| *** PsionTheory has quit IRC | 00:19 | |
| *** r-daneel has joined #openstack-telemetry | 00:23 | |
| *** farid has quit IRC | 00:37 | |
| *** chlong has joined #openstack-telemetry | 00:40 | |
| openstackgerrit | Merged openstack/ceilometer: Remove unused pytz requirement https://review.openstack.org/275350 | 00:50 |
|---|---|---|
| *** cheneydc has joined #openstack-telemetry | 00:57 | |
| *** Ephur has quit IRC | 01:40 | |
| *** r-daneel has quit IRC | 01:41 | |
| *** mgarza_ has joined #openstack-telemetry | 01:55 | |
| *** farid has joined #openstack-telemetry | 02:03 | |
| *** alejandrito has joined #openstack-telemetry | 02:20 | |
| *** ljxiash has joined #openstack-telemetry | 02:20 | |
| *** ljxiash has quit IRC | 02:25 | |
| *** ljxiash has joined #openstack-telemetry | 02:27 | |
| openstackgerrit | Merged openstack/ceilometer: devstack: use password with version discovery https://review.openstack.org/274539 | 02:43 |
| *** prashantD has quit IRC | 02:50 | |
| *** achatterjee has quit IRC | 02:52 | |
| *** thorst has joined #openstack-telemetry | 02:52 | |
| *** achatterjee has joined #openstack-telemetry | 02:53 | |
| openstackgerrit | liusheng proposed openstack/aodh: Add composite rule alarm evaluator https://review.openstack.org/261226 | 03:10 |
| *** thorst has quit IRC | 03:17 | |
| *** alejandrito has quit IRC | 03:17 | |
| *** thorst has joined #openstack-telemetry | 03:17 | |
| openstackgerrit | liusheng proposed openstack/aodh: Add composite rule alarm evaluator https://review.openstack.org/261226 | 03:25 |
| *** thorst has quit IRC | 03:27 | |
| *** thumpba has joined #openstack-telemetry | 03:34 | |
| *** cheneydc_ has joined #openstack-telemetry | 03:40 | |
| *** farid has quit IRC | 03:43 | |
| *** links has joined #openstack-telemetry | 03:49 | |
| *** thumpba has quit IRC | 03:53 | |
| *** thumpba has joined #openstack-telemetry | 03:54 | |
| *** mgarza_ has quit IRC | 04:06 | |
| *** boris-42 has quit IRC | 04:11 | |
| *** ryanpetrello has quit IRC | 04:11 | |
| *** thumpba_ has joined #openstack-telemetry | 04:12 | |
| *** datravis has quit IRC | 04:12 | |
| *** tpeoples has quit IRC | 04:12 | |
| *** thumpba has quit IRC | 04:13 | |
| *** jroll has quit IRC | 04:13 | |
| *** davidlenwell_ has joined #openstack-telemetry | 04:13 | |
| *** davidlenwell has quit IRC | 04:13 | |
| *** chlong has quit IRC | 04:13 | |
| *** ryanpetrello has joined #openstack-telemetry | 04:14 | |
| *** jroll has joined #openstack-telemetry | 04:14 | |
| *** datravis has joined #openstack-telemetry | 04:14 | |
| *** tpeoples has joined #openstack-telemetry | 04:17 | |
| *** alextricity25_ has joined #openstack-telemetry | 04:17 | |
| *** boris-42 has joined #openstack-telemetry | 04:18 | |
| *** pasquier-s has quit IRC | 04:19 | |
| *** alextricity25 has quit IRC | 04:19 | |
| *** elemoine has quit IRC | 04:19 | |
| *** pasquier-s has joined #openstack-telemetry | 04:20 | |
| *** elemoine has joined #openstack-telemetry | 04:20 | |
| *** thumpba_ has quit IRC | 04:25 | |
| *** thorst has joined #openstack-telemetry | 04:27 | |
| *** chlong has joined #openstack-telemetry | 04:28 | |
| *** thorst has quit IRC | 04:32 | |
| *** freealvin has joined #openstack-telemetry | 04:39 | |
| *** davidlenwell_ has quit IRC | 05:02 | |
| *** davidlenwell has joined #openstack-telemetry | 05:05 | |
| openstackgerrit | Zi Lian Ji proposed openstack/ceilometer: Extend validation for new generic SNMP pollster https://review.openstack.org/262660 | 05:08 |
| *** boris-42 has quit IRC | 05:13 | |
| openstackgerrit | Zi Lian Ji proposed openstack/ceilometer: Extend validation for new generic SNMP pollster https://review.openstack.org/262660 | 05:23 |
| *** liamji has joined #openstack-telemetry | 05:24 | |
| *** thorst has joined #openstack-telemetry | 05:31 | |
| *** jwcroppe has joined #openstack-telemetry | 05:33 | |
| *** thorst has quit IRC | 05:37 | |
| *** ljxiash has quit IRC | 05:42 | |
| *** ljxiash has joined #openstack-telemetry | 05:42 | |
| *** ljxiash has quit IRC | 05:46 | |
| *** ljxiash has joined #openstack-telemetry | 05:50 | |
| *** ljxiash has quit IRC | 05:55 | |
| *** ljxiash has joined #openstack-telemetry | 05:56 | |
| *** liamji has quit IRC | 05:59 | |
| *** ljxiash has quit IRC | 06:00 | |
| *** ljxiash has joined #openstack-telemetry | 06:01 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ceilometer: Imported Translations from Zanata https://review.openstack.org/273346 | 06:08 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/python-gnocchiclient: search: Ensure we didn't ignore extra chars https://review.openstack.org/275643 | 06:20 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/python-gnocchiclient: Add batching to client https://review.openstack.org/273429 | 06:20 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/python-gnocchiclient: Ensure description exists for HTTP error. https://review.openstack.org/273428 | 06:20 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/python-gnocchiclient: The search query is mandatory https://review.openstack.org/275625 | 06:20 |
| *** tongli has quit IRC | 06:30 | |
| openstackgerrit | Merged openstack/python-gnocchiclient: search: Ensure we didn't ignore extra chars https://review.openstack.org/275643 | 06:43 |
| *** jwcroppe has quit IRC | 06:51 | |
| *** jwcroppe has joined #openstack-telemetry | 06:52 | |
| *** jwcroppe has quit IRC | 06:56 | |
| *** belmoreira has joined #openstack-telemetry | 07:02 | |
| *** safchain has joined #openstack-telemetry | 07:37 | |
| *** thorst has joined #openstack-telemetry | 07:43 | |
| *** _nadya_ has joined #openstack-telemetry | 07:43 | |
| *** ljxiash has quit IRC | 07:47 | |
| *** ljxiash has joined #openstack-telemetry | 07:48 | |
| *** thorst has quit IRC | 07:48 | |
| *** ljxiash has quit IRC | 07:52 | |
| *** eglynn has joined #openstack-telemetry | 07:54 | |
| *** ljxiash has joined #openstack-telemetry | 08:02 | |
| *** sinese has joined #openstack-telemetry | 08:05 | |
| *** ljxiash has quit IRC | 08:07 | |
| *** eglynn has quit IRC | 08:17 | |
| *** eglynn has joined #openstack-telemetry | 08:17 | |
| *** ljxiash has joined #openstack-telemetry | 08:29 | |
| *** ljxiash has quit IRC | 08:30 | |
| *** ljxiash has joined #openstack-telemetry | 08:30 | |
| *** achatterjee has quit IRC | 08:32 | |
| *** thorst has joined #openstack-telemetry | 08:46 | |
| *** thorst has quit IRC | 08:52 | |
| * sileht writes a simple migration script that need one more times to be tweaked by stupid mysql limitation... | 08:54 | |
| *** achatterjee has joined #openstack-telemetry | 09:01 | |
| openstackgerrit | Julien Danjou proposed openstack/ceilometer: Use overtest to setup functional backends https://review.openstack.org/275738 | 09:01 |
| *** rbak has quit IRC | 09:04 | |
| openstackgerrit | Julien Danjou proposed openstack/ceilometer: sample: remove deprecated option name. https://review.openstack.org/276102 | 09:05 |
| openstackgerrit | Julien Danjou proposed openstack/ceilometer: network: remove deprecated option name https://review.openstack.org/276104 | 09:06 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/python-gnocchiclient: The search query is mandatory https://review.openstack.org/275625 | 09:15 |
| *** liamji has joined #openstack-telemetry | 09:17 | |
| *** openstackgerrit has quit IRC | 09:17 | |
| *** openstackgerrit has joined #openstack-telemetry | 09:18 | |
| *** yassine has joined #openstack-telemetry | 09:21 | |
| *** ljxiash has quit IRC | 09:24 | |
| *** efoley has joined #openstack-telemetry | 09:24 | |
| *** ljxiash has joined #openstack-telemetry | 09:25 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Add some resource types tests https://review.openstack.org/270419 | 09:27 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Implements resource attribute number https://review.openstack.org/270091 | 09:27 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Implements resource attribute bool https://review.openstack.org/270418 | 09:27 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Implements resource attribute uuid https://review.openstack.org/270090 | 09:27 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Don't create Ceilometer resource types by default. https://review.openstack.org/270322 | 09:27 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Move legacy Ceilometer resource into indexer. https://review.openstack.org/270266 | 09:27 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Move resource type into their own sql table https://review.openstack.org/269843 | 09:27 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Implements resource type CRUD. https://review.openstack.org/269844 | 09:27 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Implements resource attribute string https://review.openstack.org/269888 | 09:27 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Remove useless indexes https://review.openstack.org/276110 | 09:27 |
| sileht | jd__, can you take a look to this: https://review.openstack.org/276110 | 09:27 |
| *** ljxiash has quit IRC | 09:29 | |
| jd__ | sileht: lgtm; added gordc just in acse | 09:29 |
| sileht | jd__, the issue was pointed by him :) | 09:29 |
| jd__ | I thought so | 09:31 |
| openstackgerrit | Merged openstack/python-gnocchiclient: The search query is mandatory https://review.openstack.org/275625 | 09:36 |
| *** liamji has quit IRC | 09:46 | |
| *** thorst has joined #openstack-telemetry | 09:51 | |
| *** thorst has quit IRC | 09:57 | |
| *** cheneydc has quit IRC | 10:02 | |
| *** efoley_ has joined #openstack-telemetry | 10:04 | |
| *** efoley has quit IRC | 10:05 | |
| *** ildikov has quit IRC | 10:07 | |
| *** lsmola_ has quit IRC | 10:12 | |
| openstackgerrit | Merged openstack/gnocchi: tests: remove measures reporting test https://review.openstack.org/275063 | 10:17 |
| openstackgerrit | Merged openstack/gnocchi: swift, ceph: take smaller batch of new measures to process https://review.openstack.org/274843 | 10:22 |
| *** idegtiarov has joined #openstack-telemetry | 10:23 | |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: indexer: replace get_metrics() by list_metrics() https://review.openstack.org/274683 | 10:30 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: rest: implement groupby in resource/metric aggregation https://review.openstack.org/274684 | 10:30 |
| *** jdandrea_ has joined #openstack-telemetry | 10:37 | |
| *** jdandrea has quit IRC | 10:40 | |
| *** thorst has joined #openstack-telemetry | 10:55 | |
| *** bdossant has joined #openstack-telemetry | 11:00 | |
| *** thorst has quit IRC | 11:03 | |
| *** eglynn has quit IRC | 11:04 | |
| *** efoley__ has joined #openstack-telemetry | 11:06 | |
| *** efoley__ is now known as efoley | 11:07 | |
| *** efoley_ has quit IRC | 11:08 | |
| efoley | Can Aodh alarms be used with Gnocchi as the metric source? | 11:10 |
| *** lsmola has joined #openstack-telemetry | 11:19 | |
| *** _nadya_ has quit IRC | 11:25 | |
| *** ljxiash has joined #openstack-telemetry | 11:25 | |
| *** ljxiash has quit IRC | 11:30 | |
| *** efoley has quit IRC | 11:40 | |
| *** efoley has joined #openstack-telemetry | 11:41 | |
| openstackgerrit | Merged openstack/ceilometer: Fix wrong capitalization https://review.openstack.org/275533 | 11:54 |
| *** cdent has joined #openstack-telemetry | 12:00 | |
| *** efoley_ has joined #openstack-telemetry | 12:01 | |
| sileht | efoley: of course | 12:02 |
| *** efoley has quit IRC | 12:05 | |
| *** _nadya_ has joined #openstack-telemetry | 12:08 | |
| *** ildikov has joined #openstack-telemetry | 12:09 | |
| *** liamji has joined #openstack-telemetry | 12:27 | |
| *** efoley_ has quit IRC | 12:28 | |
| *** efoley_ has joined #openstack-telemetry | 12:32 | |
| *** thorst has joined #openstack-telemetry | 12:38 | |
| *** efoley_ is now known as efoley | 12:41 | |
| efoley | sileht: Cool, thanks, just want to make sure before I go and cause a lot of trouble | 12:43 |
| *** bdossant has quit IRC | 12:45 | |
| *** eglynn has joined #openstack-telemetry | 12:46 | |
| *** elemoine- has joined #openstack-telemetry | 12:52 | |
| *** achatterjee has quit IRC | 12:53 | |
| *** guillaume__ has quit IRC | 12:53 | |
| *** elemoine has quit IRC | 12:53 | |
| *** guillaume__ has joined #openstack-telemetry | 12:54 | |
| *** bdossant has joined #openstack-telemetry | 13:05 | |
| *** bdossant has quit IRC | 13:10 | |
| *** bdossant has joined #openstack-telemetry | 13:10 | |
| *** peristeri has joined #openstack-telemetry | 13:15 | |
| *** peristeri has quit IRC | 13:15 | |
| *** peristeri has joined #openstack-telemetry | 13:16 | |
| *** gordc has joined #openstack-telemetry | 13:22 | |
| *** sinese has quit IRC | 13:28 | |
| *** sinese has joined #openstack-telemetry | 13:29 | |
| *** sinese has quit IRC | 13:29 | |
| openstackgerrit | gordon chung proposed openstack/ceilometer: use retrying to attempt to rejoin group https://review.openstack.org/275806 | 13:30 |
| *** rbak has joined #openstack-telemetry | 13:31 | |
| *** efoley_ has joined #openstack-telemetry | 13:33 | |
| *** cdent has quit IRC | 13:33 | |
| *** ildikov has quit IRC | 13:34 | |
| *** efoley has quit IRC | 13:35 | |
| openstackgerrit | gordon chung proposed openstack/ceilometer: use retrying to attempt to rejoin group https://review.openstack.org/275806 | 13:36 |
| *** david-lyle has quit IRC | 13:38 | |
| *** pradk has joined #openstack-telemetry | 13:42 | |
| *** nicodemus_ has joined #openstack-telemetry | 13:46 | |
| *** _nadya_ has quit IRC | 13:53 | |
| *** links has quit IRC | 13:56 | |
| *** ildikov has joined #openstack-telemetry | 14:01 | |
| nicodemus_ | hello gordc, do you have a moment? | 14:01 |
| *** barra204 has quit IRC | 14:04 | |
| *** shakamunyi has joined #openstack-telemetry | 14:06 | |
| *** bdossant_ has joined #openstack-telemetry | 14:07 | |
| *** julim has joined #openstack-telemetry | 14:07 | |
| *** bdossant has quit IRC | 14:10 | |
| *** efoley_ has quit IRC | 14:11 | |
| gordc | nicodemus_: sure. whatsup? | 14:11 |
| nicodemus_ | I wanted to ask you a question about an open bug: https://bugs.launchpad.net/ceilometer/+bug/1457440 | 14:12 |
| openstack | Launchpad bug 1457440 in Ceilometer "Compute agent virDomainGetBlockInfo error for instances with RBD backend" [Medium,Triaged] | 14:12 |
| nicodemus_ | would you happen to know if there were any sort of advance after the last comment? | 14:13 |
| gordc | nicodemus_: darn. that got lost during holidays. let me try pinging someone | 14:13 |
| *** yassine has quit IRC | 14:13 | |
| *** alextricity25_ is now known as alextricity25 | 14:14 | |
| nicodemus_ | gordc: Thanks! Cause I'm seeing that behavior as well | 14:14 |
| *** bdossant_ has quit IRC | 14:15 | |
| *** bdossant has joined #openstack-telemetry | 14:16 | |
| gordc | stuff like this would be a lot easier if nova folks owned their own metrics rather than rely on telemetry folks to be experts in all other projects... sigh | 14:17 |
| gordc | nicodemus_: my nova contact isn't online. i'll need to get back to you. | 14:18 |
| nicodemus_ | gordc: sure, thanks! | 14:19 |
| openstackgerrit | Igor Degtiarov proposed openstack/ceilometer: [MongoDB] exchange compound index with single field indexes https://review.openstack.org/276262 | 14:21 |
| *** guillaume__ has quit IRC | 14:26 | |
| gordc | peristeri: is vlad around? | 14:38 |
| openstackgerrit | Igor Degtiarov proposed openstack/ceilometer-specs: Transformers for events pipeline https://review.openstack.org/162167 | 14:40 |
| peristeri | gordc, I | 14:42 |
| peristeri | gordc, I'm working from home | 14:42 |
| gordc | peristeri: nice. nm then. | 14:43 |
| peristeri | gordc, when are you passing by Mtl? | 14:43 |
| gordc | had a libvirt question. | 14:43 |
| gordc | peristeri: when via rail gets a train that wasn't made pre-WW1.lol | 14:44 |
| gordc | peristeri: did they roll out new subways yet? | 14:45 |
| nicodemus_ | jd__, I'm doing some performance tests today | 14:47 |
| nicodemus_ | let me know if you'd like to know the results | 14:48 |
| *** ljxiash has joined #openstack-telemetry | 14:49 | |
| *** _nadya_ has joined #openstack-telemetry | 14:49 | |
| peristeri | gordc, sometime this year maybe. | 14:50 |
| gordc | lol | 14:50 |
| *** efoley_ has joined #openstack-telemetry | 14:51 | |
| jd__ | nicodemus_: obviously :) | 14:52 |
| nicodemus_ | briefly what I see so far: four instances running two metricd workers each, can't keep up with the measures produced by two ceilometer-agent-compute and one ceilometer-agent-central | 14:54 |
| nicodemus_ | I'm looking for the magic number of workers/metricd instances | 14:55 |
| *** efoley_ has quit IRC | 14:56 | |
| *** pradk has quit IRC | 14:56 | |
| *** pradk has joined #openstack-telemetry | 14:56 | |
| gordc | nicodemus_: with current master? | 14:56 |
| nicodemus_ | gordc: from Jan 27 | 14:57 |
| gordc | nicodemus_: you may want to try latest. we have a patch to attempt to handle the large backlog issue | 14:58 |
| gordc | not a full solution but a first step | 14:59 |
| gordc | meeting in 1 minutes folks | 14:59 |
| *** ityaptin has joined #openstack-telemetry | 14:59 | |
| nicodemus_ | gordc: interesting, didn't know that. However, I had to delete the *huge* 7,5M backlog I had | 14:59 |
| nicodemus_ | now I have a clean, empty ceph gnocchi pool | 14:59 |
| *** ljxiash has quit IRC | 14:59 | |
| gordc | ack. i think it will still grow regardless (once it hits a tipping point of 'this backlog is too big') | 15:00 |
| gordc | jd__: sileht: pradk: | 15:02 |
| jd__ | nicodemus_: yeah, what gordc just said | 15:03 |
| jd__ | nicodemus_: we wrote a few patches to improve that situation | 15:03 |
| *** efoley has joined #openstack-telemetry | 15:03 | |
| jd__ | so I'd suggest you give it another try with latest master :) | 15:03 |
| nicodemus_ | jd__: will do, I'll finish collecting perf numbers with the current version I have as to compare later :) | 15:04 |
| jd__ | nicodemus_: good idea | 15:04 |
| *** KrishR has joined #openstack-telemetry | 15:07 | |
| *** ljxiash has joined #openstack-telemetry | 15:09 | |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: indexer: replace get_metrics() by list_metrics() https://review.openstack.org/274683 | 15:12 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: rest: implement groupby in resource/metric aggregation https://review.openstack.org/274684 | 15:12 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: implement full listing for new measures https://review.openstack.org/276289 | 15:12 |
| ljxiash | o/ | 15:14 |
| *** jwcroppe has joined #openstack-telemetry | 15:14 | |
| *** guillaume__ has joined #openstack-telemetry | 15:24 | |
| *** mgarza has joined #openstack-telemetry | 15:29 | |
| *** ljxiash has quit IRC | 15:32 | |
| *** ljxiash has joined #openstack-telemetry | 15:32 | |
| *** julim has quit IRC | 15:33 | |
| *** julim has joined #openstack-telemetry | 15:35 | |
| ljxiash | gordc: hmm.. yeah, our plan is if we can get it all in in this cycle, it would be very good, because we already have basic code. We will find more people review it. | 15:50 |
| gordc | ljxiash: ok. that's fine. as long as we can get it up for review soon. the last week of cycle is usually impossible to merge anything as gate is usually broken | 15:50 |
| gordc | so really there's only 2 weeks left. | 15:51 |
| ljxiash | gordc: Thanks for reminding me this. And thanks for your support ! :) | 15:52 |
| gordc | ljxiash: np | 15:53 |
| *** ddieterly has joined #openstack-telemetry | 16:00 | |
| *** _nadya_ has quit IRC | 16:00 | |
| *** liusheng has quit IRC | 16:02 | |
| *** r-daneel has joined #openstack-telemetry | 16:04 | |
| *** liamji has quit IRC | 16:11 | |
| *** liamji has joined #openstack-telemetry | 16:12 | |
| *** farid_ has joined #openstack-telemetry | 16:14 | |
| *** ljxiash has quit IRC | 16:15 | |
| *** ljxiash has joined #openstack-telemetry | 16:16 | |
| openstackgerrit | gordon chung proposed openstack/ceilometer: set higher batching requirement https://review.openstack.org/276330 | 16:22 |
| *** david-lyle has joined #openstack-telemetry | 16:24 | |
| *** shardy has joined #openstack-telemetry | 16:30 | |
| *** belmoreira has quit IRC | 16:34 | |
| *** liamji has quit IRC | 16:36 | |
| *** eglynn has quit IRC | 16:36 | |
| *** ljxiash has quit IRC | 16:41 | |
| openstackgerrit | Pradeep Kilambi proposed openstack/ceilometer: Skip duplicate meter definitions https://review.openstack.org/274860 | 16:49 |
| * pradk pokes sileht | 16:50 | |
| pradk | :) | 16:50 |
| pradk | sileht, so how do i enable gnocchi in functional tests? | 16:50 |
| openstackgerrit | Igor Degtiarov proposed openstack/ceilometer-specs: Transformers for events pipeline https://review.openstack.org/162167 | 16:50 |
| *** shardy has quit IRC | 16:52 | |
| gordc | pradk: you'll need to edit setup-tests.sh | 16:54 |
| gordc | i think in theory you can probably mismash the two setup-tests from gnocchiclient and aodhclient | 16:54 |
| *** KrishR1 has joined #openstack-telemetry | 16:58 | |
| pradk | gordc, hmm i see.. i guess we could make that a separate script in aodhclient and invoke it from setup script | 16:58 |
| *** KrishR has quit IRC | 16:58 | |
| pradk | gordc, i'll play around with it and see where i get | 16:58 |
| *** lsmola has quit IRC | 16:59 | |
| jd__ | sileht: sooooo I've applied a quick RLE compression to Carbonara files | 17:03 |
| jd__ | goes down from 252 Kb to 83 Kb | 17:03 |
| jd__ | 77% compression \o/ | 17:04 |
| *** bdossant_ has joined #openstack-telemetry | 17:04 | |
| jd__ | I'll polish that tomorrow | 17:05 |
| *** bdossant_ has quit IRC | 17:07 | |
| *** bdossant has quit IRC | 17:08 | |
| *** jaosorior_ has joined #openstack-telemetry | 17:08 | |
| *** jaosorior_ is now known as jaosorior | 17:09 | |
| *** _nadya_ has joined #openstack-telemetry | 17:09 | |
| *** _nadya_ has quit IRC | 17:10 | |
| pradk | gordc, we also need keystone running ? AuthorizationFailure: Authorization Failed: Unable to establish connection to http://localhost:5000/v2.0/tokens | 17:21 |
| jd__ | sileht: so I'm done from ~17 bytes per datapoint to ~6 bytes – now we have to achieve the ~3 bytes from InfluxDB :D should be fun | 17:23 |
| *** krotscheck1 has joined #openstack-telemetry | 17:24 | |
| krotscheck1 | jd__: The CORS patch I tried to land is currently failing because you downvoted a patch that it used to be dependent on. | 17:25 |
| jd__ | krotscheck1: on aodh? | 17:25 |
| krotscheck1 | Long story short: If i use the generic paste.ini file that comes with aodh, it automatically includes keystonemiddleware. | 17:25 |
| krotscheck1 | jd__: The tests assume keystone's not there. | 17:26 |
| krotscheck1 | So as soon as I add in the regular paste.ini file, they start to fail. | 17:26 |
| krotscheck1 | I created a gabbi-specific paste.ini file, which is the patch you downvoted. | 17:26 |
| jd__ | krotscheck1: just don't change gabbi and don't add those tests? | 17:27 |
| jd__ | we don't really care about testing a middleware I'd say | 17:27 |
| krotscheck1 | It's not the middleware that I'm concerned with, it's the proper integration of said middleware into the service. | 17:28 |
| krotscheck1 | These tests make sure the middleware is active. They don't test functionality. | 17:28 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: serialize AggregatedTimeSerie using RLE https://review.openstack.org/276365 | 17:28 |
| jd__ | which we don't really care about since api-paste.ini is a user supplied config file | 17:29 |
| jd__ | it can be deactivated | 17:30 |
| jd__ | gordc: ^ ? | 17:30 |
| gordc | reading | 17:30 |
| gordc | pradk: i'd need to check. it should be disabled in tests | 17:31 |
| gordc | jd__: you want to remove keystonemiddleware from default api-paste? | 17:31 |
| krotscheck1 | gordc: No. | 17:33 |
| pradk | gordc, this happens on the server side when validate_alarms is called http://paste.openstack.org/show/485999/ .. for aodh to authenticate to gnocchi | 17:34 |
| krotscheck1 | gordc: Patch set 5 on this patch: https://review.openstack.org/#/c/265330/5 | 17:34 |
| krotscheck1 | gordc: I created a gabbi-only paste file, just like in ceilometer, that excludes keystone. | 17:35 |
| krotscheck1 | gordc: But that was -2'd because it diverges, which seems odd to me because it brings aodh more in line with how things are done in ceilometer. | 17:36 |
| krotscheck1 | So, I'd like to know how you'd like me to proceed. Drop the feature altogether, do a gabbi-specific paste file, not test the middleware.... | 17:37 |
| krotscheck1 | I'm strongly against the latter. | 17:37 |
| gordc | krotscheck1: but the tests fail in that link... when you say middleware? which one are you talking about? CORS, keystonemiddleware? something else? | 17:38 |
| krotscheck1 | gordc: Patchset 5 passed. | 17:38 |
| krotscheck1 | gordc: Ok, so goal #1 is to provide gabbi with a testing surface that includes the entire paste chain. | 17:39 |
| krotscheck1 | If I do that, it breaks existing tests that were built without keystone auth tokens in mind. | 17:40 |
| krotscheck1 | I'm not even worried about the CORS middleware right now. | 17:41 |
| gordc | what pipeline is aodh running against now for functional tests? no middleware? | 17:41 |
| krotscheck1 | So there's two approaches that can be taken there. 1- create a different paste.ini file without keystonemiddleware in it. 2- figure out how to mock out the bits in keystonemiddleware in order to bypass the inevitable 401 | 17:41 |
| krotscheck1 | gordc: it runs against the raw application, no pipeline. | 17:43 |
| krotscheck1 | (app._app) | 17:43 |
| gordc | krotscheck1: ah i see. | 17:44 |
| gordc | krotscheck1: yeah, i'm inclined to say just create a new test paste. i'm not sure we want to change the default one to drop keystonemiddleware (i'd expect there might be residual side effects) | 17:45 |
| krotscheck1 | gordc: Ok, so I can revive patchset 5? | 17:45 |
| gordc | sure. i don't really have an alternative (i'd have to dig around to see how to mock out keystone) | 17:46 |
| openstackgerrit | Michael Krotscheck proposed openstack/aodh: gabbi's own paste.ini file https://review.openstack.org/265330 | 17:50 |
| *** _nadya_ has joined #openstack-telemetry | 17:50 | |
| krotscheck1 | gordc, jd__ ^^ Restored, updated. | 17:51 |
| gordc | pradk: sigh... i guess we have to enable that too? i don't know of any dummy parameters to make it ignore | 17:52 |
| gordc | will need to ask sileht about that. | 17:52 |
| gordc | biab. grabbing lunch | 17:53 |
| pradk | yea looks like it | 17:53 |
| nicodemus_ | jd__, gordc, I've pulled gnocchi master, redeployed and tested again with the ceilometer-agents | 17:58 |
| nicodemus_ | still, it appears that the unprocessed metrics are still building up in the pool | 17:59 |
| *** ddieterly has quit IRC | 17:59 | |
| *** yassine has joined #openstack-telemetry | 18:05 | |
| stevelle | ceilometer question: I have messages stuck in notifications.sample, where should I be looking | 18:06 |
| nicodemus_ | stevelle: is ceilometer-collector running? | 18:06 |
| stevelle | nicodemus_: yes, gnocchi as meter storage but nothing is happening | 18:07 |
| stevelle | lb says no calls to gnocchi | 18:07 |
| nicodemus_ | do you see consumers connected on the metering.sample queue? collector should be there consuming the messages | 18:08 |
| stevelle | correction: I see "GET /v1/capabilities/ HTTP/1.1" coming from collector on restart | 18:08 |
| nicodemus_ | on ceilometer.conf do you have a [dispatcher_gnocchi] section with the gnocchi URL? | 18:11 |
| stevelle | nicodemus_: yes, and there are two connections from collector to rabbit | 18:11 |
| nicodemus_ | stevelle: one time I had 50k messages stuck in the queue, eventually had to empty it by hand for processing to resume | 18:15 |
| nicodemus_ | ceilometer-collector.log should have some hint about what's happening | 18:15 |
| stevelle | it's the notifications queue where I see a buildup though | 18:16 |
| *** prashantD has joined #openstack-telemetry | 18:16 | |
| stevelle | and collector log is entirely non-helpful, though I was seeing this for a while yesterday http://paste.openstack.org/show/ZqkrNL6P2YyXCjK3ewFN/ | 18:17 |
| nicodemus_ | stevelle: oops my bad, I first read metering instead of notifications :/ | 18:21 |
| *** yassine has quit IRC | 18:25 | |
| *** tongli has joined #openstack-telemetry | 18:31 | |
| gordc | stevelle: notifications.sample comes from polling agents (not sure if any other service started publishing to that topic) | 18:40 |
| gordc | the notification agent should consume it (assuming it's coming from ceilometer) | 18:41 |
| gordc | nicodemus_: does it keep increasing at the same rate? | 18:41 |
| nicodemus_ | gordc: roughly yes. I'm trying increasing the num of workers | 18:42 |
| gordc | nicodemus_: same scenario? 2 metricd workers? you know roughly how many datapoints you publishing per second? | 18:42 |
| *** _nadya_ has quit IRC | 18:42 | |
| nicodemus_ | gordc: I'm listing all the objects in the ceph pool that start with the "measure_" prefix every 5 secs | 18:43 |
| *** krotscheck1 has quit IRC | 18:45 | |
| stevelle | per agent log: "INFO ceilometer.pipeline [-] detected decoupled pipeline config format" is this a problem? | 18:45 |
| nicodemus_ | gordc: there are about 300 new "measure_" objects per minute | 18:46 |
| gordc | stevelle: no. that's normal. i'd run rabbitmqctl list_consumers to see if anything is listening to notifications.sample | 18:47 |
| gordc | if not, that's a problem. (assuming this is liberty/mitaka) | 18:47 |
| gordc | nicodemus_: how many datapoints are you sending? 300 per minute doesn't seem bad (seems like adding an additional agent(s) might manage it. | 18:49 |
| stevelle | gordc: mitaka, and no listeners on that one | 18:49 |
| gordc | stevelle: odd. you might want to try to start another notification agent. | 18:50 |
| gordc | it should be handling that queue. | 18:50 |
| nicodemus_ | gordc: how can I measure the datapoints? | 18:50 |
| gordc | nicodemus_: errr. how frequent are your polling intervals, and how many instances do you have? | 18:51 |
| gordc | i guess one way would be to turn off ceilometer-collector and see how quickly metering.sample queue grows | 18:51 |
| nicodemus_ | gordc: polling interval is set at 60 seconds, and there are 36 instances on two hypervisors | 18:52 |
| nicodemus_ | that can be done, give me a minute and I'll have a number | 18:52 |
| nicodemus_ | gordc: there are 30 new messages on the metering.sample queue per minute | 19:02 |
| gordc | nicodemus_: when collector is not running? | 19:03 |
| nicodemus_ | correct | 19:03 |
| gordc | nicodemus_: you have a custom pipeline? | 19:04 |
| gordc | nicodemus_: that seems abnormally low since for every vm, there's multiple metrics captured for it (unless the vms are off?) | 19:05 |
| gordc | i'd expect a few hundred/min... doesn't really make sense 30 datapoints turns into 300 backlog | 19:06 |
| nicodemus_ | gordc: yes, we have a custom pipeline and only one of those VMs is off, all other VMs are up and running | 19:06 |
| nicodemus_ | gordc: http://paste.openstack.org/show/486005/ <--- this is the pipeline.yaml we're using | 19:07 |
| gordc | yeah, you should be getting way more than 30/min if polling 36instance every minute | 19:08 |
| *** efoley has quit IRC | 19:09 | |
| gordc | not sure why that's happening to be honest. | 19:09 |
| nicodemus_ | gordc: I'm checking the queue using rabbitmq management plugin, let me double-check if the num of messages in the queue is being reported correctly | 19:10 |
| *** david-lyle has quit IRC | 19:11 | |
| gordc | nicodemus_: kk. i ususally use 'sudo rabbitmqctl list_queues | grep -e \\.sample -e \\.info' | 19:16 |
| *** jaosorior has quit IRC | 19:19 | |
| nicodemus_ | gordc: is it possible that each message on the queue contains multiple metrics? I redid the test and yes, there are 30 messages/min | 19:23 |
| nicodemus_ | those 30 messages turn into between 250-300 "measure_" objects in the ceph gnocchi pool | 19:24 |
| gordc | on the notification.sample queue, the message may have multiple metrics... but at metering.sample it should be individual decomposed into individual metrics (unless you have accumulator transformer enabled, which it doesn't seem you do) | 19:24 |
| gordc | are you getting all the datapoints you'd expect... based on your pipeline, you should be getting everything http://docs.openstack.org/admin-guide-cloud/telemetry-measurements.html | 19:25 |
| *** _nadya_ has joined #openstack-telemetry | 19:27 | |
| nicodemus_ | yes | 19:28 |
| gordc | nicodemus_: i'm still very confused how you are getting a faster growing backlog than the number of datapoints you are actually writing.lol seems very odd | 19:32 |
| nicodemus_ | gordc: could you point me where to look? I assumed it was correct for each rabbitmq message to be stored as several "measure_" objects | 19:34 |
| gordc | nicodemus_: right. so when the collector receives a datapoint it will send it to gnocchi and it will store unprocessed data in measure_* obj | 19:42 |
| gordc | depending on whether you have batching enabled in collector (and how well the batching happens), it should be roughly a 1:1 value i think | 19:43 |
| gordc | 1 metering.sample message = 1 measure_* entry in ceph | 19:43 |
| gordc | assuming ceilometer is the only thing writing to gnocchi | 19:44 |
| gordc | you can try turning off collector and maybe using rabbitmq webui to see what the 30 messages in metering.sapmle queue are. | 19:44 |
| nicodemus_ | gordc: got it. I'll dig a little deeper, to see what's happening | 19:47 |
| *** mragupat has joined #openstack-telemetry | 19:48 | |
| nicodemus_ | gordc: btw, this is one message being submitted in the queue: http://paste.openstack.org/show/486008/ | 19:49 |
| gordc | metering.sample queue? | 19:50 |
| gordc | or notifications.sample queue? | 19:51 |
| nicodemus_ | metering.sample queue | 19:53 |
| nicodemus_ | apparently, a single message on that queue results in 21 "measure_" objects in the ceph pool | 19:53 |
| gordc | that's weird. that's strange. i'd imagine notification agent would break that up. | 19:55 |
| gordc | i'll need to dig into that | 19:55 |
| nicodemus_ | One sec... I was just using compute-agent, agent-central and collector. I'm not running ceilometer-agent-notification... should I? | 19:57 |
| gordc | nicodemus_: liberty or newer? | 19:57 |
| nicodemus_ | liberty | 19:58 |
| gordc | yeah, you need them... i'm not sure what is writing to metering.sample to be honest... only the notification agent should be writing to that queu | 19:58 |
| gordc | yeah, you need them... i'm not sure what is writing to metering.sample to be honest... only the notification agent should be writing to that queue | 19:59 |
| *** jwcroppe has quit IRC | 20:01 | |
| *** jwcroppe has joined #openstack-telemetry | 20:02 | |
| *** jwcroppe has quit IRC | 20:07 | |
| *** KrishR1 has quit IRC | 20:08 | |
| *** KrishR has joined #openstack-telemetry | 20:10 | |
| *** safchain has quit IRC | 20:17 | |
| *** safchain has joined #openstack-telemetry | 20:31 | |
| stevelle | so I have 4 channels, only 3 connections, and 2 consumers coming from ceilometer. Agent is not consuming notifications.samples. I'm stumped | 20:41 |
| *** ityaptin_ has joined #openstack-telemetry | 20:41 | |
| *** wolsen_ has joined #openstack-telemetry | 20:41 | |
| *** Taytay has joined #openstack-telemetry | 20:45 | |
| *** ityaptin has quit IRC | 20:47 | |
| *** nicodemus_ has quit IRC | 20:47 | |
| *** titilambert has quit IRC | 20:47 | |
| *** wolsen has quit IRC | 20:47 | |
| *** nicodemus_ has joined #openstack-telemetry | 20:49 | |
| *** nicodemus_ has quit IRC | 20:50 | |
| *** gordc has quit IRC | 20:52 | |
| *** safchain has quit IRC | 20:56 | |
| *** nicodemus_ has joined #openstack-telemetry | 21:03 | |
| openstackgerrit | Pradeep Kilambi proposed openstack/python-aodhclient: WIP: Add gnocchi alarm rules https://review.openstack.org/266073 | 21:04 |
| *** nicodemus_ has quit IRC | 21:13 | |
| *** _nadya_ has quit IRC | 21:16 | |
| *** guillaume__ has quit IRC | 21:27 | |
| *** farid_ has quit IRC | 21:27 | |
| pradk | sileht, you around? | 21:27 |
| *** guillaume__ has joined #openstack-telemetry | 21:42 | |
| *** pradk has quit IRC | 21:46 | |
| *** pradk has joined #openstack-telemetry | 21:48 | |
| *** gordc has joined #openstack-telemetry | 21:49 | |
| *** julim has quit IRC | 21:55 | |
| *** julim has joined #openstack-telemetry | 21:55 | |
| openstackgerrit | Merged openstack/ceilometer: set higher batching requirement https://review.openstack.org/276330 | 22:00 |
| *** julim has quit IRC | 22:02 | |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: serialize AggregatedTimeSerie using RLE https://review.openstack.org/276365 | 22:02 |
| *** thorst has quit IRC | 22:04 | |
| *** davidlenwell has quit IRC | 22:22 | |
| *** davidlenwell has joined #openstack-telemetry | 22:23 | |
| *** thorst has joined #openstack-telemetry | 22:29 | |
| *** thorst has quit IRC | 22:34 | |
| *** leitan has quit IRC | 22:34 | |
| openstackgerrit | gordon chung proposed openstack/gnocchi: add randomness/chaos to metricd - POC https://review.openstack.org/276485 | 22:43 |
| *** david-lyle has joined #openstack-telemetry | 22:48 | |
| pradk | sileht, gordc, have you seen this erro before http://paste.openstack.org/show/486025/ | 22:48 |
| pradk | seems related to recent change https://github.com/openstack/gnocchi/commit/3acea2d12a1d479f547630f1c7fc60ab4a28433f | 22:49 |
| gordc | pradk: what does your paste look like? | 22:49 |
| pradk | gordc, https://ci.centos.org/artifacts/rdo/weirdo-promote-mitaka-puppet-openstack-scenario001/49/puppet-openstack/logs/etc/gnocchi/api-paste.ini.txt.gz | 22:50 |
| pradk | pretty much whats in master | 22:50 |
| gordc | pradk: ... your pipeline has gnocchi in it. | 22:51 |
| gordc | it should be either gnocchi+auth or gnocchi+noauth? | 22:51 |
| pradk | yea so looks like we need to update puppet to use gnocchi=noauth? | 22:52 |
| gordc | or gnocchiv1 | 22:52 |
| pradk | cool | 22:52 |
| gordc | either or i guess | 22:52 |
| *** david-lyle has quit IRC | 22:52 | |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: serialize AggregatedTimeSerie using RLE https://review.openstack.org/276365 | 22:57 |
| *** ska has joined #openstack-telemetry | 23:02 | |
| ska | I see unique UUID's for "binary": "neutron-openvswitch-agent", entries in Ceilometer heartbeats. Are those related to host or the service that heartbeat is monitoring? | 23:04 |
| ska | One example was for "agent_type": "Open vSwitch agent" | 23:05 |
| *** guillaume__ has quit IRC | 23:06 | |
| jd__ | gordc: https://en.wikipedia.org/wiki/Run-length_encoding | 23:11 |
| gordc | jd__: wow. we're actually computer scientist now. | 23:11 |
| jd__ | YEAH | 23:12 |
| jd__ | maybe I'll get a degree | 23:12 |
| gordc | maybe. i'll keep using my hammer. | 23:12 |
| gordc | ska: what are ceilometer heartbeats? | 23:13 |
| jd__ | gordc: any interesting stuff wrt nico? | 23:13 |
| gordc | i'm not sure. mentioned it was no difference. | 23:14 |
| gordc | but he also didn't have a notification agent running so i'm not exactly sure how he was even sending data to gnocchi | 23:14 |
| gordc | unless theres a way to go polling->collector->gnocchi? | 23:15 |
| *** mgarza has quit IRC | 23:27 | |
| gordc | signing off | 23:27 |
| *** gordc has quit IRC | 23:28 | |
| *** pradk has quit IRC | 23:28 | |
| openstackgerrit | Merged openstack/ceilometer: do not configure worker specific items in init https://review.openstack.org/273792 | 23:28 |
| *** chlong has quit IRC | 23:30 | |
| *** KrishR has quit IRC | 23:39 | |
| *** mragupat_ has joined #openstack-telemetry | 23:45 | |
| *** mragupat has quit IRC | 23:49 | |
| *** mragupat_ has quit IRC | 23:49 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!