| *** promulo has joined #openstack-ceilometer | 00:05 | |
| *** sdake_1 has quit IRC | 00:15 | |
| *** matsuhashi has joined #openstack-ceilometer | 00:26 | |
| *** openstackgerrit has joined #openstack-ceilometer | 00:30 | |
| *** gordc has quit IRC | 00:45 | |
| *** matsuhashi has quit IRC | 00:47 | |
| *** matsuhashi has joined #openstack-ceilometer | 00:48 | |
| *** sdake has joined #openstack-ceilometer | 00:48 | |
| *** sdake has joined #openstack-ceilometer | 00:48 | |
| *** matsuhashi has quit IRC | 00:52 | |
| *** matsuhashi has joined #openstack-ceilometer | 00:57 | |
| openstackgerrit | liusheng proposed a change to openstack/ceilometer: Fix the floatingip pollster https://review.openstack.org/83676 | 01:00 |
|---|---|---|
| *** jaypipes has quit IRC | 01:08 | |
| openstackgerrit | liusheng proposed a change to openstack/ceilometer: Fix the return of statistic with getting no sample https://review.openstack.org/80467 | 01:09 |
| openstackgerrit | liusheng proposed a change to openstack/ceilometer: Fix the floatingip pollster https://review.openstack.org/83676 | 01:22 |
| *** nosnos has joined #openstack-ceilometer | 01:47 | |
| *** changbl has joined #openstack-ceilometer | 01:47 | |
| *** flwang has quit IRC | 02:17 | |
| *** flwang has joined #openstack-ceilometer | 02:55 | |
| *** flwang has quit IRC | 03:03 | |
| *** flwang has joined #openstack-ceilometer | 03:04 | |
| openstackgerrit | Jia Dong proposed a change to openstack/ceilometer: Sync rpc modules from oslo-incubator https://review.openstack.org/84609 | 03:08 |
| *** matsuhashi has quit IRC | 03:21 | |
| *** nosnos has quit IRC | 03:37 | |
| *** _nadya_ has joined #openstack-ceilometer | 03:41 | |
| *** _nadya_ has quit IRC | 03:59 | |
| *** gordc has joined #openstack-ceilometer | 04:17 | |
| *** piyushmasrani has joined #openstack-ceilometer | 04:29 | |
| *** nosnos has joined #openstack-ceilometer | 04:31 | |
| *** matsuhashi has joined #openstack-ceilometer | 04:36 | |
| *** matsuhashi has quit IRC | 04:50 | |
| *** gordc has quit IRC | 04:55 | |
| *** matsuhashi has joined #openstack-ceilometer | 05:00 | |
| *** ildikov_ has quit IRC | 05:23 | |
| *** inc0_ has joined #openstack-ceilometer | 05:52 | |
| *** inc0_ has quit IRC | 05:53 | |
| *** inc0_ has joined #openstack-ceilometer | 05:53 | |
| openstackgerrit | Jenkins proposed a change to openstack/ceilometer: Imported Translations from Transifex https://review.openstack.org/84359 | 06:04 |
| *** mihgen has joined #openstack-ceilometer | 06:12 | |
| *** ildikov_ has joined #openstack-ceilometer | 06:13 | |
| *** shardy_afk is now known as shardy | 06:33 | |
| *** saju_m has joined #openstack-ceilometer | 06:39 | |
| *** tasdomas` has joined #openstack-ceilometer | 07:25 | |
| *** tasdomas has quit IRC | 07:25 | |
| *** saju_m has quit IRC | 07:42 | |
| *** boris-42 has quit IRC | 07:50 | |
| *** boris-42 has joined #openstack-ceilometer | 07:51 | |
| *** nacim has joined #openstack-ceilometer | 07:51 | |
| *** saju_m has joined #openstack-ceilometer | 07:52 | |
| *** piyushmasrani has quit IRC | 08:25 | |
| *** wayneeseguin has quit IRC | 08:28 | |
| *** wayneeseguin has joined #openstack-ceilometer | 08:28 | |
| *** yassine has joined #openstack-ceilometer | 08:38 | |
| *** eouti has joined #openstack-ceilometer | 08:39 | |
| *** saju_m has quit IRC | 08:48 | |
| *** saju_m has joined #openstack-ceilometer | 08:49 | |
| *** flwang has quit IRC | 09:11 | |
| *** saju_m has quit IRC | 09:24 | |
| openstackgerrit | Ilya Tyaptin proposed a change to openstack/ceilometer: Add unit test for checking right order of sample list https://review.openstack.org/84674 | 09:25 |
| *** saju_m has joined #openstack-ceilometer | 09:26 | |
| *** saju_m has quit IRC | 09:31 | |
| openstackgerrit | A change was merged to openstack/ceilometer: Disable oslo.messaging debug logs https://review.openstack.org/82671 | 09:32 |
| *** alexpilotti has joined #openstack-ceilometer | 09:34 | |
| *** admin0 has joined #openstack-ceilometer | 09:36 | |
| *** saju_m has joined #openstack-ceilometer | 09:42 | |
| *** matsuhashi has quit IRC | 09:50 | |
| *** matsuhashi has joined #openstack-ceilometer | 09:57 | |
| *** flwang has joined #openstack-ceilometer | 09:59 | |
| *** eglynn has joined #openstack-ceilometer | 10:16 | |
| eglynn | jay_t_c: around? | 10:17 |
| eglynn | jay_t_c_2: ^^^ | 10:17 |
| *** aswadrangnekar has joined #openstack-ceilometer | 10:17 | |
| eglynn | jay_t_c_2: finally I got around to looking at your heat problem, but only had a spare RHEL VM to play with | 10:18 |
| aswadrangnekar | Hi | 10:19 |
| eglynn | jay_t_c_2: ... so I used the latest incarnation of the havana-derived Red Hat opensatck distro (as opposed to the canonical distro for ubuntu) | 10:19 |
| eglynn | jay_t_c_2: ... all went swimmingly with your template | 10:19 |
| eglynn | jay_t_c_2: ... the alarm_actions as expected are a pre-signed http URL | 10:20 |
| aswadrangnekar | I am trying to use raw_request method from ceilometer client http.py file | 10:20 |
| eglynn | jay_t_c_2: ... [u'http://172.16.12.60:8000/v1/signal/arn%3Aopenstack%3Aheat%3A%3A1088d2...'] | 10:20 |
| eglynn | jay_t_c_2: ... so for some reason looks like the alarm action is being truncated somewhere in your case (i.e. either on the heat or ceilo side) | 10:21 |
| eglynn | jay_t_c_2: ... what storage driver are you using? | 10:21 |
| aswadrangnekar | but I an not able to access this method after initializing the client | 10:22 |
| aswadrangnekar | is there a usecase where python ceilometer client is used by other projects? | 10:24 |
| eglynn | aswadrangnekar: yes | 10:24 |
| eglynn | aswadrangnekar: heat | 10:24 |
| eglynn | aswadrangnekar: and horizon | 10:25 |
| aswadrangnekar | eglynn: thanks I will take a look | 10:25 |
| eglynn | aswadrangnekar: BTW ceilometer itself used python-ceilometerclient | 10:25 |
| eglynn | aswadrangnekar: (in the alarm-evaluator) | 10:25 |
| *** ildikov_ has quit IRC | 10:26 | |
| eglynn | aswadrangnekar: https://github.com/openstack/ceilometer/blob/master/ceilometer/alarm/service.py#L72 | 10:26 |
| eglynn | aswadrangnekar: https://github.com/openstack/heat/blob/master/heat/engine/clients.py#L250 | 10:27 |
| aswadrangnekar | eglynn: thank for the pointer :) , let me have a look at it | 10:27 |
| eglynn | aswadrangnekar: np! ... BTW the horizon client is instantiated here: https://github.com/openstack/horizon/blob/master/openstack_dashboard/api/ceilometer.py#L257 | 10:29 |
| *** kun_huang has joined #openstack-ceilometer | 10:41 | |
| *** matsuhashi has quit IRC | 11:00 | |
| *** nosnos has quit IRC | 11:04 | |
| *** openstackgerrit has quit IRC | 11:08 | |
| *** openstackgerrit has joined #openstack-ceilometer | 11:08 | |
| *** matsuhashi has joined #openstack-ceilometer | 11:16 | |
| *** julim has joined #openstack-ceilometer | 11:16 | |
| *** kun_huang has quit IRC | 11:44 | |
| *** admin0 has quit IRC | 11:52 | |
| *** mengxd has joined #openstack-ceilometer | 12:04 | |
| *** jay_t_c_2 has quit IRC | 12:09 | |
| *** admin0 has joined #openstack-ceilometer | 12:12 | |
| *** saju_m has quit IRC | 12:17 | |
| *** admin0 has quit IRC | 12:22 | |
| *** admin0 has joined #openstack-ceilometer | 12:26 | |
| *** kun_huang has joined #openstack-ceilometer | 12:26 | |
| *** jdob has joined #openstack-ceilometer | 12:28 | |
| *** vrovachev has joined #openstack-ceilometer | 12:29 | |
| *** eouti has quit IRC | 12:32 | |
| *** admin0 has quit IRC | 12:35 | |
| *** ildikov_ has joined #openstack-ceilometer | 12:36 | |
| *** admin0 has joined #openstack-ceilometer | 12:40 | |
| *** promulo has quit IRC | 12:41 | |
| *** saju_m has joined #openstack-ceilometer | 12:47 | |
| *** kun_huang has quit IRC | 12:50 | |
| *** kun_huang has joined #openstack-ceilometer | 12:54 | |
| *** gordc has joined #openstack-ceilometer | 12:58 | |
| *** matsuhashi has quit IRC | 13:02 | |
| *** matsuhashi has joined #openstack-ceilometer | 13:02 | |
| *** prad_ has joined #openstack-ceilometer | 13:04 | |
| eglynn | folks, FYI there's an alternative summit being organized in APAC | 13:05 |
| eglynn | ... perhaps we should think of submitting some ceilo proposals? | 13:05 |
| eglynn | ... for details see http://closedstack.org/ | 13:05 |
| eglynn | ;) | 13:05 |
| *** admin0_ has joined #openstack-ceilometer | 13:06 | |
| *** matsuhashi has quit IRC | 13:07 | |
| eglynn | ... /me is a day late and a dollar shy ;) | 13:07 |
| *** admin0 has quit IRC | 13:08 | |
| *** admin0_ is now known as admin0 | 13:08 | |
| *** matsuhashi has joined #openstack-ceilometer | 13:10 | |
| eglynn | gordc: got time for a super-quick question? | 13:11 |
| eglynn | gordc: about https://review.openstack.org/79962 | 13:11 |
| eglynn | gordc: so I was thinking of carrying this as a additional patch in the havana-derived RH openstack distro | 13:12 |
| eglynn | (... seeing as it's not on stable/havana and prolly not suitable for backport) | 13:12 |
| eglynn | gordc: ... but remind me of problematic aspect of this discussed last week, it was a postgres-only issue right? | 13:13 |
| *** gordc has quit IRC | 13:24 | |
| *** promulo has joined #openstack-ceilometer | 13:24 | |
| *** admin0 has quit IRC | 13:35 | |
| *** saju_m has quit IRC | 13:36 | |
| *** admin0 has joined #openstack-ceilometer | 13:37 | |
| *** admin0 has quit IRC | 13:40 | |
| *** admin0 has joined #openstack-ceilometer | 13:41 | |
| *** admin0 has quit IRC | 13:45 | |
| *** admin0 has joined #openstack-ceilometer | 13:48 | |
| eglynn | jd__: do you know where we specify the mongodb version used by the scenario tests in the gate? | 13:49 |
| eglynn | ... or is it just latest available in whatever package repo? | 13:49 |
| jd__ | option #2 | 13:49 |
| jd__ | so the mongodb tests do not run on py27 because it's Oldbuntu | 13:49 |
| jd__ | only on py26 where CentOS is used and has a more recent MongoDB | 13:49 |
| eglynn | yeah py26 is the problem ... /me is trying to figure out this py26 specific fail on stable/havana | 13:50 |
| eglynn | http://logs.openstack.org/98/84098/2/check/gate-ceilometer-python26/abc5ad6/console.html | 13:50 |
| *** admin0 has quit IRC | 13:50 | |
| eglynn | ... can't reproduce it on RHEL tho' with mongodb 2.4.6 installed | 13:50 |
| jd__ | eglynn: looks like it's because 1 is different than 2 | 13:50 |
| jd__ | you're welcome. | 13:50 |
| jd__ | :-D | 13:50 |
| *** admin0 has joined #openstack-ceilometer | 13:50 | |
| eglynn | jd__: darn! why didn't I think of that ;) | 13:51 |
| jd__ | eglynn: replace self.assertEqual(len(resources), 2) by self.assertEqual(len(resources), len(resources)) and that should work | 13:51 |
| jd__ | you're welcome, AGAIN. | 13:51 |
| eglynn | jd__: LOL :) | 13:52 |
| jd__ | more seriously that's weird that it suddenty happens on stable | 13:52 |
| eglynn | jd__: ... well only for the patch that ups the "implicit" mongo requirement to 2.4 | 13:52 |
| jd__ | hm | 13:54 |
| jd__ | I don't think the nodes are different between py26 for stable and master | 13:55 |
| jd__ | maybe the patch cherry-pick misses something? :( | 13:55 |
| eglynn | jd__: ... to be fair, the patch also changes the resource collection somewhat, so not necessarily related to versioning | 13:55 |
| eglynn | jd__: ... yep, always poss that I screwed that up ;) | 13:55 |
| jd__ | yeah I don't think it's versioning at all | 13:55 |
| eglynn | jd__: cough, nothing to see here ;) | 13:56 |
| eglynn | jd__: ... yep bad cherry pick | 13:56 |
| eglynn | jd__: ... /me scurries off to fix | 13:56 |
| jd__ | eglynn: you know what? | 13:57 |
| eglynn | jd__: shoot | 13:57 |
| jd__ | YOU ARE WELCOME FOR THE THIRD TIME! | 13:57 |
| jd__ | ;) | 13:58 |
| eglynn | jd__: my belated but sincere thanks, good sir! | 13:58 |
| jd__ | hehe | 13:58 |
| *** aswadrangnekar has quit IRC | 14:04 | |
| *** mengxd has quit IRC | 14:09 | |
| *** admin0_ has joined #openstack-ceilometer | 14:09 | |
| *** admin0 has quit IRC | 14:10 | |
| *** admin0_ is now known as admin0 | 14:10 | |
| openstackgerrit | Nadya Privalova proposed a change to openstack/ceilometer: Make entities (Resource, User, Project) able to store lists https://review.openstack.org/78244 | 14:12 |
| *** inc0_ has quit IRC | 14:15 | |
| openstackgerrit | Nadya Privalova proposed a change to openstack/ceilometer: Make entities (Resource, User, Project) able to store lists https://review.openstack.org/78244 | 14:17 |
| *** aswadrangnekar has joined #openstack-ceilometer | 14:18 | |
| *** mriedem has joined #openstack-ceilometer | 14:21 | |
| *** mriedem has left #openstack-ceilometer | 14:22 | |
| *** htruta has joined #openstack-ceilometer | 14:24 | |
| htruta | hello, guys! i'm trying to get some ceilometer information in nova. Is that possible? can i get it through ceilometer client or the file generated by ceilometer? | 14:25 |
| *** admin0 has quit IRC | 14:26 | |
| *** aswadrangnekar has left #openstack-ceilometer | 14:29 | |
| *** nealph has joined #openstack-ceilometer | 14:30 | |
| *** admin0 has joined #openstack-ceilometer | 14:35 | |
| *** gordc has joined #openstack-ceilometer | 14:38 | |
| gordc | eglynn: sorry about that. internet cut out at my apartment. sent you an email. | 14:39 |
| eglynn | gordc: yeah thanks for that, I was just responding to your mail | 14:39 |
| eglynn | gordc: ... but easier to just say it on IRC | 14:40 |
| gordc | eglynn: ok. | 14:40 |
| eglynn | gordc: ... so in the meantime I was reminding myself of the last weekly meeting discussion and IIUC seems to be indicating that the issue impacts postgres only and *not* mysql | 14:40 |
| eglynn | http://eavesdrop.openstack.org/meetings/ceilometer/2014/ceilometer.2014-03-26-21.01.log.html | 14:40 |
| eglynn | ... e.g. <gordc> jd__: agreed. there's something wonky happening which doesn't let postgres work with multiple connection but let's mysql | 14:41 |
| eglynn | gordc: ^^^ is that in reference to the same multi-worker collector patch? | 14:41 |
| eglynn | gordc: ... or am I conflating two separate issues? | 14:41 |
| gordc | eglynn: right. so it's sort of two different issues. there's that item... | 14:41 |
| gordc | and then there's also the fact that you'll get deadlock/rollback errors in both sql backends as well even with the postgres patch. | 14:42 |
| openstackgerrit | Nadya Privalova proposed a change to openstack/ceilometer: Make entities (Resource, User, Project) able to store lists https://review.openstack.org/78244 | 14:43 |
| eglynn | gordc: so currently on master, what's the status in terms of brokeness against either mysql or postgres? | 14:43 |
| gordc | eglynn: https://review.openstack.org/#/c/80461/ this is the patch i have that resolves it but the real fix is to fix the data moel | 14:44 |
| gordc | eglynn: you should probably stick with a single worker... it's not committing fast enough with the above patch. | 14:44 |
| gordc | s/with the/without the/ | 14:45 |
| eglynn | gordc: so without https://review.openstack.org/80461 ... are *both* sqla DBs problematic on master with multiple collectors? | 14:46 |
| eglynn | (i.e. either multiple workers or multiple processes) | 14:46 |
| gordc | eglynn: i didn't see the issue as much (or at all) with multiple collector services... that said i didn't test as much that scenario. | 14:47 |
| gordc | eglynn: you'll definitely see it under decent loads if you spawn multiple workers under single service. | 14:48 |
| eglynn | gordc: sounds like we need to release-note this as a know issue with the sqla driver? | 14:49 |
| eglynn | gordc: ... i.e. instruct users not to spawn multiple worker threads in the sqla case | 14:49 |
| gordc | eglynn: yeah. we'll definitely need a backport at some point. | 14:50 |
| eglynn | gordc: k, let's discuss in more detail at the weekly meeting what needs to be explicitly captured in the icehouse release notes (prior to future backports onto stable/icehouse) | 14:52 |
| eglynn | gordc: ... in the meantime, sounds like your multi-worker patch is safe to carry in RHOS as we only support mongo | 14:52 |
| *** matsuhashi has quit IRC | 14:52 | |
| *** jmckind has joined #openstack-ceilometer | 14:52 | |
| gordc | eglynn: ok. give it a try just to confirm but all the issues i saw were sql related. | 14:54 |
| eglynn | gordc: cool, will do | 14:54 |
| gordc | eglynn: i actually won't be at meeting tomorrow (will be out of town) so i'll send you guys a note if i recall any other issues. | 14:54 |
| eglynn | gordc: great, thank you sir! | 14:55 |
| htruta | hello? anyone? | 14:57 |
| eglynn | htruta: anyone what? | 15:02 |
| nprivalova | eglynn, gordc, I need you thoughts about https://bugs.launchpad.net/ceilometer/+bug/1301371 | 15:08 |
| eglynn | nprivalova: so what's the story, same resource but different sources? | 15:10 |
| nprivalova | eglynn: same meters but different sources | 15:10 |
| eglynn | nprivalova: but meters are per-resource, right? | 15:11 |
| nprivalova | eglynn: yep | 15:11 |
| nprivalova | eglynn: let me find a short example... | 15:12 |
| eglynn | nprivalova: so "same meters but different sources" is equivalent to "individual resource(s) each associated with multiple sources"? | 15:12 |
| nprivalova | eglynn: if we have | 15:13 |
| nprivalova | eglynn: {'user_id': u'user-id-alternate', 'name': 'instance', 'resource_id': u'resource-id-alternate', 'source': 'test-2', 'project_id': u'project-id', 'type': 'cumulative', 'unit': ''} | 15:13 |
| nprivalova | eglynn: {'user_id': u'user-id-alternate', 'name': 'instance', 'resource_id': u'resource-id-alternate', 'source': 'test-3', 'project_id': u'project-id', 'type': 'cumulative', 'unit': ''} | 15:13 |
| nprivalova | eglynn: these objects are not distinguished | 15:13 |
| eglynn | nprivalova: ... so only delta is the source value, right? | 15:14 |
| nprivalova | eglynn: yep | 15:14 |
| eglynn | nprivalova: when we discussed this yesterday (IIRC) my instinct was ... | 15:14 |
| eglynn | ... since this scenario is not going to come up in real-life | 15:14 |
| eglynn | ... => we shouldn't sweat it too much | 15:14 |
| nprivalova | eglynn: my problem is that I need to maintain this behavior in HBase :) | 15:15 |
| *** jaypipes has joined #openstack-ceilometer | 15:15 | |
| eglynn | nprivalova: can you explain why? do you have a definite use-case for the multiple sources? | 15:15 |
| nprivalova | eglynn: no, we have unit-tests with it :( | 15:15 |
| nprivalova | eglynn: "scenario-tests" | 15:16 |
| eglynn | nprivalova: so the test referred to in the bug https://bugs.launchpad.net/ceilometer/+bug/1301371 | 15:16 |
| eglynn | nprivalova: ... that's passing on master, but failing on your development branch? | 15:17 |
| gordc | nprivalova: i think we should probably hold off on the hbase updates... i think we need to really finalise the data model we want to store... ie. multiple sources needed. | 15:17 |
| nprivalova | eglynn: yep, but there are several tests, not only this. I made a hack to make it work but meter-list works too long because of it | 15:18 |
| *** mihgen has quit IRC | 15:18 | |
| eglynn | gordc: agree we need closure on the unfinished discussion started at the last summit https://wiki.openstack.org/wiki/Ceilometer/blueprints/tighten-model | 15:19 |
| eglynn | nprivalova: "works too long"? | 15:19 |
| gordc | eglynn: agreed. no point wasting time fixing for something that may be wrong. | 15:19 |
| eglynn | nprivalova: i.e. the test takes too long to run? | 15:20 |
| nprivalova | gordc, eglynn, I would like meter-list works correct in Icehouse. Maybe long but correct | 15:20 |
| nprivalova | eglynn: on real env, not test | 15:20 |
| gordc | nprivalova: i think there's also another problem with meter-list. | 15:21 |
| nprivalova | eglynn, gordc, actually my complains about bug 1301371 is for discussions on summit | 15:21 |
| eglynn | nprivalova: ok, two things ... 1. getting another changes into icehouse would need an RC2 to be cut, right? | 15:21 |
| *** openstackgerrit has quit IRC | 15:21 | |
| *** openstackgerrit has joined #openstack-ceilometer | 15:21 | |
| eglynn | nprivalova: and 2. meter-list working unexpectedly with multi-source resources in real env isn't really a problem if "real" samples always respect a 1:1 relationship between resource:source | 15:23 |
| gordc | nprivalova: right now, we store meters in a list with resource in mongo/hbase. that list may contain stale meters if we ever run ttl since we only ever add and never remove.... might also be a problem in sql. | 15:23 |
| nprivalova | eglynn: I need only one patch https://review.openstack.org/#/c/78244/9 . It repairs meter-list and doesn't affect any other modules except HBase | 15:23 |
| nprivalova | gordc: I fixed it :) | 15:23 |
| gordc | nprivalova: really? cool! | 15:24 |
| nprivalova | gordc: stop :) | 15:24 |
| eglynn | nprivalova: would you consider https://bugs.launchpad.net/ceilometer/+bug/1288284 a release-critical bug? | 15:25 |
| gordc | nprivalova: i would think only critical items would justify a RC2 build htough. if not, it should be a backport | 15:25 |
| eglynn | nprivalova: (... marked as Medium currently) | 15:25 |
| eglynn | nprivalova: ... if not release-critical, I'd suspect we'd have a job persuading ttx to cut an RC2 | 15:25 |
| eglynn | nprivalova: ... likely he'd pushback to just punt to 1st stable/icehouse release | 15:26 |
| eglynn | (i.e. wot gordc just said) | 15:26 |
| nprivalova | eglynn: let me try to explain. Now get_meters returns very cut version of meters. "very" means only 10% | 15:26 |
| nprivalova | eglynn: about | 15:26 |
| eglynn | nprivalova: ... so reading between the lines, do you consider this bug to be more important than Medium? | 15:27 |
| nprivalova | eglynn: It's hard to evaluate how critical it is... From one hand it's just API output | 15:27 |
| eglynn | ("medium" == "Fix when convenient, or schedule to fix later") | 15:28 |
| nprivalova | eglynn: from other hand meter-list is very common request | 15:28 |
| gordc | nprivalova: for me, critical is 'this doesn't start or this crashes' | 15:28 |
| nprivalova | gordc: yep, I mostly agree | 15:29 |
| eglynn | gordc: +1 | 15:29 |
| admin0 | +1 | 15:29 |
| gordc | nprivalova: i guess one way to look at it is if you were a deployer and saw this error about this at 4am in the morning. would you fix it or would you sleep for another 3 hrs.lol... | 15:30 |
| gordc | that said, i'd probably always choose to sleep for another 3 hrs. | 15:30 |
| eglynn | nprivalova: I could be wrong, but I'm thinking incomplete info returned by GET /v2/meters doesn't warrant the Critical tag ... but perhaps High? | 15:31 |
| nprivalova | gordc: hehe :) Maybe it's not normal but my dreams were about this issue today :) | 15:31 |
| nprivalova | eglynn: yep, High is better that Medium or Critical | 15:31 |
| eglynn | nprivalova: ... so you could re-prioritize it as "High" and see if we can get the patch landed on master first at least | 15:32 |
| gordc | eglynn: agreed, non-critical to me. | 15:32 |
| nprivalova | eglynn: ok, will do | 15:32 |
| eglynn | nprivalova: ... without it being on master, there's only a very weak case to persuade ttx we need an RC2 | 15:32 |
| gordc | nprivalova: i've had a few of those dreams... scary that i dream about ceilometer now. | 15:32 |
| eglynn | nprivalova: ... but even with it on master, he still may baulk at cutting another RC and instead prolly would punt to stable/icehouse | 15:33 |
| nprivalova | eglynn: ok. I have some plans about performance against backends and wanted to use master-HBase for it... And maybe prepare some demo for summit about HBase to show how it works. I guess that it's the main reason (for me) to have it it Icehouse. But if think once more it's not critical actually | 15:36 |
| nprivalova | gordc: I'm trying to understand your ttl example. Do you mean that we should clean meters in Resource table if meter becomes removed? | 15:39 |
| gordc | nprivalova: right. well if i remember correctly, we currently have a a list of meters under resource which we append to (if new). but if we run ttl, there's a chance that the sample (and meter) will be removed but it won't be removed in list of meters under resource. | 15:41 |
| gordc | nprivalova: so if we call meter-list we'll get back meters that may have been cleaned by ttl (or whatever other reason, ie. archiving) | 15:42 |
| gordc | nprivalova: did that make sense? basically the list we are storing under resource is a historical list and not one that reflects what currently exists in db... (disclaimer: i've yet to verify this) | 15:46 |
| nprivalova | gordc: hmm, hard to imagine remove of meter if there is no "meter" table, only "sample" actually...But it looks like everything should be ok in HBase. We may determine ttl for all tables and entries. If all samples related to one meter disappears after 100 minutes then entry that corresponds this meter in resource table will disappear too because it's "age" equals to the latest sample that corresponds meter | 15:49 |
| nprivalova | gordc: now we store "columns" in HBase, not list | 15:50 |
| gordc | nprivalova: ok cool. yeah i haven't really seen your new code but i think adding columns instead of list makes more sense for hbase. | 15:51 |
| gordc | nprivalova: this is just based on previously looking at hbase/mongo code. | 15:51 |
| *** Vic has joined #openstack-ceilometer | 15:51 | |
| nprivalova | gordc: yes. actually there are a lot of things that may be done in HBase. E.g. now we use only one filter. But HBase contains 10th of them | 15:53 |
| gordc | nprivalova: you proposing a session on hbase? i know Alexei_987 has one for data model i think we should make sure we implement the data model specific to the backend. | 15:56 |
| nprivalova | gordc: unfortunately I cannot go to summit. I hope Alexey will go. In this case he will be responsible for HBase demo :) Hope he will not mind | 15:58 |
| *** Ruetobas has quit IRC | 16:01 | |
| eglynn | nprivalova: "And maybe prepare some demo for summit about HBase to show how it works" ==> I'll look forward to it! :) | 16:02 |
| eglynn | nprivalova: "unfortunately I cannot go to summit" ==> that's a pity! :( | 16:02 |
| *** Ruetobas has joined #openstack-ceilometer | 16:03 | |
| *** thomasem has joined #openstack-ceilometer | 16:03 | |
| nprivalova | eglynn: it's better to home O:-) | 16:04 |
| nprivalova | eglynn: *to be | 16:04 |
| eglynn | nprivalova: yeap, I hear ya! :) | 16:04 |
| *** samcdona has joined #openstack-ceilometer | 16:06 | |
| *** Ruetobas has quit IRC | 16:08 | |
| htruta | eglynn: i'm trying to get some ceilometer information in nova. Is that possible? can i get it through ceilometer client or the file generated by ceilometer? | 16:09 |
| eglynn | htruta: "get some ceilometer information in nova" == "get some information about nova resources from ceilometer"? | 16:10 |
| eglynn | htruta: here are nova meters that ceilometer collects http://docs.openstack.org/developer/ceilometer/measurements.html#compute-nova | 16:11 |
| eglynn | htruta: you can access these programmatically via the python-ceilometerclient library | 16:11 |
| eglynn | htruta: or via the CLI | 16:11 |
| *** admin0 has left #openstack-ceilometer | 16:12 | |
| eglynn | htruta: e.g. $ ceilometer statistics -m cpu_util -q resource=$INSTANCE_ID | 16:12 |
| htruta | eglynn: exactly! that's what i want | 16:12 |
| eglynn | htruta: ^^^ to get the CPU utilization for a particular instance | 16:12 |
| eglynn | htruta: you need to ensure that the ceilometer compute agent is running on every nova-compute node | 16:13 |
| *** Ruetobas has joined #openstack-ceilometer | 16:14 | |
| htruta | can i instance the ceilometer client inside nova? | 16:14 |
| eglynn | htruta: I don't understand your question | 16:18 |
| htruta | eglynn: is it possible to get access to the ceilometer via client on the nova? | 16:21 |
| htruta | programmatically, i mean | 16:22 |
| eglynn | htruta: when you say "on the nova" do you mean "on the nova compute node"? | 16:22 |
| *** sphoorti_ has joined #openstack-ceilometer | 16:23 | |
| htruta | on the nova, for each compute node... i want the statistics about every single instance | 16:23 |
| eglynn | htruta: you can run the ceilometer client from any host that has access to keystone and the ceilometer-api service | 16:24 |
| eglynn | htruta: ... just like you can run the clients for any other openstack service | 16:25 |
| *** nacim has quit IRC | 16:38 | |
| *** admin0 has joined #openstack-ceilometer | 16:41 | |
| *** Vic has quit IRC | 16:41 | |
| *** admin0 has quit IRC | 16:44 | |
| htruta | eglynn: i didnt know this usage of the clients. thanks! | 16:55 |
| *** eglynn has quit IRC | 16:55 | |
| *** vrovachev has quit IRC | 17:04 | |
| *** sphoorti_ has left #openstack-ceilometer | 17:04 | |
| *** htruta has quit IRC | 17:13 | |
| *** jmckind has quit IRC | 17:38 | |
| *** nprivalova has quit IRC | 17:38 | |
| *** nprivalova has joined #openstack-ceilometer | 17:47 | |
| *** kun_huang has quit IRC | 18:02 | |
| *** mihgen has joined #openstack-ceilometer | 18:06 | |
| *** zul has quit IRC | 18:09 | |
| *** zul has joined #openstack-ceilometer | 18:12 | |
| *** urulama has joined #openstack-ceilometer | 18:24 | |
| *** gordc has quit IRC | 18:33 | |
| openstackgerrit | A change was merged to openstack/ceilometer: ensure dispatcher service is configured before rpc https://review.openstack.org/83595 | 18:38 |
| *** mihgen has quit IRC | 18:39 | |
| *** yassine has quit IRC | 18:45 | |
| *** _nadya_ has joined #openstack-ceilometer | 18:45 | |
| *** alexpilotti has quit IRC | 18:50 | |
| *** _nadya_ has quit IRC | 18:52 | |
| *** urulama has quit IRC | 18:56 | |
| *** rwsu has quit IRC | 19:01 | |
| *** jaypipes has quit IRC | 19:05 | |
| *** rwsu has joined #openstack-ceilometer | 19:07 | |
| *** _nadya_ has joined #openstack-ceilometer | 19:07 | |
| *** _nadya_ has quit IRC | 19:46 | |
| *** gordc has joined #openstack-ceilometer | 19:57 | |
| *** thomasem_ has joined #openstack-ceilometer | 20:38 | |
| *** thomasem has left #openstack-ceilometer | 20:38 | |
| *** thomasem_ is now known as thomasem | 20:38 | |
| *** gordc has quit IRC | 20:47 | |
| *** ok_delta has joined #openstack-ceilometer | 20:59 | |
| *** jdob has quit IRC | 21:00 | |
| *** yjiang5_1 has quit IRC | 21:01 | |
| *** yjiang5 has joined #openstack-ceilometer | 21:07 | |
| *** yjiang5 has quit IRC | 21:07 | |
| *** yjiang5 has joined #openstack-ceilometer | 21:35 | |
| *** promulo has quit IRC | 21:45 | |
| *** thomasem has quit IRC | 21:47 | |
| *** prad_ has quit IRC | 22:48 | |
| *** ok_delta has quit IRC | 22:56 | |
| openstackgerrit | Jenkins proposed a change to openstack/ceilometer: Updated from global requirements https://review.openstack.org/77639 | 23:11 |
| *** bada has quit IRC | 23:26 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!