceilometer22:08
*** shakamunyi has joined #openstack-ceilometer22:10
*** ddieterly has quit IRC22:23
*** nati_ueno has joined #openstack-ceilometer22:32
*** `jpg has joined #openstack-ceilometer22:32
*** nati_uen_ has joined #openstack-ceilometer22:42
*** nati_ueno has quit IRC22:42
*** fnaval has quit IRC22:45
*** `jpg has quit IRC22:45
*** changbl has quit IRC22:47
*** gordc has quit IRC22:52
*** promulo has joined #openstack-ceilometer22:52
r0j4z0hi all23:32
r0j4z0someone online?23:32
r0j4z0i dont understand how ceilometer process the notifications and convert it to events...23:33
r0j4z0i cant get anything from ceilometer event-list23:33
r0j4z0also i cant view anything on the anotification log or collector...23:34
r0j4z0i have /etc/ceilometer/event_definitions.yaml but is like ceilometer dont use it...23:34
r0j4z0notification_driver = ceilometer.compute.nova_notifier23:35
r0j4z0also i have these 2 notification drivers on nova.conf23:35
r0j4z0the central agent is working fine, i can view all the meters without problem23:37
r0j4z0any idea?23:37
r0j4z0im using devstack23:38
r0j4z0the compute agent is working also23:40
r0j4z0i get the data and is pipelined correctly23:40
*** ddieterly has joined #openstack-ceilometer23:49
*** ddieterly has quit IRC23:53

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!