| *** Kennan2 is now known as Kennan | 00:33 | |
| *** promulo__ has joined #openstack-ceilometer | 00:44 | |
| *** promulo has quit IRC | 00:48 | |
| *** liusheng has quit IRC | 01:18 | |
| *** Longgeek_ has joined #openstack-ceilometer | 01:46 | |
| *** Longgeek has quit IRC | 01:49 | |
| *** Longgeek_ has quit IRC | 02:47 | |
| *** Longgeek has joined #openstack-ceilometer | 02:47 | |
| *** Longgeek_ has joined #openstack-ceilometer | 03:16 | |
| *** Longgeek has quit IRC | 03:18 | |
| *** ddieterly has quit IRC | 03:57 | |
| *** ildikov has quit IRC | 04:06 | |
| *** deepthi has joined #openstack-ceilometer | 04:08 | |
| *** liusheng has joined #openstack-ceilometer | 04:41 | |
| *** ddieterly has joined #openstack-ceilometer | 04:58 | |
| *** ddieterly has quit IRC | 05:03 | |
| *** dikonoor has joined #openstack-ceilometer | 05:06 | |
| *** _nadya_ has joined #openstack-ceilometer | 05:28 | |
| *** _nadya_ has quit IRC | 05:45 | |
| *** ig0r__ has joined #openstack-ceilometer | 05:52 | |
| *** ig0r_ has quit IRC | 05:55 | |
| *** _nadya_ has joined #openstack-ceilometer | 06:14 | |
| *** ddieterly has joined #openstack-ceilometer | 06:28 | |
| *** ddieterly has quit IRC | 06:32 | |
| *** jwcroppe has joined #openstack-ceilometer | 06:43 | |
| *** belmoreira has joined #openstack-ceilometer | 06:53 | |
| *** exploreshaifali has joined #openstack-ceilometer | 06:53 | |
| *** ankita_wagh has joined #openstack-ceilometer | 07:23 | |
| *** _nadya_ has quit IRC | 07:24 | |
| *** mikedillion has joined #openstack-ceilometer | 07:26 | |
| *** exploreshaifali has quit IRC | 07:26 | |
| *** mikedillion has left #openstack-ceilometer | 07:29 | |
| *** ddieterly has joined #openstack-ceilometer | 07:29 | |
| *** _nadya_ has joined #openstack-ceilometer | 07:32 | |
| *** ddieterly has quit IRC | 07:33 | |
| *** ankita_wagh has quit IRC | 07:34 | |
| *** Ala has joined #openstack-ceilometer | 07:36 | |
| *** ildikov has joined #openstack-ceilometer | 07:42 | |
| *** _nadya_ has quit IRC | 07:46 | |
| *** safchain has joined #openstack-ceilometer | 07:54 | |
| *** exploreshaifali has joined #openstack-ceilometer | 08:12 | |
| *** _nadya_ has joined #openstack-ceilometer | 08:18 | |
| *** _nadya_ has quit IRC | 08:28 | |
| *** ddieterly has joined #openstack-ceilometer | 08:30 | |
| *** Ala has quit IRC | 08:33 | |
| *** ddieterly has quit IRC | 08:34 | |
| *** tshtilma has joined #openstack-ceilometer | 08:36 | |
| *** arunman has joined #openstack-ceilometer | 08:37 | |
| *** yassine_ has joined #openstack-ceilometer | 08:37 | |
| *** Ala has joined #openstack-ceilometer | 08:42 | |
| *** _nadya_ has joined #openstack-ceilometer | 08:46 | |
| *** tshtilma has quit IRC | 08:58 | |
| jd__ | the sileht is back!! | 09:24 |
|---|---|---|
| sileht | jd__, :) | 09:30 |
| *** ddieterly has joined #openstack-ceilometer | 09:31 | |
| *** ddieterly has quit IRC | 09:35 | |
| *** yassine_ has quit IRC | 09:55 | |
| *** ildikov has quit IRC | 09:59 | |
| *** eglynn has joined #openstack-ceilometer | 10:02 | |
| *** Kennan has quit IRC | 10:05 | |
| *** lan_ has quit IRC | 10:07 | |
| *** Ala has quit IRC | 10:08 | |
| *** ildikov has joined #openstack-ceilometer | 10:14 | |
| *** lan_ has joined #openstack-ceilometer | 10:19 | |
| *** Kennan has joined #openstack-ceilometer | 10:21 | |
| *** ddieterly has joined #openstack-ceilometer | 10:31 | |
| *** ddieterly has quit IRC | 10:36 | |
| *** lan__ has joined #openstack-ceilometer | 10:37 | |
| openstackgerrit | Merged openstack/gnocchi: sqlalchemy: fix lazy-loading of attributes https://review.openstack.org/198306 | 10:38 |
| *** Kennan2 has joined #openstack-ceilometer | 10:39 | |
| *** arunman has quit IRC | 10:40 | |
| *** lan_ has quit IRC | 10:40 | |
| *** Kennan has quit IRC | 10:41 | |
| *** Longgeek_ has quit IRC | 10:41 | |
| *** Longgeek has joined #openstack-ceilometer | 10:42 | |
| *** cdent has joined #openstack-ceilometer | 10:44 | |
| *** yatin has joined #openstack-ceilometer | 10:48 | |
| *** Kennan2 is now known as Kennan | 10:48 | |
| *** yatin has quit IRC | 10:57 | |
| *** yassine_ has joined #openstack-ceilometer | 11:11 | |
| *** yassine_ has quit IRC | 11:11 | |
| *** yatin has joined #openstack-ceilometer | 11:14 | |
| *** deepthi has quit IRC | 11:29 | |
| *** ildikov has quit IRC | 11:31 | |
| *** ddieterly has joined #openstack-ceilometer | 11:32 | |
| *** jaosorior has joined #openstack-ceilometer | 11:34 | |
| *** ddieterly has quit IRC | 11:37 | |
| jd__ | don't forget to review aodh patches | 11:41 |
| *** thorst has joined #openstack-ceilometer | 11:42 | |
| *** ildikov has joined #openstack-ceilometer | 11:44 | |
| _nadya_ | jd__: starting now :) | 12:02 |
| _nadya_ | jd__: looks like we don't have enough tests on gate, right? | 12:06 |
| _nadya_ | jd__: sorry, I see 802. That's not bad | 12:07 |
| *** tshtilma has joined #openstack-ceilometer | 12:07 | |
| *** ildikov has quit IRC | 12:22 | |
| *** edmondsw has joined #openstack-ceilometer | 12:24 | |
| jd__ | _nadya_: we have everything except devstack-gate/tempest | 12:26 |
| *** gordc has joined #openstack-ceilometer | 12:27 | |
| cdent | looks like prad and _nadya_ (with a little bit of llu) are the dependent variables on the midcycle schedule. This is what we end up with to match their schedules: https://www.google.com/calendar/embed?src=cXZqcTY2ZXJjcnYxbG9kaGtoMHZwNmxrNDRAZ3JvdXAuY2FsZW5kYXIuZ29vZ2xlLmNvbQ | 12:27 |
| cdent | That's not fixed yet, need prad to check my work. | 12:28 |
| _nadya_ | cdent: Let me check | 12:30 |
| _nadya_ | cdent: google tells me that testing is in 7am | 12:30 |
| cdent | yes, which is within one of the zones you blocked out: 7-9, 10-1, 3-6 | 12:32 |
| *** ddieterly has joined #openstack-ceilometer | 12:33 | |
| openstackgerrit | Merged openstack/ceilometer: Switch to using pbr's autodoc capability https://review.openstack.org/120441 | 12:37 |
| *** ddieterly has quit IRC | 12:38 | |
| _nadya_ | cdent: looks like time converter lied to me. ok, I will be available. But if my schedule is a problem, I can fix it. And actually I want if it doesn't break the schedule | 12:38 |
| *** ildikov has joined #openstack-ceilometer | 12:39 | |
| cdent | We can still make adjustments, that schedule is not set in stone, that's just what I was able to fit in with the requirements. If you've got a different _actual_ schedule than if you could add a nadya2 line to the doodle poll I'll see what i can make fit in a more civilised fashion: http://doodle.com/gzhumwaakrqds38h | 12:40 |
| _nadya_ | cdent: I'll check now | 12:41 |
| _nadya_ | cdent: aha, the final calendar is in UTC too. I thought Google has used my TZ | 12:43 |
| _nadya_ | cdent: everything is ok then | 12:43 |
| cdent | cool, when prad shows up I'll make sure he agrees with my assessment and then we can make any remaining adjustments if necessary | 12:45 |
| gordc | that's in UTC? | 12:47 |
| *** exploreshaifali has quit IRC | 12:49 | |
| *** yatin has quit IRC | 12:49 | |
| * gordc is going to coin flip whether or not i wake up at 3am... | 12:49 | |
| gordc | also will coinflip whether i wake up for coinflip | 12:49 |
| _nadya_ | gordc: yep, UTC :) | 12:49 |
| _nadya_ | gordc: I believe 1 hour for tests is enough so the second day you may sleep one hour more | 12:50 |
| *** exploreshaifali has joined #openstack-ceilometer | 12:51 | |
| gordc | lol thanks | 12:51 |
| cdent | the controlling factor on the tests being then is _nadya_, gordc so perhaps you can do some smooth talking to get a reschedule? | 12:55 |
| gordc | cdent: i'm ok. either way it's 3am something. | 12:57 |
| gordc | functional testing for me seems pretty straight forward so i'm fine with timing. | 12:57 |
| *** Ala has joined #openstack-ceilometer | 12:58 | |
| jasonamyers | good morning | 12:59 |
| _nadya_ | gordc, cdent actually I have two questions: 1. how are we going to test all pieces together and 2. Rally tests | 12:59 |
| *** exploreshaifali has quit IRC | 13:00 | |
| cdent | _nadya_ sounds like you got the beginnings of an etherpad to add here: https://wiki.openstack.org/wiki/Meetings/Ceilometer/Liberty_Virtual_Mid-Cycle#Topics | 13:00 |
| gordc | all pieces as in all ceilometer pieces? | 13:01 |
| _nadya_ | gordc, cdent so maybe we move it to the end of second day? | 13:01 |
| _nadya_ | gordc: yep | 13:01 |
| jasonamyers | woot for 2AM starts! | 13:01 |
| _nadya_ | jasonamyers: it's utc | 13:01 |
| cdent | jasonamyers: your flexibility has made the scheduling _much_ easier | 13:02 |
| jasonamyers | haha I'm nobody you should schedule like you like | 13:02 |
| *** julim has joined #openstack-ceilometer | 13:03 | |
| jasonamyers | gonna be fun, really looking forward to this | 13:03 |
| gordc | jasonamyers: it's typical euro-centric views. :) | 13:03 |
| jasonamyers | I won't snore too much | 13:03 |
| cdent | gordc: I'll have you know that I paid strict attention to the voting on both doodles and topics | 13:04 |
| cdent | DEMOCRACY FOR THE WIN | 13:04 |
| * gordc forgot about camera... will not be sigining in | 13:04 | |
| jasonamyers | dat camera at 2am ... gonna be awesome | 13:06 |
| *** yassine_ has joined #openstack-ceilometer | 13:08 | |
| openstackgerrit | Merged openstack/aodh: Fix options registeration in tests https://review.openstack.org/197071 | 13:19 |
| *** annasort has joined #openstack-ceilometer | 13:24 | |
| *** rbrooker has joined #openstack-ceilometer | 13:29 | |
| openstackgerrit | Merged openstack/aodh: Replaces methods deprecated in pymongo3.0 https://review.openstack.org/196611 | 13:29 |
| _nadya_ | jd__: ping | 13:29 |
| openstackgerrit | Merged openstack/aodh: Move aodh.alarm.storage to aodh.storage https://review.openstack.org/196722 | 13:29 |
| _nadya_ | cdent: | 13:30 |
| openstackgerrit | Merged openstack/aodh: Change the default api server port https://review.openstack.org/198210 | 13:30 |
| cdent | hi _nadya_ | 13:30 |
| _nadya_ | cdent: hi :) | 13:30 |
| _nadya_ | cdent: I hae some errors after applying "Move aodh.alarm.storage to aodh.storage " | 13:30 |
| cdent | there's a fair few dependent patches in that stack, yeah? | 13:31 |
| cdent | brb | 13:31 |
| _nadya_ | cdent: there are no dependency as I see... | 13:32 |
| * cdent will check when he's b from the brb | 13:32 | |
| _nadya_ | cdent: anyway, https://gist.github.com/anonymous/48992116a8764bf2120a | 13:32 |
| _nadya_ | cdent: I'll run it again in new venv | 13:33 |
| _nadya_ | cdent: to be sure | 13:33 |
| *** claudiub has joined #openstack-ceilometer | 13:34 | |
| *** liusheng_ has joined #openstack-ceilometer | 13:36 | |
| *** exploreshaifali has joined #openstack-ceilometer | 13:37 | |
| *** claudiub has quit IRC | 13:38 | |
| *** gordc_internal has joined #openstack-ceilometer | 13:40 | |
| *** promulo__ has quit IRC | 13:40 | |
| *** Longgeek has quit IRC | 13:43 | |
| *** [1]tshtilma has joined #openstack-ceilometer | 13:53 | |
| *** prad has joined #openstack-ceilometer | 13:53 | |
| cdent | _nadya_: I think that error is expected when there is not coordination backend set | 13:53 |
| *** tshtilma has quit IRC | 13:54 | |
| *** [1]tshtilma is now known as tshtilma | 13:54 | |
| *** r-daneel has joined #openstack-ceilometer | 13:56 | |
| *** rbak has quit IRC | 13:56 | |
| cdent | prad: hola, I've made a first stab at a schedule. If you could check my math that would be dandy: https://www.google.com/calendar/embed?src=cXZqcTY2ZXJjcnYxbG9kaGtoMHZwNmxrNDRAZ3JvdXAuY2FsZW5kYXIuZ29vZ2xlLmNvbQ | 13:56 |
| cdent | the thing I'm seeing _nadya_, jd__ is that the postgresql tests aren't working at all "too many clients already" during the table creation process | 13:58 |
| *** tongli has joined #openstack-ceilometer | 13:59 | |
| jd__ | cdent: in aodh? yeah we should fix it and gate on that | 13:59 |
| cdent | so known issue then? | 13:59 |
| prad | cdent, morning | 14:00 |
| prad | cdent, sure i'll take a look now | 14:00 |
| prad | ildikov, thx for the updates to the abstract, looks good to me | 14:01 |
| prad | cdent, time is UTC i presume? | 14:01 |
| cdent | prad, on that link, yes | 14:03 |
| ildikov | prad: hi. great, thanks | 14:04 |
| prad | cdent, hmm the morning schedule on that link is a bit rough on US folks.. ? | 14:05 |
| cdent | a) no us folks said they were required to attend those sessions | 14:06 |
| cdent | b) the sessions which you said you needed to attend are in the afternoon | 14:06 |
| cdent | c) basically the entire schedule is based on what you and _nadya_ said in terms of availability and "required to attend" | 14:06 |
| cdent | So using the voting that people did that's the schedule that maps | 14:07 |
| *** r-daneel has quit IRC | 14:07 | |
| cdent | c.5) everyone else had long windows | 14:09 |
| prad | cdent, yea understand, its fine with me | 14:09 |
| prad | cdent, wasnt sure if people who are interested but havnt added themselves to the doodle want to dial in | 14:10 |
| *** ddieterly has joined #openstack-ceilometer | 14:10 | |
| prad | but we can only satisfy only so many | 14:10 |
| *** rbak has joined #openstack-ceilometer | 14:11 | |
| cdent | you don't vote, you don't git | 14:11 |
| *** Ephur has joined #openstack-ceilometer | 14:11 | |
| prad | yea | 14:11 |
| prad | cdent, looks good to me then | 14:12 |
| cdent | gordc: can you make a final declaration of finally declared, finally? | 14:12 |
| *** liusheng_ has quit IRC | 14:17 | |
| *** logan2 has quit IRC | 14:17 | |
| *** r-daneel has joined #openstack-ceilometer | 14:20 | |
| *** logan2 has joined #openstack-ceilometer | 14:41 | |
| *** Longgeek has joined #openstack-ceilometer | 14:43 | |
| *** jruano has joined #openstack-ceilometer | 14:47 | |
| *** Longgeek has quit IRC | 14:49 | |
| *** ddieterly has quit IRC | 14:54 | |
| *** ddieterly has joined #openstack-ceilometer | 14:57 | |
| thorst | cdent: have a min to discuss the powervm-compute-inspector blueprint? | 14:57 |
| *** ddieterly has quit IRC | 14:58 | |
| *** Ephur has quit IRC | 15:01 | |
| *** ddieterly has joined #openstack-ceilometer | 15:02 | |
| *** serverascode has quit IRC | 15:05 | |
| *** serverascode has joined #openstack-ceilometer | 15:05 | |
| *** logan2 has quit IRC | 15:11 | |
| *** logan2 has joined #openstack-ceilometer | 15:19 | |
| *** belmoreira has quit IRC | 15:27 | |
| *** obedmr has joined #openstack-ceilometer | 15:33 | |
| *** gordc_internal has quit IRC | 15:34 | |
| *** gordc_internal has joined #openstack-ceilometer | 15:34 | |
| gordc | cdent: are the topics ordered by popularity? | 15:36 |
| *** gordc_internal has quit IRC | 15:36 | |
| *** gordc_internal has joined #openstack-ceilometer | 15:36 | |
| cdent | gordc: there are ordered by popularity here: https://docs.google.com/spreadsheets/d/1mKLlRtTUWEq62ZrIi4Xr7EoV7SwGUNPORa7yauXYCbk/edit#gid=0 | 15:37 |
| *** gordc_internal has quit IRC | 15:37 | |
| cdent | on the calendar they are ordered by where they fit for the people who said they need to be there on here: https://etherpad.openstack.org/p/ceilometer-liberty-midcycle | 15:37 |
| cdent | thorst: still around | 15:37 |
| cdent | ? | 15:37 |
| thorst | yep | 15:38 |
| cdent | what's up? | 15:38 |
| thorst | wanted to talk some process stuff real quick | 15:38 |
| *** exploreshaifali has quit IRC | 15:38 | |
| cdent | gordc might be better for that, but we're both here, so go for it. | 15:38 |
| thorst | so I was initially targetting to put the powervm driver in ceilometer core. Had a great discussion with dhellman. Turns out I just didn't understand the plugin model 100% | 15:39 |
| gordc | cdent: kk. was going to ask for schematisation to be flipped with delete/deprecate | 15:40 |
| thorst | he recommended that we do it initially in a stackforge project (ceilometer-powervm) since its tied to pypowervm which is pretty mature, but not 1.0 yet | 15:40 |
| thorst | wanted to make sure that was OK from a process side? And if so, I could propose an update to the blueprint to reflect that | 15:41 |
| cdent | gordc: the reason for d or do being in the afternoon is to make sure that it includes prad | 15:42 |
| cdent | thorst if you're happy to have it as an external plugin as far as I'm concerned that is a much better way to make it go | 15:42 |
| cdent | and I think everyone else would agree too | 15:43 |
| thorst | cdent: I don't mind it while it matures. I would love to propose it into core perhaps the next release? | 15:43 |
| gordc | thorst: is the blocker that pypowervm isn't going to get into requirements? | 15:43 |
| cdent | thorst: If you like I'm sure that would be fine but I think over the long term we'll be encouraging more and more people to externally plug-in | 15:44 |
| thorst | gordc: yeah. Which had a great discussion | 15:44 |
| cdent | (or at least we should) | 15:44 |
| gordc | cdent: oh... i don't see his name on that topic. maybe it got deleted. | 15:44 |
| thorst | honestly don't mind the external plugin approach at the moment | 15:44 |
| cdent | gordc: It's not but that's because I made an executive decision: he needs to be there whether he realizes it or not because of TheEmployer™ | 15:45 |
| thorst | just want to make sure I can still doc it properly, get reviews (second set of eyes to make sure we're doing things nicely), etc... | 15:45 |
| cdent | also, I made the afternoon sessions long on purpose | 15:45 |
| gordc | thorst: ok with plugin approach | 15:45 |
| cdent | thorst: if you make it externally you'll need to make sure that you explicitly invited ceilo people to the review as people won't see it on their radar | 15:46 |
| gordc | cdent: what's the debate on delete/deprecate? or just in anticipation of one? | 15:46 |
| thorst | yeah, I don't mind doing that. Just don't want to frustrate anyone with excess reviews | 15:46 |
| thorst | but I would find a lot of value in having people review. | 15:46 |
| cdent | gordc: it's not so much a debate as listing off and acknowledging all the things that will need to happen: devstack plugins, tempest plugins, grenade plugins, notification of downstream people (if we delete) | 15:47 |
| cdent | but if you're cool with that stuff already being acknowledge then no problem moving it | 15:47 |
| *** zhiyan has quit IRC | 15:47 | |
| *** yassine_ has quit IRC | 15:47 | |
| *** zhiyan has joined #openstack-ceilometer | 15:47 | |
| cdent | I just had the impression that we hadn't really been thinking that hard about the details of what it really means to be making all these changes | 15:47 |
| cdent | and wanted to make sure we set aside some proper time | 15:47 |
| cdent | and the schematisation stuff doesn't need anytime because nothing will actually happen ;) | 15:48 |
| gordc | cdent: ... i ... can't argue with that. | 15:48 |
| *** eglynn is now known as eglynn-afk | 15:49 | |
| cdent | you can do it gordc, I know you can! | 15:49 |
| gordc | cdent: me 2 years ago thought that... i will be fooled no more. | 15:50 |
| * sileht will write a bot to count how many time he see that : 'schematisation stuff doesn't need anytime because nothing will actually happen' | 15:50 | |
| cdent | I'm only parroting what I've already seen | 15:50 |
| * cdent would love for the people who want it to get it | 15:50 | |
| * sileht too | 15:50 | |
| cdent | but I no longer believe it is necessary. In fact I think it is an anti-pattern | 15:51 |
| gordc | you both lie. | 15:51 |
| * cdent likes to say "anti-pattern" | 15:51 | |
| cdent | sileht: now that you're back you okay with that schedule, have we left out something precious to you? | 15:52 |
| cdent | gordc: you can keep on thinking about it until tomorrow? and then we make a final decision? If you want to make some changes check with prad, _nadya_ and llu. Everyone else is quite flexible. | 15:53 |
| sileht | cdent, I'm not sure I have totally understand what happen, I guess yes :p | 15:54 |
| cdent | :) | 15:55 |
| gordc | cdent: i'm ok with schedule. it's going to be less than ideal for americas regardless. | 15:55 |
| *** Ala has quit IRC | 15:55 | |
| gordc | until we get more participants from here, i can't really change anything. | 15:56 |
| cdent | we had a vote to decide a) what topics we would mid-cycle about d) what hours during the day people are available. Then we decided how many votes were necessary for a topic to be relevant. | 15:56 |
| cdent | ^ sileht | 15:56 |
| cdent | gordc: get on that gordc, we need more participants | 15:56 |
| gordc | fly me to south america. i'll do some recruiting down there. | 15:57 |
| gordc | i hear rio is a hotspot for devs | 15:57 |
| *** tshtilma has quit IRC | 15:57 | |
| gordc | and the mountains in chile. | 15:57 |
| cdent | I'll gladly go recruiting in patagonia if you like | 15:58 |
| gordc | no! i already offered. | 15:58 |
| sileht | cdent, the planning looks for me | 15:58 |
| * sileht is adjusting its alarm clock | 15:59 | |
| sileht | *looks good* | 16:00 |
| *** Longgeek has joined #openstack-ceilometer | 16:00 | |
| *** ddieterly has quit IRC | 16:01 | |
| *** thorst is now known as thorst_afk | 16:02 | |
| *** ddieterly has joined #openstack-ceilometer | 16:02 | |
| *** rwsu has joined #openstack-ceilometer | 16:04 | |
| *** Longgeek has quit IRC | 16:05 | |
| *** safchain has quit IRC | 16:08 | |
| *** ankita_wagh has joined #openstack-ceilometer | 16:17 | |
| *** exploreshaifali has joined #openstack-ceilometer | 16:24 | |
| *** ankita_wagh has quit IRC | 16:25 | |
| *** jwcroppe_ has joined #openstack-ceilometer | 16:27 | |
| *** jwcroppe has quit IRC | 16:30 | |
| *** jwcroppe_ has quit IRC | 16:32 | |
| *** pauldix has joined #openstack-ceilometer | 16:32 | |
| *** jaosorior has quit IRC | 16:36 | |
| *** packet has joined #openstack-ceilometer | 16:39 | |
| *** tshtilma has joined #openstack-ceilometer | 16:40 | |
| *** tongli has quit IRC | 16:42 | |
| *** tongli has joined #openstack-ceilometer | 16:42 | |
| *** ankita_wagh has joined #openstack-ceilometer | 16:44 | |
| *** rbrooker has quit IRC | 16:46 | |
| *** ankita_wagh has quit IRC | 16:46 | |
| *** ankita_wagh has joined #openstack-ceilometer | 16:46 | |
| *** tongli has quit IRC | 16:47 | |
| *** ildikov has quit IRC | 16:48 | |
| *** ondergetekende_ has quit IRC | 16:50 | |
| *** ondergetekende has joined #openstack-ceilometer | 16:50 | |
| prad | gordc, hey | 16:51 |
| prad | gordc, thx for the review, regarding couple of your comments | 16:51 |
| gordc | prad: sure | 16:51 |
| prad | gordc, when you say make each definition an individual end point | 16:52 |
| prad | gordc, do you see an issue with current way of initializing each definition ? or just suggesting an alternative | 16:53 |
| openstackgerrit | Matt Riedemann proposed openstack/ceilometer: delete python bytecode before every test run https://review.openstack.org/198815 | 16:54 |
| gordc | an alternative... it's not a big deal i think. so currently, all our 'code definitions' are endpoints and they each deal with one meter. | 16:54 |
| gordc | yours is a single endpoint which deals with multiple meters. | 16:54 |
| * cdent sighs | 16:54 | |
| prad | right | 16:55 |
| cdent | (at deleting *pyc) | 16:55 |
| openstackgerrit | Matt Riedemann proposed openstack/ceilometer: delete python bytecode before every test run https://review.openstack.org/198815 | 16:55 |
| gordc | it's more inline with how event definitions work so that's fine... i'm not even sure if it's a performance boost anyways so you can ignore | 16:55 |
| prad | ok cool | 16:55 |
| prad | gordc, secondly, | 16:55 |
| prad | the vol=1 comment | 16:56 |
| prad | the reason i have it the way it is is cos jsonpath_rw hates numbers .. so i future we might need some plugin based approach | 16:56 |
| prad | for now this seemed like a simple way to default to 1 until we deprecate these meters | 16:57 |
| *** dscottsf has joined #openstack-ceilometer | 16:57 | |
| openstackgerrit | Matt Riedemann proposed openstack/ceilometer: delete python bytecode before every test run https://review.openstack.org/198815 | 16:57 |
| *** dscottsf has quit IRC | 16:58 | |
| gordc | prad: you mean my comment on the get_volume method? | 16:58 |
| prad | gordc, yea volume not optional comment | 16:58 |
| prad | "that said, i feel like vol should be a required attribute" | 16:59 |
| gordc | you can't add a definition that says volume: 1? | 17:00 |
| *** _nadya_ has quit IRC | 17:00 | |
| *** _cjones_ has joined #openstack-ceilometer | 17:00 | |
| *** Longgeek has joined #openstack-ceilometer | 17:01 | |
| gordc | prad: can't you change parse_fields to detect if it's an int? | 17:01 |
| prad | gordc, not with the way it is now.. jsonpath_rw doesnt like it.. I could have a separate nonmeters class and override it perhaps | 17:02 |
| prad | gordc, perhaps , but then we'll run into cases where vol is an arithmetic expression.. for just this case i can do an int check, but thats not solving the whole problem. Thought i save this for a plugin approach later | 17:03 |
| prad | but i'm cool with doing it just for vol =1 meters for now and worry about others later | 17:03 |
| gordc | prad: yeah, we should have better support in parse_fields (if you want to continue supporting volume=1 stuff) | 17:04 |
| gordc | it's pretty weird to say: 'don't put a volume if it's a number' | 17:05 |
| *** Longgeek has quit IRC | 17:06 | |
| prad | gordc, ok np, i'll look into something more elegant | 17:06 |
| *** thorst_afk is now known as thorst | 17:07 | |
| openstackgerrit | gordon chung proposed openstack/ceilometer: add mandatory limit value to meter list https://review.openstack.org/198818 | 17:10 |
| *** tongli has joined #openstack-ceilometer | 17:10 | |
| *** promulo has joined #openstack-ceilometer | 17:17 | |
| *** logan2 has quit IRC | 17:21 | |
| *** akgunjal has joined #openstack-ceilometer | 17:25 | |
| *** akgunjal has quit IRC | 17:37 | |
| *** logan2 has joined #openstack-ceilometer | 17:50 | |
| *** jwcroppe has joined #openstack-ceilometer | 17:53 | |
| *** jaosorior has joined #openstack-ceilometer | 17:59 | |
| *** ankita_w_ has joined #openstack-ceilometer | 18:00 | |
| *** ankita_wagh has quit IRC | 18:03 | |
| *** Kennan2 has joined #openstack-ceilometer | 18:03 | |
| *** Kennan has quit IRC | 18:04 | |
| prad | gordc, so regarding having the whole metadata as part of the sample, i kept it for backward compatibility.. i know there are cases we look at some info in the metadata. | 18:05 |
| prad | gordc, you want to trim this somehow or skip it entirely? | 18:06 |
| *** pauldix has quit IRC | 18:11 | |
| *** pauldix has joined #openstack-ceilometer | 18:12 | |
| *** tshtilma has quit IRC | 18:22 | |
| gordc | prad: i was thinking more that we could have a defineable metadata section. so similar to 'traits' section in events | 18:23 |
| gordc | if you don't define a metadata, you get nothing. | 18:23 |
| gordc | so in case of compute meters, we define a static set of metadata | 18:23 |
| *** dikonoor has quit IRC | 18:24 | |
| gordc | prad: or like how we define resources in gnocchi (with specific attributes) | 18:25 |
| prad | gordc, hmm that mean the user needs full understanding of a specific event structure? is it ok to put that burden on the user? | 18:25 |
| prad | i'm not sure | 18:26 |
| prad | s/event/metadata | 18:27 |
| gordc | fair. the problem with dumping entire message is that it's not resource metadata | 18:28 |
| gordc | it's just a raw dump. | 18:28 |
| gordc | by associating that dump with resource and sample is going to cause db to explode. | 18:28 |
| gordc | there needs to be separation i think. | 18:29 |
| gordc | i should add this is probably outside scope of your patch... but i think being able to define your metadata is useful. | 18:31 |
| prad | gordc, yea i understand your view.. with my patch we'll definitely make it easier to do this kind of a change in future | 18:32 |
| prad | i'm happy to look into trimming the metadata as a separate issue | 18:32 |
| prad | regarding the tests, i agree with you that its not very efficient to test every possible sample generated for each meter | 18:33 |
| gordc | prad: yeah, that point i was hoping it's possible to continue forward with. | 18:36 |
| prad | yep | 18:42 |
| *** ddieterly has quit IRC | 18:44 | |
| ankita_w_ | Hi gordc, So meters are getting built on the compute nodes and then sent to the message bus right ? | 18:47 |
| gordc | ankita_w_: right. so it's two different paths. | 18:47 |
| ankita_w_ | and notification agent listens to notifications which are bascially events | 18:47 |
| gordc | from polling agent, they poll data -> build meter -> publish to queue -> collector stores it | 18:47 |
| openstackgerrit | Pradeep Kilambi proposed openstack/ceilometer: Declarative meters support https://review.openstack.org/197633 | 18:47 |
| gordc | from notification agent, services(nova/cinder/etc) publish data -> notification agent picks it up -> build meter/event -> publish to queue -> collector stores it | 18:48 |
| ankita_w_ | I have a plugin for notifications which modifies the data when published on end points | 18:48 |
| ankita_w_ | I want to use that plugin for meters built by the compute agent as well | 18:49 |
| ankita_w_ | Is there a way to do that | 18:49 |
| gordc | the data from polling agents (currently) do not have any connection to the notification you get in notification agent | 18:49 |
| ankita_w_ | if I add the metering topic to notification_topics, in ceilometer.conf , why is it not working ? | 18:50 |
| gordc | ankita_w_: cdent had a proposal to do exactly that | 18:50 |
| gordc | https://review.openstack.org/#/c/192286/ | 18:50 |
| ankita_w_ | notification_topics = notifications, metering | 18:50 |
| gordc | so that has nothing to do with polling agents (compute agent) | 18:50 |
| gordc | compute agents publish to queue to be consumed by collector | 18:51 |
| ankita_w_ | so who listens to the topics we specify in the conf ? | 18:51 |
| ankita_w_ | why cant we make the notification agent listen to the topic on which polling agent is publishing? | 18:53 |
| gordc | that is for internal queue. which the collector consumes | 18:53 |
| gordc | http://docs.openstack.org/developer/ceilometer/architecture.html#high-level-architecture | 18:54 |
| gordc | that is the basic architecture as it currently stands... cdents proposal is required to do what you want to do. | 18:55 |
| cdent | ankita_w_: the changes I have in progress will let you do what you want | 18:55 |
| gordc | the polling agent does not communicate with notification agent... it communicates with collector. | 18:55 |
| cdent | basically the pollsters will send notifications rather than doing rpc to the collector | 18:55 |
| cdent | but it will be a while yet | 18:55 |
| ankita_w_ | yeah got it....thanku gordc and cdent | 18:56 |
| cdent | presumably there would be some way to publish to something that made a notification | 18:56 |
| cdent | rather than dispatch to the collector? | 18:56 |
| cdent | BTW gordc: I have my prototype working™ but testing it effectively in the context of the existing ceilo tests is proving a real bear | 18:56 |
| gordc | cdent: delete all the tests | 18:57 |
| ankita_w_ | just to clarify- notification_agent listens to the notification_topics which are specified in the conf ? or is it the collector ? | 18:58 |
| cdent | which conf do you mean ankita_w_ ? | 19:00 |
| ankita_w_ | ceilometer.conf on ceilometer server | 19:00 |
| cdent | okay, just wanted to be sure you weren't talking about nova's conf or something like that... | 19:01 |
| *** tongli has quit IRC | 19:01 | |
| ankita_w_ | ok...who listens to those topics ? | 19:01 |
| *** tongli has joined #openstack-ceilometer | 19:02 | |
| *** Longgeek has joined #openstack-ceilometer | 19:02 | |
| cdent | ankita_w_: I'm just looking into, to make sure I give you the actually correct answer rather than my guess | 19:02 |
| gordc | ankita_w_: sorry. metering_topic or notification+topic? | 19:04 |
| ankita_w_ | notification_topic | 19:04 |
| cdent | ankita_w_: so yeah, the notification agent lists for notification_topics. In fact, each plugin sets up a Target for each topic | 19:04 |
| ankita_w_ | the default metering_topic is set to metering, where all the meters are published right ? | 19:05 |
| gordc | yes | 19:05 |
| gordc | that is what all the agents publish to | 19:06 |
| ankita_w_ | if I set notification_topics = notifications, metering, wont the notification agent also receive meters published on metering topic ? | 19:06 |
| *** tongli has quit IRC | 19:06 | |
| gordc | no... we publish to metering.sample. | 19:06 |
| gordc | notification agent only listens to xyz.info and xyz.debug | 19:06 |
| ankita_w_ | got it :) | 19:07 |
| ankita_w_ | thanks a bunch | 19:07 |
| cdent | this is all far too opaque | 19:07 |
| gordc | ankita_w_: you can add sample. but you'll have notification agent and collector both grabbing from the queue | 19:07 |
| *** tongli has joined #openstack-ceilometer | 19:07 | |
| gordc | cdent: job security | 19:08 |
| ankita_w_ | the only reason am trying to do is ...my plugin is already working to modify the notifications in the format our system understands. I want to meters to go thru the same path instead of building a new plugin | 19:09 |
| ankita_w_ | the meters* | 19:09 |
| gordc | ankita_w_: you can specify a custom topic when publishing. | 19:10 |
| gordc | ankita_w_: have the notification agent listen to that topic... you'll still need to listen to .sample level though | 19:11 |
| ankita_w_ | so can I specify notification_topics = notifications, metering.sample or does it require additional changes | 19:12 |
| *** tongli has quit IRC | 19:12 | |
| gordc | https://github.com/openstack/ceilometer/blob/master/ceilometer/event/endpoint.py | 19:13 |
| gordc | in the same way we define, info and error level, you need a sample level | 19:13 |
| ankita_w_ | got it...thanks a lot..this solved my problem | 19:15 |
| gordc | ankita_w_: cool cool | 19:15 |
| *** Longgeek has quit IRC | 19:27 | |
| openstackgerrit | Chris Dent proposed openstack/ceilometer: [WIP] Spike at making the pollsters send notifications https://review.openstack.org/198865 | 19:28 |
| cdent | gordc I could with some commentary on that ^ | 19:28 |
| cdent | at your leisure, I'm done for tonight | 19:28 |
| * cdent waves | 19:29 | |
| *** cdent has quit IRC | 19:29 | |
| *** Ephur has joined #openstack-ceilometer | 19:30 | |
| *** tongli has joined #openstack-ceilometer | 19:38 | |
| *** tongli has quit IRC | 19:44 | |
| *** pauldix has quit IRC | 19:45 | |
| *** dencaval has quit IRC | 19:54 | |
| *** pauldix has joined #openstack-ceilometer | 19:56 | |
| *** pauldix has quit IRC | 19:58 | |
| *** exploreshaifali has quit IRC | 20:10 | |
| *** promulo__ has joined #openstack-ceilometer | 20:15 | |
| *** promulo has quit IRC | 20:16 | |
| openstackgerrit | Pradeep Kilambi proposed openstack/ceilometer: Declarative meters support https://review.openstack.org/197633 | 20:19 |
| ankita_w_ | gordc : I tried what we had just discussed about adding sample level in endpoint. I am still not getting the meters. I am running ceilometer juno | 20:20 |
| *** ddieterly has joined #openstack-ceilometer | 20:21 | |
| *** Longgeek has joined #openstack-ceilometer | 20:24 | |
| *** packet has quit IRC | 20:25 | |
| *** pauldix has joined #openstack-ceilometer | 20:30 | |
| gordc | ankita_w_: do you have collector running?/ | 20:39 |
| ankita_w_ | yes | 20:39 |
| gordc | notification agent and collector would be listening to same queue | 20:40 |
| gordc | and fighting over messages. | 20:40 |
| gordc | if you run rabbitmqctl list_queues do you see anthing on metering.sample queue? | 20:40 |
| ankita_w_ | lemme check | 20:40 |
| ankita_w_ | metering.info0 | 20:42 |
| ankita_w_ | metering.sample0 | 20:42 |
| ankita_w_ | metering.sample.info0 | 20:42 |
| gordc | did you put 'metering.sample' as your notification_topic? | 20:42 |
| *** tongli has joined #openstack-ceilometer | 20:44 | |
| ankita_w_ | I changed it to metering | 20:44 |
| ankita_w_ | afterwards | 20:44 |
| ankita_w_ | and restarted | 20:44 |
| *** ildikov has joined #openstack-ceilometer | 20:45 | |
| *** pauldix has quit IRC | 20:46 | |
| gordc | ankita_w_: ah i see... if you have both collector/notification agent running they'll be fighting on same queue. | 20:47 |
| gordc | ankita_w_: i need to step out for a bit. i can probably look at it later tonight if you want to send email... or i can look at it tomorrow | 20:48 |
| ankita_w_ | ok thanks | 20:48 |
| *** tongli has quit IRC | 20:48 | |
| *** tongli has joined #openstack-ceilometer | 20:49 | |
| *** gordc has quit IRC | 20:49 | |
| *** Longgeek has quit IRC | 20:50 | |
| *** thorst has quit IRC | 21:07 | |
| *** jaosorior has quit IRC | 21:16 | |
| *** fawadkhaliq has joined #openstack-ceilometer | 21:31 | |
| *** julim has quit IRC | 21:34 | |
| *** jruano has quit IRC | 21:35 | |
| *** pauldix has joined #openstack-ceilometer | 21:36 | |
| *** promulo has joined #openstack-ceilometer | 21:41 | |
| *** promulo__ has quit IRC | 21:41 | |
| *** eglynn-afk has quit IRC | 21:42 | |
| *** dscottsf has joined #openstack-ceilometer | 21:45 | |
| *** rjaiswal has joined #openstack-ceilometer | 21:46 | |
| *** Longgeek has joined #openstack-ceilometer | 21:47 | |
| *** dscottsf has quit IRC | 21:47 | |
| *** fawadkhaliq has quit IRC | 21:52 | |
| *** eglynn-afk has joined #openstack-ceilometer | 21:55 | |
| *** Kennan has joined #openstack-ceilometer | 22:02 | |
| *** Kennan2 has quit IRC | 22:03 | |
| *** Longgeek has quit IRC | 22:06 | |
| *** ildikov has quit IRC | 22:06 | |
| *** ddieterly has quit IRC | 22:08 | |
| *** ddieterly has joined #openstack-ceilometer | 22:09 | |
| *** ddieterly has quit IRC | 22:12 | |
| *** ddieterl_ has joined #openstack-ceilometer | 22:13 | |
| *** ddieterl_ has quit IRC | 22:13 | |
| *** prad has quit IRC | 22:15 | |
| *** ddieterly has joined #openstack-ceilometer | 22:16 | |
| *** ddieterl_ has joined #openstack-ceilometer | 22:17 | |
| *** ddieter__ has joined #openstack-ceilometer | 22:18 | |
| *** ddieterly has quit IRC | 22:19 | |
| *** ddieter__ has quit IRC | 22:20 | |
| *** ddieterly has joined #openstack-ceilometer | 22:21 | |
| *** ddieterl_ has quit IRC | 22:21 | |
| *** jruano has joined #openstack-ceilometer | 22:27 | |
| *** tongli has quit IRC | 22:33 | |
| *** tongli has joined #openstack-ceilometer | 22:34 | |
| *** eglynn-afk has quit IRC | 22:35 | |
| *** tongli has quit IRC | 22:37 | |
| *** tongli has joined #openstack-ceilometer | 22:37 | |
| *** tongli has quit IRC | 22:42 | |
| *** pauldix has quit IRC | 22:45 | |
| *** edmondsw has quit IRC | 22:47 | |
| *** tongli has joined #openstack-ceilometer | 22:48 | |
| *** ddieterly has quit IRC | 22:52 | |
| *** tongli has quit IRC | 22:53 | |
| openstackgerrit | Rohit Jaiswal proposed openstack/ceilometer: Implementation of dynamically reloadable pipeline https://review.openstack.org/190842 | 22:56 |
| *** jruano has quit IRC | 22:57 | |
| *** tongli has joined #openstack-ceilometer | 22:57 | |
| *** tongli has quit IRC | 23:02 | |
| *** rbak has quit IRC | 23:02 | |
| *** ankita_wagh has joined #openstack-ceilometer | 23:04 | |
| *** ankita_w_ has quit IRC | 23:07 | |
| *** Ephur has quit IRC | 23:07 | |
| *** ankita_wagh has quit IRC | 23:12 | |
| *** ankita_wagh has joined #openstack-ceilometer | 23:12 | |
| *** tongli has joined #openstack-ceilometer | 23:18 | |
| *** tongli has quit IRC | 23:23 | |
| *** thorst has joined #openstack-ceilometer | 23:24 | |
| *** thorst has quit IRC | 23:30 | |
| *** Longgeek has joined #openstack-ceilometer | 23:32 | |
| *** Longgeek has quit IRC | 23:50 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!