| *** fnaval has quit IRC | 00:17 | |
| *** harlowja_away is now known as harlowja | 00:30 | |
| *** rbowen has joined #openstack-ceilometer | 00:56 | |
| *** sbfox has joined #openstack-ceilometer | 00:58 | |
| *** _cjones_ has quit IRC | 01:02 | |
| *** rbowen has quit IRC | 01:03 | |
| *** _cjones_ has joined #openstack-ceilometer | 01:03 | |
| *** rbowen has joined #openstack-ceilometer | 01:04 | |
| *** sbfox has quit IRC | 01:06 | |
| *** _cjones_ has quit IRC | 01:07 | |
| *** sbfox has joined #openstack-ceilometer | 01:10 | |
| openstackgerrit | shengjie min proposed a change to openstack/ceilometer-specs: Add the instance uptime as a Ceilometer Meter https://review.openstack.org/104784 | 01:30 |
|---|---|---|
| openstackgerrit | shengjie min proposed a change to openstack/ceilometer-specs: Add the instance uptime as a Ceilometer Meter https://review.openstack.org/104784 | 01:36 |
| *** rbowen has quit IRC | 01:40 | |
| *** nosnos has joined #openstack-ceilometer | 01:58 | |
| *** thomasem has joined #openstack-ceilometer | 01:59 | |
| *** thomasem has quit IRC | 01:59 | |
| *** Ruetobas has quit IRC | 02:08 | |
| *** sbfox has quit IRC | 02:11 | |
| *** Ruetobas has joined #openstack-ceilometer | 02:17 | |
| *** fnaval has joined #openstack-ceilometer | 02:24 | |
| *** Ruetobas has quit IRC | 02:24 | |
| *** Ruetobas has joined #openstack-ceilometer | 02:28 | |
| *** changbl has joined #openstack-ceilometer | 02:36 | |
| *** Ruetobas has quit IRC | 02:50 | |
| *** Ruetobas has joined #openstack-ceilometer | 02:56 | |
| *** nati_ueno has quit IRC | 02:57 | |
| *** Ruetobas has quit IRC | 03:00 | |
| *** Ruetobas has joined #openstack-ceilometer | 03:06 | |
| *** harlowja is now known as harlowja_away | 03:16 | |
| *** sbfox has joined #openstack-ceilometer | 03:33 | |
| *** nosnos has quit IRC | 03:43 | |
| openstackgerrit | A change was merged to openstack/ceilometer: Transform sample_cnt type to int https://review.openstack.org/100441 | 03:44 |
| *** fnaval has quit IRC | 04:00 | |
| *** fnaval has joined #openstack-ceilometer | 04:01 | |
| *** nosnos has joined #openstack-ceilometer | 04:32 | |
| *** yatin has joined #openstack-ceilometer | 04:39 | |
| *** underyx is now known as Underyx | 04:56 | |
| *** Underyx is now known as underyx | 05:08 | |
| *** asvechnikov has quit IRC | 05:14 | |
| *** psharma has joined #openstack-ceilometer | 05:18 | |
| *** sbfox has quit IRC | 05:31 | |
| *** sbfox has joined #openstack-ceilometer | 05:35 | |
| *** ildikov has quit IRC | 05:39 | |
| *** fnaval has quit IRC | 06:00 | |
| *** _nadya_ has joined #openstack-ceilometer | 06:04 | |
| *** _nadya_ has quit IRC | 06:07 | |
| *** ildikov has joined #openstack-ceilometer | 06:07 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ceilometer: Imported Translations from Transifex https://review.openstack.org/104974 | 06:08 |
| *** eglynn has joined #openstack-ceilometer | 06:16 | |
| *** eglynn_ has joined #openstack-ceilometer | 06:25 | |
| *** eglynn has quit IRC | 06:25 | |
| *** nati_ueno has joined #openstack-ceilometer | 06:26 | |
| *** yatin has quit IRC | 06:41 | |
| *** yatin_ has joined #openstack-ceilometer | 06:41 | |
| *** _nadya_ has joined #openstack-ceilometer | 06:50 | |
| *** underyx is now known as Underyx | 07:01 | |
| *** asvechnikov has joined #openstack-ceilometer | 07:04 | |
| *** d0ugal has quit IRC | 07:07 | |
| *** d0ugal has joined #openstack-ceilometer | 07:07 | |
| *** eglynn_ has quit IRC | 07:07 | |
| *** sbfox has quit IRC | 07:20 | |
| *** yatin_ has left #openstack-ceilometer | 07:29 | |
| *** _nadya_ has quit IRC | 07:29 | |
| *** _nadya_ has joined #openstack-ceilometer | 07:30 | |
| *** asvechnikov has quit IRC | 07:49 | |
| *** ChanServ changes topic to "OpenStack Telemetry (Ceilometer) | https://wiki.openstack.org/wiki/Ceilometer" | 07:54 | |
| *** asvechnikov has joined #openstack-ceilometer | 07:56 | |
| *** nati_ueno has quit IRC | 07:56 | |
| *** eglynn_ has joined #openstack-ceilometer | 08:02 | |
| *** eglynn_ is now known as eglynn | 08:07 | |
| openstackgerrit | A change was merged to openstack/ceilometer: Remove redundant parentheses https://review.openstack.org/106345 | 08:08 |
| boris-42 | eglynn hi | 08:08 |
| boris-42 | eglynn something wired is happening with this patch | 08:08 |
| boris-42 | eglynn https://review.openstack.org/#/c/103610/ | 08:08 |
| *** nacim has joined #openstack-ceilometer | 08:08 | |
| boris-42 | eglynn I mean take a look at rechecks | 08:09 |
| boris-42 | eglynn I mean I even don't know why this is happaning=) | 08:09 |
| eglynn | boris-42: py26 is timing out http://logs.openstack.org/10/103610/1/check/gate-ceilometer-python26/32b2b87/console.html | 08:09 |
| boris-42 | eglynn so it's common issue? | 08:10 |
| eglynn | looks like the same timeout error that's been plaguing the tests since the oslo-messaging switchover | 08:10 |
| boris-42 | eglynn maybe rebase can somehow help? | 08:10 |
| eglynn | boris-42: a rebase on what? | 08:11 |
| boris-42 | eglynn that patch | 08:11 |
| boris-42 | eglynn this one https://review.openstack.org/#/c/103610/ | 08:11 |
| eglynn | boris-42: I don't really see how rebasing could help | 08:11 |
| boris-42 | =) | 08:11 |
| boris-42 | eglynn I remember that one time it helped ) | 08:11 |
| eglynn | sileht: are you around? | 08:12 |
| boris-42 | eglynn it was related to some merge crap | 08:12 |
| boris-42 | eglynn e.g. it was merged successfully but with some magic changes=) | 08:12 |
| eglynn | boris-42: yeah, well rebase it anyway for good hygiene | 08:12 |
| boris-42 | eglynn and is this ok https://review.openstack.org/#/c/103420/ ? | 08:12 |
| eglynn | boris-42: ... but that looks like the eventlet issue once again | 08:12 |
| boris-42 | eglynn eventlet strikes again! =) | 08:13 |
| nsaje | eglynn boris-42 : this might be fixed after the oslo.messaging 1.3.1 release | 08:13 |
| boris-42 | nsaje ok nice | 08:13 |
| boris-42 | nsaje btw not sure | 08:13 |
| *** alexpilotti has joined #openstack-ceilometer | 08:13 | |
| boris-42 | nsaje in gates we are using from master oslo.messaging as I know | 08:13 |
| eglynn | boris-42: OK your TODO makes a bit more sense now, ready to fly | 08:14 |
| boris-42 | eglynn and yes I am going to make plugins in future=) | 08:14 |
| boris-42 | eglynn for collecting keystone stuff | 08:14 |
| eglynn | boris-42: https://github.com/openstack/ceilometer/blob/master/requirements.txt#L20 | 08:14 |
| boris-42 | eglynn hm but at least in dsvm stuff | 08:14 |
| boris-42 | eglynn we are using oslo.messaging from master | 08:15 |
| boris-42 | eglynn not sure about py jobs | 08:15 |
| eglynn | boris-42: the timeout failure is in the py26/27 jobs | 08:15 |
| sileht | eglynn, o/ | 08:16 |
| *** safchain has joined #openstack-ceilometer | 08:17 | |
| eglynn | sileht: hey, welcome back :) | 08:17 |
| sileht | eglynn, thanks | 08:17 |
| eglynn | sileht: ... just a heads-up that we're still seeing the oslo-messaging/eventlet timeouts e.g. http://logs.openstack.org/10/103610/1/check/gate-ceilometer-python26/32b2b87/console.html | 08:18 |
| eglynn | sileht: ... but nsaje also found an interesting bug in the oslo-msg fake driver that may be related | 08:19 |
| eglynn | sileht: ... FYI https://bugs.launchpad.net/oslo.messaging/+bug/1342088 | 08:19 |
| sileht | eglynn, looks like a bug I have introduced when I have fixed the reason timeout issue of our tests in oslo-msg...damm | 08:20 |
| eglynn | sileht: sound plausible that it could be the root cause of the more recent failures? | 08:21 |
| *** _nadya_ has quit IRC | 08:22 | |
| sileht | eglynn, yes that depending of who create the fake exchange point if eventlet schedule the server first the test fail, if the client is scheduled first the test pass | 08:24 |
| sileht | eglynn, before my fix when the server start first, the test can pass, but no it always fail ... | 08:25 |
| sileht | no/now | 08:25 |
| eglynn | sileht: a-ha, OK ... so hopefully we'll have a olso-msg 1.3.1 release today to confirm in the gate | 08:25 |
| eglynn | sileht: the other thing is just to confirm that we can declare vistory on https://blueprints.launchpad.net/ceilometer/+spec/dedicated-alarm-database for j2? | 08:26 |
| eglynn | sileht: ... looks like everything that's needed is already landed? :) | 08:26 |
| sileht | eglynn, I have pull out of my big patch every thing I can, now I have to prepare the final one | 08:27 |
| sileht | eglynn, all reviews I have pull out of my bug patch have been merged | 08:28 |
| eglynn | sileht: a-ha, cool, so the final one is do-able for juno-2 do you think? | 08:28 |
| eglynn | juno-2 == July 24th | 08:28 |
| sileht | eglynn, I need to check if I have no hurry thing to do I have missed during my holiday first, but it looks reasonable, I hope to submit the latest patch before monday afternoon | 08:31 |
| boris-42 | eglynn btw I have one question | 08:31 |
| *** inc0 has joined #openstack-ceilometer | 08:31 | |
| boris-42 | eglynn I was asked from community to cover one more use cases related to profiler | 08:31 |
| boris-42 | eglynn e.g. they would like to use "CLI" | 08:31 |
| boris-42 | eglynn and get profiling info | 08:32 |
| boris-42 | eglynn cinder --profile volume create 1 | 08:32 |
| eglynn | sileht: cool, monday would be tight to get the patch shepharded thru gerrit before thurs, but should be doable if the gate doesn't explode on us | 08:32 |
| boris-42 | and it will print what is that (trace id) | 08:32 |
| boris-42 | eglynn so they would like to be able to retrieve this data | 08:33 |
| *** ildikov has quit IRC | 08:33 | |
| boris-42 | eglynn and currently they can only retrieve notification lists | 08:33 |
| boris-42 | eglynn so could we add some command in ceilometer python client that will call osprofiler.parse_notifications() that transforms this list to the tree structure ? | 08:33 |
| eglynn | boris-42: why have a dependency on osprofiler in the ceilo client? | 08:34 |
| boris-42 | eglynn yep that is not super cool | 08:34 |
| eglynn | boris-42: i.e. why not just write a simple python-osprofilerclient? | 08:34 |
| boris-42 | eglynn but I don't know where to put this parser | 08:34 |
| boris-42 | eglynn ou | 08:35 |
| boris-42 | eglynn btw | 08:35 |
| boris-42 | eglynn not so bad idea | 08:35 |
| boris-42 | eglynn I have a lot of stuff that can be put there | 08:35 |
| eglynn | boris-42: cool :) | 08:35 |
| boris-42 | eglynn thanks for idea | 08:35 |
| eglynn | np! | 08:35 |
| *** aviau has quit IRC | 08:44 | |
| *** yatin has joined #openstack-ceilometer | 08:44 | |
| *** aviau has joined #openstack-ceilometer | 08:46 | |
| jd__ | eglynn: https://review.openstack.org/#/c/105185/ waiting for you :) | 08:49 |
| eglynn | jd__: thank you sir! | 08:50 |
| eglynn | jd__: I'll spin up a fresh gnocchi install and give it a whirl this morning | 08:51 |
| jd__ | eglynn: cool! | 08:51 |
| *** IvanBerezovskiy has joined #openstack-ceilometer | 09:23 | |
| *** manish has joined #openstack-ceilometer | 09:24 | |
| manish | Hi... does any one knows, that to get the SWIFT related data, ceilometer uses which REST API? ( if not using the Notification mechanism, but using pollster) | 09:25 |
| manish | is it HEAD method on Account level? | 09:25 |
| openstackgerrit | Nejc Saje proposed a change to openstack/ceilometer: Improves pipeline transformer documentation https://review.openstack.org/107023 | 09:34 |
| *** cdent has joined #openstack-ceilometer | 09:36 | |
| *** yatin_ has joined #openstack-ceilometer | 09:44 | |
| eglynn | nsaje: thanks for the docco patch! | 09:44 |
| eglynn | nsaje: ... though on reading your description of the "Aggregator transformer", I got a sinking feeling :( | 09:45 |
| eglynn | nsaje: ... seems the logic is broken for cumulative meters, or? | 09:45 |
| DinaBelova | eglynn, btw, /me in progress of retention testing in opentsdb :D the thing is that retention param for the timeserie is counted with 'days' so the exact result I'll get ~11pm local time :D | 09:46 |
| *** yatin has quit IRC | 09:47 | |
| nsaje | eglynn: let me check the code | 09:47 |
| DinaBelova | eglynn, I hope to publish the draft before end of week :) | 09:47 |
| eglynn | DinaBelova: a-ha, interesting, so multiples of 1 day is the only way to set the rentention policy in opentsdb? | 09:48 |
| DinaBelova | eglynn, at least it looks so now | 09:48 |
| nsaje | eglynn: could be, though I must admit the purpose of this transformer is not crystal clear to me | 09:48 |
| eglynn | nsaje: https://github.com/openstack/ceilometer/blob/master/ceilometer/transformer/conversions.py#L243 | 09:48 |
| DinaBelova | eglynn, that feature is still developing (for the 2.0 and higher), I'll try to ping some opentsdb folks to understand what they're going to do with it further | 09:49 |
| nsaje | eglynn: I don't see what the use case for it would be with a cumulative meter | 09:49 |
| DinaBelova | eglynn, as 2.x is smth new and currently developing in the opentsdb | 09:49 |
| eglynn | nsaje: ... so I'm guess that should first check if the meter type is cumulative and if so do a straight assignment: self.samples[key].volume = self._scale(sample) | 09:49 |
| eglynn | DinaBelova: cool | 09:49 |
| eglynn | I'm *guessing that should first ... | 09:50 |
| nsaje | eglynn: ah, that could work, yes, do you want me to file a bug and submit a patch? | 09:50 |
| eglynn | nsaje: that would be great :) ... thanl you sir! | 09:50 |
| eglynn | *thank | 09:50 |
| openstackgerrit | A change was merged to openstack/ceilometer: Add keystone control exchange https://review.openstack.org/103420 | 09:51 |
| *** yatin_ has quit IRC | 09:52 | |
| eglynn | boris-42: ^^^ kick back and light a cigar ;) | 09:53 |
| boris-42 | eglynn yaaaa | 09:53 |
| boris-42 | eglynn nice=) | 09:53 |
| eglynn | lesson learned ... if you rebase often enough, eventually the gating Gods will smile upon you! :) | 09:54 |
| cdent | The gate is a dark and murky place | 09:58 |
| *** alexpilotti has quit IRC | 10:09 | |
| *** _nadya_ has joined #openstack-ceilometer | 10:14 | |
| *** eglynn is now known as eglynn-commute | 10:18 | |
| openstackgerrit | Christian Berendt proposed a change to openstack/python-ceilometerclient: Calculate a suitable column width for positional arguments https://review.openstack.org/97881 | 10:19 |
| *** eglynn-commute has quit IRC | 10:24 | |
| openstackgerrit | Christian Berendt proposed a change to openstack/python-ceilometerclient: Calculate a suitable column width for positional arguments https://review.openstack.org/97881 | 10:29 |
| *** nacim has quit IRC | 10:35 | |
| *** vrovachev has joined #openstack-ceilometer | 10:38 | |
| *** yatin_ has joined #openstack-ceilometer | 10:46 | |
| openstackgerrit | A change was merged to openstack/ceilometer: Improve a bit performance of Ceilometer https://review.openstack.org/103610 | 10:51 |
| *** eglynn has joined #openstack-ceilometer | 10:53 | |
| *** yatin_ has quit IRC | 10:55 | |
| openstackgerrit | Chris Dent proposed a change to openstack/ceilometer: Implement consuming ipmi notifications from Ironic https://review.openstack.org/105486 | 10:55 |
| *** Ruetobas has quit IRC | 11:07 | |
| *** Ruetobas has joined #openstack-ceilometer | 11:13 | |
| *** giroro_ has joined #openstack-ceilometer | 11:16 | |
| *** Ruetobas has quit IRC | 11:17 | |
| *** dmsimard has joined #openstack-ceilometer | 11:20 | |
| *** Administrator_ has joined #openstack-ceilometer | 11:25 | |
| *** robsparker has quit IRC | 11:27 | |
| openstackgerrit | Rikimaru Honjo proposed a change to openstack/ceilometer: Split bunch of samples into multiple messages https://review.openstack.org/94155 | 11:40 |
| *** robsparker has joined #openstack-ceilometer | 11:43 | |
| *** ildikov has joined #openstack-ceilometer | 11:47 | |
| *** r3pl4y has joined #openstack-ceilometer | 11:47 | |
| *** Longgeek has joined #openstack-ceilometer | 11:49 | |
| openstackgerrit | Christian Berendt proposed a change to openstack/python-ceilometerclient: Calculate a suitable column width for positional arguments https://review.openstack.org/97881 | 11:51 |
| openstackgerrit | Rikimaru Honjo proposed a change to openstack/ceilometer: Add validate alarm_actions schema in alarm API https://review.openstack.org/107330 | 11:51 |
| *** nacim has joined #openstack-ceilometer | 11:58 | |
| openstackgerrit | Rikimaru Honjo proposed a change to openstack/ceilometer: Replace datetime of time_constraints by aware object https://review.openstack.org/107332 | 12:00 |
| *** jdob has joined #openstack-ceilometer | 12:09 | |
| openstackgerrit | Christian Berendt proposed a change to openstack/pycadf: Bump hacking to version 0.9.2 https://review.openstack.org/107346 | 12:20 |
| *** rbowen has joined #openstack-ceilometer | 12:22 | |
| openstackgerrit | Nejc Saje proposed a change to openstack/ceilometer: Fixes aggregator transformer timestamp and user input handling https://review.openstack.org/107347 | 12:26 |
| *** _nadya_ has quit IRC | 12:38 | |
| *** promulo has quit IRC | 12:40 | |
| *** nosnos has quit IRC | 12:49 | |
| *** julim has joined #openstack-ceilometer | 12:49 | |
| *** cdent_ has joined #openstack-ceilometer | 12:54 | |
| *** cdent has quit IRC | 12:55 | |
| *** cdent_ is now known as cdent | 12:55 | |
| *** thomasem has joined #openstack-ceilometer | 13:13 | |
| openstackgerrit | Eoghan Glynn proposed a change to openstack/ceilometer: Add testcase for multiple discovery-driven sources https://review.openstack.org/105139 | 13:17 |
| *** promulo has joined #openstack-ceilometer | 13:22 | |
| *** fnaval has joined #openstack-ceilometer | 13:22 | |
| *** yatin has joined #openstack-ceilometer | 13:25 | |
| openstackgerrit | Chris Dent proposed a change to openstack/ceilometer: Implement consuming ipmi notifications from Ironic https://review.openstack.org/105486 | 13:32 |
| cdent | I reckon this is good to go if/when it gets some +2s (fixes a rare race in the gate): https://review.openstack.org/#/c/106564/ | 13:35 |
| *** yatin_ has joined #openstack-ceilometer | 13:37 | |
| DinaBelova | eglynn, o/ | 13:39 |
| eglynn | DinaBelova: o/ | 13:39 |
| DinaBelova | eglynn, do you know something about central agents HA status? :) | 13:40 |
| eglynn | DinaBelova: fabiog is on vacation | 13:40 |
| DinaBelova | eglynn, who's working on it? | 13:40 |
| DinaBelova | fabiog - yes, but he was going to ignore tooz that time.. | 13:40 |
| DinaBelova | does he plan to impl it as it was supposed to? | 13:40 |
| *** yatin has quit IRC | 13:41 | |
| eglynn | DinaBelova: we had a long, long discussion about it at mid-cycle on Friday afternoon | 13:41 |
| eglynn | DinaBelova: that was my recommendation | 13:41 |
| DinaBelova | eglynn, yes, I remember this discussion - but I don't remember his answer here.. | 13:41 |
| eglynn | DinaBelova: ... he'll come back with an updated BP when he returns from vacation | 13:41 |
| DinaBelova | eglynn, cool, that's nice :) | 13:42 |
| DinaBelova | eglynn - does he have vacation this week only or some time too? | 13:42 |
| eglynn | DinaBelova: I don't know exactly, but I think he said 2 weeks | 13:43 |
| eglynn | DinaBelova: ... i.e. this week and last week | 13:43 |
| DinaBelova | eglynn, oh, ok | 13:44 |
| ildikov | eglynn: DinaBelova: yeap, he said 10 days, so two weeks, he should be available again next week | 13:44 |
| *** inc0 has quit IRC | 13:45 | |
| DinaBelova | ildikov, eglynn - cool! I just want to say, that if he'll need any help here, central agent HA is quite important for me, and as j2 has passed, I'm really full of energy to help him - or finally do that myself if it'll be impossible for him :) | 13:45 |
| eglynn | DinaBelova: cool, I think having a second contributor involved would be a good think | 13:46 |
| eglynn | good *thing | 13:46 |
| *** psharma has quit IRC | 13:46 | |
| DinaBelova | eglynn, yeah, exactly... I completely miss his vacation from my mind (my fault...) that's why I became really worried.. | 13:47 |
| DinaBelova | if it'll be done at all :D | 13:47 |
| jd__ | eglynn: did you spin up that Gnocchi instance? | 13:48 |
| eglynn | DinaBelova: so if you listened in on does mid-cycle discussions, did my description of horizontal versus vertical slicing thru' the target resource sets make sense to you? | 13:48 |
| *** amalagon has quit IRC | 13:49 | |
| DinaBelova | eglynn, yes, it still seem logical for me - as it'll make uniform load on all the agents running | 13:49 |
| eglynn | jd__: actually I did, but I've also screwed up something dumb in my swift setup :( ... digging into it now | 13:49 |
| eglynn | jd__: I should have something to report shortly | 13:49 |
| jd__ | eglynn: no pressure | 13:50 |
| eglynn | DinaBelova: cool :) | 13:50 |
| DinaBelova | eglynn, as far as I remember idea was to give unifromly pack of tasks to be done to each agent with different target resources | 13:51 |
| DinaBelova | eglynn, not by resource - but multiple resources for each agent | 13:51 |
| *** pradk has joined #openstack-ceilometer | 13:51 | |
| DinaBelova | eglynn, are my rememberings true? :) | 13:51 |
| eglynn | DinaBelova: yep, so instead of agent1 doing all the SNMP polling, agent2 doing all the neutron polling, ... | 13:52 |
| DinaBelova | eglynn, yes, ok - I asked just to be sure everything is ok with my memory :) | 13:53 |
| eglynn | DinaBelova: we have agent1 doing some SNMP polling & some neutron, agent2 doing some neutron polling & some SNMP, ... etc | 13:53 |
| eglynn | DinaBelova: the idea was to drive that slicing via the discovery extensions grabbing from the taskboard (or equivalent mechanism for allocation) | 13:53 |
| DinaBelova | eglynn - that also won't allow to have some time when pack of snmp tasks was not processed at all during some 'fault-tolerant' time | 13:53 |
| DinaBelova | eglynn, that's even more HA that deviding by resource | 13:54 |
| DinaBelova | dividing* | 13:54 |
| DinaBelova | the only problem we may meet here is with efficient zookeeper working, but that's solvable imho (my experience with zookeeper tells me that it likes separated nodes for its work, that's not really comfortable for the small envs - although it was a while ago, now zookeeper is more tolerant) | 13:56 |
| DinaBelova | anyway, that'll be defined during the work :d | 13:57 |
| eglynn | DinaBelova: yep, so smaller allocations of resource type, means say all SNMP polling isn't interrupted during a failover | 13:57 |
| DinaBelova | eglynn, exactly | 13:57 |
| eglynn | DinaBelova: yeah I'm not an expert on ZK, and suitability for small as well as large deployments | 13:57 |
| eglynn | jd__: has there been much adoption of tooz d'ya know? (outside of ceilo) | 13:58 |
| * eglynn just wondering about the experiences with deploying/packaging ZK | 13:59 | |
| *** nati_ueno has joined #openstack-ceilometer | 14:00 | |
| jd__ | eglynn: not yet, though harlowj has plan to use it in taskflow there's a bp for that merged in oslo-specs | 14:01 |
| eglynn | jd__: cool, and ZK is also used directly by nova service groups IIRC | 14:02 |
| jd__ | eglynn: yeah the plan is to change that to tooz at some point | 14:03 |
| jd__ | though nobody's working on that yet | 14:03 |
| *** fnaval has quit IRC | 14:04 | |
| eglynn | when all else fails, the lazy man runs unstack.sh && stack.sh ;) | 14:07 |
| jd__ | running ZK on a single node is pretty straighforward, I don't know about clusters though | 14:08 |
| jd__ | but harlojw might have the answer | 14:08 |
| eglynn | jd__: cool, there are initial efforts to get ZK packaged up also ... is tooz puppetified? | 14:09 |
| eglynn | well ZK being puppetified is probably the correct question to ask | 14:09 |
| jd__ | tooz is a lib so you can't puppetify it I'd say | 14:10 |
| jd__ | yep | 14:10 |
| jd__ | dunnow about that EmilienM might know | 14:10 |
| EmilienM | o/ | 14:12 |
| EmilienM | eglynn: hola | 14:12 |
| openstackgerrit | A change was merged to openstack/ceilometer: Fix call to meter-list in measurements doc https://review.openstack.org/106385 | 14:13 |
| eglynn | cdent: ^^^ \o/ finally :) | 14:13 |
| cdent | that was painful! | 14:14 |
| eglynn | EmilienM: hola! | 14:14 |
| *** ityaptin has quit IRC | 14:14 | |
| openstackgerrit | A change was merged to openstack/ceilometer: Improves pipeline transformer documentation https://review.openstack.org/107023 | 14:17 |
| *** manish has quit IRC | 14:19 | |
| EmilienM | eglynn: looks like many projects: https://forge.puppetlabs.com/tags/zookeeper | 14:19 |
| EmilienM | cdent: hey, we could schedule a meeting for the javelin work | 14:20 |
| *** amalagon has joined #openstack-ceilometer | 14:20 | |
| eglynn | EmilienM: a-ha, cool | 14:20 |
| cdent | Indeed I reckon we should EmilienM, what are we hoping to accomplish? | 14:20 |
| *** tongli has joined #openstack-ceilometer | 14:26 | |
| amalagon | eglynn: o/ | 14:29 |
| *** yatin_ has quit IRC | 14:31 | |
| *** fnaval has joined #openstack-ceilometer | 14:33 | |
| *** Longgeek has quit IRC | 14:34 | |
| *** nacim has quit IRC | 14:35 | |
| *** r3pl4y has quit IRC | 14:37 | |
| eglynn | amalagon: o/ | 14:37 |
| *** cdent_ has joined #openstack-ceilometer | 14:38 | |
| amalagon | eglynn: hey! I wanted to ask you about making the map of the custom aggregates if you have a minute | 14:39 |
| eglynn | amalagon: shoot | 14:39 |
| *** cdent has quit IRC | 14:40 | |
| *** cdent_ is now known as cdent | 14:40 | |
| amalagon | is it ok for the map to be made in the rest/app.py file when pecan is setting up the app? | 14:41 |
| eglynn | amalagon: yeah, IIUC stevedore is a fairly static mechanism for dynamic loading ;) | 14:42 |
| openstackgerrit | Igor Degtiarov proposed a change to openstack/ceilometer: Implement not typed separator in names of columns in HBase https://review.openstack.org/106376 | 14:42 |
| eglynn | amalagon: so it's fine to just load that map once initially on startup | 14:42 |
| eglynn | amalagon: doing it in each get_measures API call is overkill as the entry_points won't have changed .. https://review.openstack.org/#/c/104415/10/gnocchi/rest/__init__.py | 14:43 |
| amalagon | eglynn: yeah...jd pointed that out too - ok, thanks, I'll try reworking it | 14:44 |
| eglynn | amalagon: now that we've got our noses in the same code ... | 14:45 |
| eglynn | amalagon: note also jd__'s comment on "So that means that all custom aggregates are based on mean?" | 14:45 |
| *** nati_ueno has quit IRC | 14:45 | |
| eglynn | amalagon: ... so that was the point I was making before about the custome aggregator *itself* calling down to the storage driver to get the raw data | 14:45 |
| amalagon | eglynn: yes, that is my bad - it should be calling raw data since granularity is 1 (the aggregation='mean' is not doing anything) | 14:46 |
| eglynn | amalagon: ... as only the custome aggregator itself knows whether same the mean-of-means is OK, or whether say mean-of-raw is needed | 14:46 |
| eglynn | amalagon: ... so how about pushing that responsibility to call down to the storage driver into each custom aggregator? | 14:47 |
| eglynn | amalagon: ... the pecan.request.storage object could be passed as param to compute() | 14:48 |
| amalagon | good point | 14:48 |
| cdent | biab | 14:48 |
| eglynn | cool | 14:48 |
| amalagon | although, just to quibble... | 14:50 |
| amalagon | I think that all the custom aggregators are taking in raw data as of now | 14:50 |
| amalagon | ... I forgot what the rest of my argument was | 14:51 |
| amalagon | nm | 14:51 |
| eglynn | amalagon: :) | 14:51 |
| eglynn | amalagon: you could be correct though | 14:51 |
| eglynn | amalagon: my thought was that *some* aggregations could possibly optimize if they know they don't need the most granular data available | 14:52 |
| eglynn | amalagon: but I'm not sure how realistic that notion is | 14:53 |
| *** _nadya_ has joined #openstack-ceilometer | 14:53 | |
| *** _nadya_ has quit IRC | 14:53 | |
| amalagon | eglynn: the optimization would come from not calling the raw data? | 14:54 |
| eglynn | amalagon: grabbing fewer datapoints from the storage backend, getting it to roll-up before pushing that data back to gnocchi | 14:55 |
| eglynn | amalagon: ... so say moving-window average was generally calculated as a mean-of-means | 14:55 |
| amalagon | but then if the other custom aggregates are going to need the raw data (ewma, moving-variance) | 14:55 |
| eglynn | amalagon: sure, so each one can decide itself what it needs | 14:55 |
| eglynn | amalagon: because the call down to the storage driver could occur *with* the custom aggregator itself | 14:56 |
| eglynn | amalagon: ... as opposed to being done for it by the API layer | 14:56 |
| amalagon | eglynn: ah yes, sorry...I keep on thinking the aggregates are working on the same retrieved data, but it has to be re-retrieved for each request..I have a very persistent wrong mental picture | 14:57 |
| eglynn | amalagon: think of each API request as a kind of clean slate | 14:57 |
| amalagon | gotcha | 14:58 |
| amalagon | I think I'm convinced now; I'll try working on this a bit | 15:00 |
| eglynn | amalagon: excellent :) | 15:01 |
| eglynn | amalagon: ... if you wanna to play around a bit with those ideas in code, just push a new patchset & ping me whenever you're ready for another review pass | 15:01 |
| amalagon | good | 15:02 |
| *** r3pl4y has joined #openstack-ceilometer | 15:04 | |
| *** ifarkas has quit IRC | 15:15 | |
| *** nati_ueno has joined #openstack-ceilometer | 15:28 | |
| *** vrovachev has quit IRC | 15:35 | |
| *** nati_ueno has quit IRC | 15:35 | |
| *** nati_ueno has joined #openstack-ceilometer | 15:38 | |
| *** ildikov_ has joined #openstack-ceilometer | 15:41 | |
| *** nati_ueno has quit IRC | 15:43 | |
| *** ildikov has quit IRC | 15:43 | |
| *** nati_ueno has joined #openstack-ceilometer | 15:43 | |
| EmilienM | cdent: we need first to fix javelin | 15:48 |
| EmilienM | so I did a couple of patches with features | 15:48 |
| EmilienM | but the basic does not work well in the OpenStack CI | 15:48 |
| cdent | Have you seen the recent changes from Sean that add a -d arg to javelin? | 15:49 |
| EmilienM | cdent: looking | 15:52 |
| EmilienM | cdent: the imgdir? | 15:52 |
| * cdent nods | 15:53 | |
| EmilienM | cdent: I missed that one, good to know | 15:53 |
| cdent | that seems pretty important as it helps ground any paths that the resources.yaml needs | 15:53 |
| EmilienM | cdent: did you already played with javelin2 ? | 15:54 |
| EmilienM | yeah right | 15:54 |
| cdent | yes, have you seen: https://review.openstack.org/#/c/102354/ | 15:54 |
| *** _nadya_ has joined #openstack-ceilometer | 15:54 | |
| EmilienM | not yet, | 15:55 |
| EmilienM | looking | 15:55 |
| EmilienM | cdent: nice one | 15:56 |
| cdent | I've got a few queries on another review too: | 15:59 |
| cdent | https://review.openstack.org/#/c/97317/ | 15:59 |
| *** giroro_ has quit IRC | 16:01 | |
| *** nacim has joined #openstack-ceilometer | 16:02 | |
| *** Ruetobas has joined #openstack-ceilometer | 16:03 | |
| *** Ruetobas has quit IRC | 16:08 | |
| *** nati_ueno has quit IRC | 16:09 | |
| *** Ruetobas has joined #openstack-ceilometer | 16:14 | |
| eglynn | jd__: I've suggested a simple fix inline for the standard deviation issue in that gnocchi patch | 16:17 |
| *** sbfox has joined #openstack-ceilometer | 16:17 | |
| eglynn | jd__: ... but haven't had time yet to kick the tyres on the shiny new truncation logic | 16:17 |
| *** _cjones_ has joined #openstack-ceilometer | 16:18 | |
| *** yjiang5 has quit IRC | 16:29 | |
| *** safchain has quit IRC | 16:30 | |
| *** _nadya_ has quit IRC | 16:33 | |
| *** robbybb has joined #openstack-ceilometer | 16:45 | |
| *** nacim has quit IRC | 16:47 | |
| *** hjrii has joined #openstack-ceilometer | 16:55 | |
| eglynn | pradk: around? | 16:56 |
| *** sbfox has quit IRC | 17:01 | |
| *** cdent_ has joined #openstack-ceilometer | 17:06 | |
| *** cdent_ has quit IRC | 17:08 | |
| *** cdent has quit IRC | 17:10 | |
| *** sbfox has joined #openstack-ceilometer | 17:11 | |
| *** nati_ueno has joined #openstack-ceilometer | 17:11 | |
| *** cdent has joined #openstack-ceilometer | 17:11 | |
| *** jaypipes has quit IRC | 17:12 | |
| *** sbfox1 has joined #openstack-ceilometer | 17:14 | |
| *** sbfox has quit IRC | 17:15 | |
| *** openstackgerrit has quit IRC | 17:16 | |
| *** openstackgerrit has joined #openstack-ceilometer | 17:17 | |
| *** harlowja_away is now known as harlowja | 17:18 | |
| *** harlowja is now known as harlowja_away | 17:19 | |
| *** harlowja_away is now known as harlowja | 17:19 | |
| *** jaypipes has joined #openstack-ceilometer | 17:30 | |
| *** sbfox1 has quit IRC | 17:31 | |
| *** sbfox has joined #openstack-ceilometer | 17:31 | |
| *** r3pl4y has quit IRC | 17:37 | |
| *** Ruetobas has quit IRC | 17:38 | |
| *** shakamunyi has joined #openstack-ceilometer | 17:39 | |
| *** shakamunyi has quit IRC | 17:39 | |
| *** shakamunyi has joined #openstack-ceilometer | 17:40 | |
| *** eglynn has quit IRC | 17:41 | |
| *** shakamunyi has quit IRC | 17:42 | |
| pradk | eglynn: hi, thx for the mail | 17:42 |
| *** ifarkas has joined #openstack-ceilometer | 17:42 | |
| *** cdent_ has joined #openstack-ceilometer | 17:44 | |
| *** fnaval has quit IRC | 17:44 | |
| *** Ruetobas has joined #openstack-ceilometer | 17:44 | |
| *** cdent has quit IRC | 17:45 | |
| *** cdent_ is now known as cdent | 17:45 | |
| *** tongli has quit IRC | 17:46 | |
| *** Ruetobas has quit IRC | 17:48 | |
| *** Ruetobas has joined #openstack-ceilometer | 17:55 | |
| *** sbfox has quit IRC | 17:55 | |
| *** cdent has quit IRC | 17:56 | |
| *** sbfox has joined #openstack-ceilometer | 17:58 | |
| *** dmsimard is now known as dmsimard_away | 18:05 | |
| *** aviau has quit IRC | 18:13 | |
| *** aviau has joined #openstack-ceilometer | 18:14 | |
| *** dmsimard_away is now known as dmsimard | 18:23 | |
| *** _nadya_ has joined #openstack-ceilometer | 18:25 | |
| *** dmsimard is now known as dmsimard_away | 18:33 | |
| *** dmsimard_away is now known as dmsimard | 18:39 | |
| openstackgerrit | Pradeep Kilambi proposed a change to openstack/ceilometer: Use resource discovery for Network Services https://review.openstack.org/103956 | 18:52 |
| sbfox | Hey ceilometer folks, is there a specific neutron agent like the nova compute agent? | 18:58 |
| *** jmckind has joined #openstack-ceilometer | 19:09 | |
| *** ildikov has joined #openstack-ceilometer | 19:10 | |
| *** fnaval has joined #openstack-ceilometer | 19:13 | |
| *** ildikov_ has quit IRC | 19:13 | |
| *** _nadya_ has quit IRC | 19:18 | |
| *** _nadya_ has joined #openstack-ceilometer | 19:19 | |
| *** _nadya_ has quit IRC | 19:24 | |
| *** cvip3r has joined #openstack-ceilometer | 19:51 | |
| cvip3r | Hi, how can I get the total size of data gathere by the ceilometer so far? | 19:52 |
| cvip3r | any help will be appreciated.. thanks | 19:52 |
| cvip3r | Via ceilomeet CLI I am able to see the collected samples.. but when I try to connect via mongo ceilometer I get an error | 19:55 |
| cvip3r | Error: couldn't connect to server 127.0.0.1:27017 at src/mongo/shell/mongo.js:145 | 19:55 |
| cvip3r | ping anyone... | 19:56 |
| *** hjrii has left #openstack-ceilometer | 19:59 | |
| *** dmsimard is now known as dmsimard_away | 20:02 | |
| cvip3r | Got it I had to use the server mentioned in conf file | 20:06 |
| cvip3r | thanks anyways | 20:06 |
| *** ifarkas has quit IRC | 20:15 | |
| *** Ruetobas has quit IRC | 20:22 | |
| *** Ruetobas has joined #openstack-ceilometer | 20:28 | |
| *** Ruetobas has quit IRC | 20:33 | |
| *** Ruetobas has joined #openstack-ceilometer | 20:33 | |
| *** sbfox has quit IRC | 20:34 | |
| *** promulo has quit IRC | 20:37 | |
| *** thomasem has quit IRC | 20:40 | |
| *** dmsimard_away is now known as dmsimard | 20:45 | |
| *** dmsimard is now known as dmsimard_away | 20:47 | |
| *** ccrouch has quit IRC | 20:57 | |
| *** mgarza has joined #openstack-ceilometer | 21:00 | |
| *** jdob has quit IRC | 21:07 | |
| *** ccrouch has joined #openstack-ceilometer | 21:09 | |
| *** ccrouch1 has joined #openstack-ceilometer | 21:16 | |
| *** cvip3r has quit IRC | 21:18 | |
| *** ccrouch has quit IRC | 21:18 | |
| *** shardy is now known as shardy_afk | 21:20 | |
| *** jmckind has quit IRC | 21:21 | |
| *** mrmoje has joined #openstack-ceilometer | 21:31 | |
| *** joshw42 has quit IRC | 21:41 | |
| *** thomasem has joined #openstack-ceilometer | 21:42 | |
| *** mrmoje has quit IRC | 21:43 | |
| *** fnaval has quit IRC | 21:45 | |
| *** rbowen has quit IRC | 21:45 | |
| *** julim has quit IRC | 21:46 | |
| *** fnaval has joined #openstack-ceilometer | 21:53 | |
| *** Ruetobas has quit IRC | 22:04 | |
| *** Ruetobas has joined #openstack-ceilometer | 22:10 | |
| *** giroro_ has joined #openstack-ceilometer | 22:13 | |
| *** Ruetobas has quit IRC | 22:14 | |
| *** pradk has quit IRC | 22:29 | |
| *** dobson has joined #openstack-ceilometer | 22:39 | |
| *** ivan has left #openstack-ceilometer | 22:39 | |
| *** sbfox has joined #openstack-ceilometer | 22:40 | |
| *** thomasem has quit IRC | 22:53 | |
| *** thomasem has joined #openstack-ceilometer | 22:53 | |
| *** thomasem has quit IRC | 22:57 | |
| *** mrmoje has joined #openstack-ceilometer | 22:59 | |
| *** thomasem has joined #openstack-ceilometer | 23:03 | |
| *** mrmoje has quit IRC | 23:03 | |
| *** mrmoje has joined #openstack-ceilometer | 23:05 | |
| *** thomasem has quit IRC | 23:07 | |
| *** mrmoje has quit IRC | 23:10 | |
| *** fnaval has quit IRC | 23:13 | |
| *** sbfox has quit IRC | 23:21 | |
| *** ccrouch1 has quit IRC | 23:29 | |
| *** mgarza has quit IRC | 23:42 | |
| *** mgarza has joined #openstack-ceilometer | 23:43 | |
| *** mgarza has quit IRC | 23:45 | |
| *** mgarza has joined #openstack-ceilometer | 23:45 | |
| *** mgarza has quit IRC | 23:47 | |
| *** mgarza has joined #openstack-ceilometer | 23:48 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!