| *** diogogmt has joined #openstack-telemetry | 00:00 | |
| *** thorst has joined #openstack-telemetry | 00:17 | |
| *** fguillot is now known as fguillot_afk | 00:20 | |
| *** fguillot_afk has quit IRC | 00:20 | |
| *** thorst has quit IRC | 00:24 | |
| *** vishwanathj has quit IRC | 00:43 | |
| *** fguillot has joined #openstack-telemetry | 00:46 | |
| *** thorst has joined #openstack-telemetry | 01:21 | |
| *** thorst has quit IRC | 01:29 | |
| *** thorst has joined #openstack-telemetry | 01:35 | |
| *** thorst has quit IRC | 01:45 | |
| openstackgerrit | Ha Van Tu proposed openstack/ceilometer: Refactor Ceilometer event API https://review.openstack.org/370510 | 01:50 |
|---|---|---|
| *** diogogmt has quit IRC | 02:09 | |
| *** zhangguoqing has joined #openstack-telemetry | 02:14 | |
| *** fguillot is now known as fguillot_afk | 02:19 | |
| *** fguillot_afk has quit IRC | 02:19 | |
| *** fguillot has joined #openstack-telemetry | 02:23 | |
| *** thorst has joined #openstack-telemetry | 02:47 | |
| *** thorst has quit IRC | 02:56 | |
| *** fguillot has quit IRC | 03:33 | |
| *** thorst has joined #openstack-telemetry | 03:44 | |
| *** thorst has quit IRC | 03:46 | |
| openstackgerrit | zhangguoqing proposed openstack/gnocchi: Add volume_type attribute to volume https://review.openstack.org/365885 | 03:50 |
| zhangguoqing | For gnocchi, I always find that there are some resources really be deleted, but those of resource and metric items remain in the database tables. why? | 04:04 |
| *** flwang1 has quit IRC | 04:26 | |
| *** thorst has joined #openstack-telemetry | 04:29 | |
| openstackgerrit | Shengping Zhang proposed openstack/gnocchi: This patch is used to delete batch of resources https://review.openstack.org/351043 | 04:36 |
| openstackgerrit | Nguyen Phuong An proposed openstack/ceilometer: Prevent use filter(lambda ...) to meet python2,3 https://review.openstack.org/370536 | 04:37 |
| *** thorst has quit IRC | 04:39 | |
| *** adriant has quit IRC | 04:40 | |
| *** thorst has joined #openstack-telemetry | 04:45 | |
| *** thorst has quit IRC | 04:58 | |
| *** thorst has joined #openstack-telemetry | 04:58 | |
| *** donghao has quit IRC | 05:01 | |
| *** thorst has quit IRC | 05:07 | |
| *** thorst has joined #openstack-telemetry | 05:09 | |
| *** thorst has quit IRC | 05:13 | |
| *** thorst has joined #openstack-telemetry | 05:15 | |
| *** rcernin has joined #openstack-telemetry | 05:17 | |
| *** thorst has quit IRC | 05:19 | |
| *** adriant has joined #openstack-telemetry | 05:29 | |
| *** yprokule has joined #openstack-telemetry | 05:33 | |
| *** thorst has joined #openstack-telemetry | 05:34 | |
| *** thorst has quit IRC | 05:42 | |
| *** thorst has joined #openstack-telemetry | 05:44 | |
| *** vishwanathj has joined #openstack-telemetry | 05:47 | |
| *** donghao has joined #openstack-telemetry | 05:48 | |
| *** thorst has quit IRC | 05:49 | |
| *** thorst has joined #openstack-telemetry | 05:50 | |
| *** thorst has quit IRC | 05:55 | |
| *** thorst has joined #openstack-telemetry | 05:56 | |
| openstackgerrit | zhangguoqing proposed openstack/gnocchi: Add volume_type attribute to volume https://review.openstack.org/365885 | 05:59 |
| *** thorst has quit IRC | 06:01 | |
| *** thorst has joined #openstack-telemetry | 06:02 | |
| *** pcaruana has joined #openstack-telemetry | 06:04 | |
| *** thorst has quit IRC | 06:07 | |
| *** thorst has joined #openstack-telemetry | 06:08 | |
| *** pcaruana is now known as pcaruana|afk| | 06:13 | |
| *** thorst has quit IRC | 06:17 | |
| *** thorst has joined #openstack-telemetry | 06:19 | |
| *** thorst has quit IRC | 06:24 | |
| *** thorst has joined #openstack-telemetry | 06:25 | |
| *** khushbu_ has joined #openstack-telemetry | 06:29 | |
| *** thorst has quit IRC | 06:30 | |
| *** adriant has quit IRC | 06:32 | |
| *** thorst has joined #openstack-telemetry | 06:32 | |
| *** thorst has quit IRC | 06:36 | |
| *** pcaruana|afk| is now known as pcaruana | 06:40 | |
| *** thorst has joined #openstack-telemetry | 06:44 | |
| *** rwsu has joined #openstack-telemetry | 06:44 | |
| *** donghao has quit IRC | 06:47 | |
| *** khushbu_ has quit IRC | 06:51 | |
| *** thorst has quit IRC | 06:52 | |
| *** khushbu has joined #openstack-telemetry | 06:53 | |
| *** thorst has joined #openstack-telemetry | 06:54 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: tox: remove useless keyword https://review.openstack.org/368411 | 06:54 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: cleanup metricd shutdown process https://review.openstack.org/368410 | 06:54 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Add simple upgrade tests https://review.openstack.org/368412 | 06:54 |
| *** thorst has quit IRC | 06:59 | |
| *** thorst has joined #openstack-telemetry | 07:00 | |
| *** thorst has quit IRC | 07:05 | |
| *** larainema has quit IRC | 07:08 | |
| *** larainema has joined #openstack-telemetry | 07:10 | |
| *** thorst has joined #openstack-telemetry | 07:11 | |
| *** thorst has quit IRC | 07:17 | |
| *** thorst has joined #openstack-telemetry | 07:18 | |
| *** shardy has joined #openstack-telemetry | 07:20 | |
| *** thorst_ has joined #openstack-telemetry | 07:21 | |
| *** zhurong has joined #openstack-telemetry | 07:22 | |
| *** thorst has quit IRC | 07:23 | |
| *** thorst_ has quit IRC | 07:25 | |
| *** yprokule_ has joined #openstack-telemetry | 07:28 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/aodh: Imported Translations from Zanata https://review.openstack.org/367140 | 07:28 |
| *** yprokule has quit IRC | 07:29 | |
| *** yprokule_ is now known as yprokule | 07:29 | |
| *** amoralej|off is now known as amoralej | 07:34 | |
| *** tobasco has quit IRC | 07:40 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/aodh: sync tooz requirement https://review.openstack.org/370416 | 07:40 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/aodh: Cap happybase to 1.0 https://review.openstack.org/370582 | 07:40 |
| *** yassine has joined #openstack-telemetry | 08:15 | |
| *** sheeprine has quit IRC | 08:15 | |
| *** yassine is now known as Guest67165 | 08:16 | |
| *** sheeprine has joined #openstack-telemetry | 08:18 | |
| openstackgerrit | Merged openstack/gnocchi: cleanup metricd shutdown process https://review.openstack.org/368410 | 08:26 |
| *** khushbu has quit IRC | 08:32 | |
| *** tobasco has joined #openstack-telemetry | 08:38 | |
| *** khushbu has joined #openstack-telemetry | 08:38 | |
| openstackgerrit | Merged openstack/gnocchi: tox: remove useless keyword https://review.openstack.org/368411 | 08:41 |
| *** yprokule_ has joined #openstack-telemetry | 08:43 | |
| *** yprokule has quit IRC | 08:46 | |
| *** yprokule_ is now known as yprokule | 08:46 | |
| *** efoley has joined #openstack-telemetry | 09:02 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Allow to pass search filter in the query string https://review.openstack.org/370629 | 09:13 |
| *** zhurong has quit IRC | 09:24 | |
| sileht | Jenkins reaction when sileht try to fix ceilometer liberty branch for next release: https://giphy.com/gifs/laughing-scrubs-gif-3i7zenReaUuI0 | 09:28 |
| *** thorst has joined #openstack-telemetry | 09:38 | |
| *** khushbu has quit IRC | 09:44 | |
| *** khushbu has joined #openstack-telemetry | 09:45 | |
| *** efoley has quit IRC | 10:07 | |
| *** efoley has joined #openstack-telemetry | 10:09 | |
| *** khushbu has quit IRC | 10:49 | |
| *** shardy has quit IRC | 10:57 | |
| *** donghao has joined #openstack-telemetry | 11:05 | |
| *** dave-mccowan has joined #openstack-telemetry | 11:06 | |
| *** efoley_ has joined #openstack-telemetry | 11:10 | |
| *** khushbu has joined #openstack-telemetry | 11:12 | |
| *** efoley has quit IRC | 11:13 | |
| *** thorst has quit IRC | 11:27 | |
| *** khushbu has quit IRC | 11:35 | |
| *** gordc has joined #openstack-telemetry | 11:59 | |
| *** jkraj has joined #openstack-telemetry | 12:04 | |
| *** amoralej is now known as amoralej|lunch | 12:07 | |
| *** pradk has joined #openstack-telemetry | 12:13 | |
| gordc | jd__: in your testing, if you use the grouped_series patch, you might want to try enabling threading. it seems to work for me. or at least it's not worse | 12:20 |
| gordc | using ceph. haven't tried file yet. | 12:20 |
| *** Matias has joined #openstack-telemetry | 12:29 | |
| Matias | morning | 12:30 |
| jd__ | gordc: cool | 12:32 |
| sileht | jd, last ceph version have my python binding for the async API, I will release a new cradox with that stuffs and start to update the ceph driver soon | 12:41 |
| gordc | for v3? | 12:42 |
| sileht | I think we release v3 before | 12:42 |
| sileht | I haven't started the work ^^ | 12:42 |
| gordc | kk. we keep finding and improving things in gnocchi. i like it. | 12:42 |
| *** ljxiash has joined #openstack-telemetry | 12:44 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Allow to retry to delete a resource https://review.openstack.org/370783 | 12:47 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Allow to retry to delete a resource https://review.openstack.org/370783 | 12:49 |
| *** shardy has joined #openstack-telemetry | 12:50 | |
| jd__ | cool sileht | 12:50 |
| jd__ | that'll be for 3.1 ^^ | 12:50 |
| jd__ | we need to wrap up 3.0 now | 12:50 |
| *** ljxiash has quit IRC | 12:53 | |
| *** fguillot has joined #openstack-telemetry | 12:57 | |
| openstackgerrit | Merged openstack/gnocchi: share groupings across aggregates https://review.openstack.org/368226 | 13:01 |
| openstackgerrit | Merged openstack/gnocchi: cleanup carbonara - drop update method https://review.openstack.org/368922 | 13:10 |
| *** cdent has joined #openstack-telemetry | 13:13 | |
| *** fguillot has quit IRC | 13:13 | |
| *** amoralej|lunch is now known as amoralej | 13:15 | |
| jd__ | sileht: is the upgrade between 2.2 and master working? | 13:18 |
| sileht | jd__, it looks like I have testing it with commit merged yesterday and today | 13:19 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: This patch is used to delete batch of resources https://review.openstack.org/351043 | 13:20 |
| sileht | jd__, checkout/rebase my change and try it ? I'm kidding I known your run macos and that doesn't work :p | 13:20 |
| jd__ | sileht: I'm surprised that the upgrade code from gordc works | 13:21 |
| jd__ | but who knows lol | 13:21 |
| sileht | jd__, I think it mock the API and always return the same values | 13:21 |
| jd__ | haha | 13:21 |
| jd__ | [BUG] Gnocchi always returns 42 | 13:21 |
| sileht | jd__, I have seen to some 'whitespace programming language' hidden in the python code | 13:22 |
| *** shardy has quit IRC | 13:22 | |
| sileht | and I don't known what it does | 13:22 |
| gordc | if it works, yes, i did it... if not, it was a group effort. we all reviewed it. | 13:23 |
| sileht | lo | 13:25 |
| sileht | l | 13:25 |
| sileht | jd__, gordc I'm reading the async API, and we have a interesting ceph detail, ceph all to return from async operation completion when object have migrated to ceph memory (vs ondisk), so next query will read the good data even their are not yet commit to the disk | 13:32 |
| sileht | jd__, gordc is that safe for us ? | 13:32 |
| gordc | sileht: is this with the new version of Ceph (with bluestore)? | 13:33 |
| sileht | jd__, gordc in memory means 3 differents server if you configure the pool to have 3 replicats | 13:33 |
| sileht | gordc, that works with older too | 13:33 |
| sileht | gordc, but it's recently exposed in the python binding | 13:34 |
| gordc | sileht: i see | 13:34 |
| sileht | I backport that to cradox like have did before | 13:34 |
| sileht | I think for Gnocchi we don't care about having think commit to disk, we just want to be sure the backend will not loose it | 13:35 |
| gordc | 3 different memory locations is pretty durable no? | 13:35 |
| sileht | yeah I think that too | 13:35 |
| gordc | i believe a lot of other databases write to memory first (or are all memory) | 13:36 |
| gordc | i'm ok with just guaranteeing it's in memory. | 13:36 |
| sileht | gordc, its memory + journal to be more precise | 13:36 |
| gordc | oh. that | 13:36 |
| gordc | that's even better | 13:36 |
| gordc | so it saves us roundtrip time ceph takes to replicates to each OSD | 13:37 |
| sileht | gordc, it still replicate the data on each OSD, but only on they memory | 13:38 |
| sileht | currently we wait on disk | 13:38 |
| sileht | that huge waste of times :p | 13:38 |
| gordc | ah right. that's what i meant | 13:38 |
| gordc | so what is normal workflow now? it writes to joural->memory+replicate->disk+replicate? | 13:39 |
| sileht | gordc, I think the exact workflow is complex and depends on what operation you ask for | 13:40 |
| gordc | i see. nothing is simple. got it.lol | 13:41 |
| gordc | i'm ok with guarantee in memory though. i agree with you | 13:41 |
| *** vint_bra1 has joined #openstack-telemetry | 13:43 | |
| *** dstepanenko_ has quit IRC | 13:45 | |
| *** vint_bra has quit IRC | 13:46 | |
| *** dstepanenko_ has joined #openstack-telemetry | 13:46 | |
| jd__ | more perf coming I understannd, kewl | 13:52 |
| jd__ | gordc: https://review.openstack.org/#/c/370849 :) | 13:52 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: Add a S3 based storage driver https://review.openstack.org/354131 | 14:08 |
| *** rbak has joined #openstack-telemetry | 14:08 | |
| gordc | jd__: will look. | 14:11 |
| gordc | did you test serialization perf vs msgpack? | 14:12 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: ceph: fix setup extra https://review.openstack.org/369016 | 14:12 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: ceph: Fix metricd start https://review.openstack.org/369010 | 14:12 |
| jd__ | sileht: I rebased ^^^^ because they were based on abandonned patches | 14:12 |
| gordc | just for reference. right now writing the curent boundtimeserie takes about 0.002s | 14:12 |
| jd__ | gordc: no I did not compare | 14:13 |
| jd__ | I can | 14:13 |
| gordc | kk, i imagine this would help if you have large back_window? | 14:14 |
| jd__ | old serialization: 12 MB/s old unserialization: 30 MB/s new serialization: 12 MB/s new unserialization: 28 MB/s | 14:15 |
| jd__ | so it's almost the same speed, but the size is smaller | 14:15 |
| jd__ | and yeah if you have a lot of points it'll be way smaller so faster to write | 14:16 |
| jd__ | so it should improve perf a bit | 14:16 |
| *** fguillot has joined #openstack-telemetry | 14:17 | |
| gordc | cool cool. | 14:17 |
| gordc | sounds like a good optimisation. | 14:17 |
| openstackgerrit | Merged openstack/ceilometer: Use deps extra for optional requirements https://review.openstack.org/369896 | 14:18 |
| *** paw has quit IRC | 14:18 | |
| *** yassine has joined #openstack-telemetry | 14:28 | |
| *** yassine is now known as Guest31676 | 14:28 | |
| *** Guest67165 has quit IRC | 14:29 | |
| *** cdent has quit IRC | 14:37 | |
| *** pradk has quit IRC | 14:42 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: WIP: ceph: move out of xattr completly https://review.openstack.org/370896 | 14:42 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: POC: ceph: make measures processing call async https://review.openstack.org/370897 | 14:42 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: ceph: read all measures async https://review.openstack.org/370898 | 14:42 |
| *** pradk has joined #openstack-telemetry | 14:45 | |
| *** donghao has quit IRC | 14:51 | |
| *** paw has joined #openstack-telemetry | 14:51 | |
| *** diogogmt has joined #openstack-telemetry | 15:03 | |
| gordc | weird. pecan won't let me wrap this in threadpoolexecutor https://github.com/openstack/gnocchi/blob/master/gnocchi/rest/__init__.py#L1285-L1287 | 15:14 |
| jd__ | sileht: https://review.openstack.org/#/c/354757 still fails and now we lost a bunch of jobs | 15:19 |
| *** pcaruana has quit IRC | 15:19 | |
| jd__ | I am getting tired of the infra :( | 15:19 |
| jd__ | gordc: how that? | 15:19 |
| gordc | jd__: made a bug https://bugs.launchpad.net/gnocchi/+bug/1624004 | 15:20 |
| openstack | Launchpad bug 1624004 in Gnocchi "use threadpoolexecutor to handle batch" [Undecided,New] | 15:20 |
| *** diogogmt has quit IRC | 15:20 | |
| gordc | don't have time currently to look into how to do it properly (or if it's just me being stupid) | 15:20 |
| gordc | maybe the threading is expected to be handled by apache or uwsgi? *shrugs* | 15:21 |
| *** mugsie has left #openstack-telemetry | 15:25 | |
| openstackgerrit | Merged openstack/gnocchi: ceph: fix setup extra https://review.openstack.org/369016 | 15:31 |
| openstackgerrit | Merged openstack/gnocchi: ceph: Fix metricd start https://review.openstack.org/369010 | 15:31 |
| openstackgerrit | Merged openstack/gnocchi: This patch is used to delete batch of resources https://review.openstack.org/351043 | 15:31 |
| jd__ | gordc: pecan.request is thread local | 15:32 |
| jd__ | gordc: so as soon as you start a new thread it's empty | 15:33 |
| jd__ | gordc: I think doing a storage = pecan.request.storage might be enough before lambda/executor.map | 15:33 |
| jd__ | just a wild guess | 15:33 |
| gordc | jd__: hmm.. let me give that i try. i still have code i think. | 15:33 |
| *** Tamayo has joined #openstack-telemetry | 15:36 | |
| gordc | jd__: that worked. didn't really improve performance though. maybe 10% at quick glance. | 15:37 |
| jd__ | not that bad | 15:37 |
| gordc | ~17s vs ~20s for 50resourcex20metrics batch | 15:39 |
| gordc | i'll keep testing in a bit. | 15:39 |
| *** cdent has joined #openstack-telemetry | 15:46 | |
| openstackgerrit | gordon chung proposed openstack/gnocchi: lower processor job polling https://review.openstack.org/370960 | 15:47 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: ceph: read all measures async https://review.openstack.org/370898 | 15:50 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: POC: ceph: make measures processing call async https://review.openstack.org/370897 | 15:50 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: WIP: ceph: move out of xattr completly https://review.openstack.org/370896 | 15:50 |
| *** jkraj has quit IRC | 15:58 | |
| *** thumpba has joined #openstack-telemetry | 16:00 | |
| *** cdent has quit IRC | 16:00 | |
| *** rcernin has quit IRC | 16:26 | |
| *** cdent has joined #openstack-telemetry | 16:35 | |
| *** fguillot has quit IRC | 16:41 | |
| *** yprokule has quit IRC | 16:44 | |
| *** thumpba has quit IRC | 16:48 | |
| *** thumpba has joined #openstack-telemetry | 16:49 | |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: implement compressed format for BoundTimeSerie https://review.openstack.org/370849 | 16:53 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: retrieve unaggregated timeserie only once for upgrade https://review.openstack.org/371001 | 16:53 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: doc: remove the rolling upgrade documentation https://review.openstack.org/371002 | 16:53 |
| *** thumpba has quit IRC | 16:54 | |
| *** cdent has quit IRC | 16:55 | |
| jd__ | sileht: I thought pifpaf was ok now for https://review.openstack.org/#/c/368412/15/run-upgrade-tests.sh ? | 17:11 |
| sileht | jd__, gordc, http://paste.openstack.org/show/577676/ | 17:11 |
| jd__ | sileht: lol. | 17:11 |
| sileht | jd__, gordc perhaps my code is broken I have written it and then bench | 17:11 |
| *** khushbu_ has joined #openstack-telemetry | 17:11 | |
| sileht | no tests of course | 17:11 |
| sileht | I will write tests tomorrow | 17:11 |
| jd__ | sileht: well if it works… it's just that I hope there is no race condition | 17:11 |
| jd__ | sileht: would be cool to have that for 3.0 :] | 17:12 |
| *** efoley_ has quit IRC | 17:13 | |
| sileht | jd__, I have pushed the wrong branch | 17:13 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Add simple upgrade tests https://review.openstack.org/368412 | 17:13 |
| jd__ | ah | 17:13 |
| jd__ | cool | 17:14 |
| jd__ | I think we should leverage Gabbi then to run the tests | 17:14 |
| sileht | jd__, creating ton of measures with gabbi is not easy | 17:15 |
| sileht | jd__, we should generate the json with script and then use gabbi I think | 17:16 |
| jd__ | yeah | 17:16 |
| *** thumpba has joined #openstack-telemetry | 17:17 | |
| sileht | see you tomorrow guys | 17:17 |
| *** fguillot has joined #openstack-telemetry | 17:26 | |
| *** amoralej is now known as amoralej|off | 17:43 | |
| *** Tamayo has quit IRC | 17:49 | |
| *** khushbu_ has quit IRC | 18:09 | |
| *** khushbu_ has joined #openstack-telemetry | 18:10 | |
| gordc | sileht: lol i'm sure you hid a 'break/continue' somewhere in there to skip code :P | 18:16 |
| *** pcaruana has joined #openstack-telemetry | 18:22 | |
| Matias | is there any way to query the sum of the means of a certain metric over all resources matching a certain query? | 18:39 |
| openstackgerrit | gordon chung proposed openstack/gnocchi: minimise job queuing https://review.openstack.org/371052 | 18:40 |
| Matias | my issue is: in my ceilometer pipeline, I'm polling at each 15 seconds | 18:40 |
| Matias | I'm using the medium archiving policy for gnocchi, which stores 1 sample per 60 seconds for the recent history | 18:40 |
| Matias | therefore, each sample stored by gnocchi corresponds to 4 samples gathered by the polling agents | 18:40 |
| Matias | when I call /v1/aggregation/resource/instance_network_interface/metric/network.outgoing.bytes.rate?aggregation=sum&..., I get the sum over all resources matching the query, but also over all 4 samples collected by the pollster in a 60 second interval | 18:41 |
| gordc | Matias: i assume you want something like this: https://bugs.launchpad.net/gnocchi/+bug/1619412 | 18:42 |
| openstack | Launchpad bug 1619412 in Gnocchi "Cannot group by without specifying resource type" [Undecided,Triaged] | 18:42 |
| *** fguillot has quit IRC | 18:44 | |
| Matias | gordc: not really, suppose I have 2 nova instances, each one generating its own network.outgoing.bytes.rate metrics | 18:44 |
| Matias | for the first one, the pollster collects [1,1,1,1, 2,2,2,2, 1,1,1,1] | 18:44 |
| Matias | for the second one, the pollster collects [2,2,2,2, 3,3,3,3, 1,1,1,1] | 18:45 |
| Matias | now, my pollster has a sampling rate of 1 sample per 15 seconds | 18:45 |
| Matias | whereas gnocchi stores 1 point per 60 seconds | 18:45 |
| Matias | gnocchi will store, as the mean aggreation, [1, 2, 1] for the first one, and [2, 3, 1] for the second one | 18:46 |
| Matias | now suppose I want a sum aggregation across all instances, to get the total network rate over my entire cloud shown in grafana | 18:46 |
| Matias | I'd pass aggregation=sum to /v1/aggregation/resource | 18:47 |
| gordc | you would use reaggregation parameter | 18:48 |
| Matias | but it'd return [12, 20, 8] | 18:48 |
| gordc | you want [3, 5, 2] right? | 18:48 |
| Matias | yes, so &aggregation=mean&reaggregation=sum ? | 18:48 |
| gordc | correct | 18:49 |
| Matias | gordc: thanks :D | 18:49 |
| gordc | np | 18:49 |
| Matias | now just need to implement these in the grafana plugin and send a PR to sileht :) | 18:49 |
| gordc | i think there's a queue for sileht. :) | 18:50 |
| *** fguillot has joined #openstack-telemetry | 18:51 | |
| *** fguillot has quit IRC | 18:51 | |
| *** khushbu_ has quit IRC | 18:51 | |
| *** fguillot has joined #openstack-telemetry | 18:52 | |
| sileht | Matias, yeah, this is welcomed | 18:53 |
| Matias | sileht: something I wanted to ask you: I patched my grafana to add needed_overlap=0 to the aggregation API call, but I don't know if it makes sense in the general cause, or if it should be configurable | 18:54 |
| sileht | Matias, configurable is better, but I think hardcoding 0 make sense for now | 18:54 |
| Matias | sileht: because when graphing server groups or project resources, in general you want to graph the total regardless if new instances are spawn or destroyed | 18:54 |
| sileht | Matias, we already do that in many app | 18:55 |
| sileht | Matias, and even we make it configure the default will be 0 too | 18:55 |
| Matias | sileht: nice, I'll try to get some PRs pushed to you until tomorrow | 18:55 |
| Matias | sileht: thanks :) | 18:56 |
| *** vint_bra1 has quit IRC | 19:02 | |
| *** vint_bra has joined #openstack-telemetry | 19:02 | |
| *** r-mibu has quit IRC | 19:16 | |
| *** r-mibu has joined #openstack-telemetry | 19:16 | |
| *** ametts has joined #openstack-telemetry | 19:26 | |
| *** pcaruana has quit IRC | 19:32 | |
| *** davidlenwell has quit IRC | 19:34 | |
| *** AJaeger has joined #openstack-telemetry | 19:37 | |
| AJaeger | aodh cores, could you import translations in time for RC1, please? https://review.openstack.org/367140 | 19:37 |
| *** donghao has joined #openstack-telemetry | 19:38 | |
| *** donghao has quit IRC | 19:39 | |
| *** david-lyle has quit IRC | 19:40 | |
| *** david-lyle has joined #openstack-telemetry | 19:40 | |
| gordc | AJaeger: done. | 19:43 |
| *** davidlenwell has joined #openstack-telemetry | 19:47 | |
| *** rcernin has joined #openstack-telemetry | 19:53 | |
| *** ametts has quit IRC | 20:01 | |
| jd__ | gordc: sileht: we're gonna need to do a RC1 on Monday for Ceilometer/Panko/Aodh so we need to wrap up at this stage, if you can swipe the patches once for all | 20:11 |
| jd__ | gordc: sileht: and we need to go ahead for Gnocchi 3.0 end of next week I'd say so let's hurry | 20:11 |
| jd__ | NO PRESSURE :P | 20:11 |
| jd__ | now I'm gonna spend my evening review patches | 20:11 |
| *** ametts has joined #openstack-telemetry | 20:13 | |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: doc: remove the rolling upgrade documentation https://review.openstack.org/371002 | 20:14 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: implement compressed format for BoundTimeSerie https://review.openstack.org/370849 | 20:14 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: retrieve unaggregated timeserie only once for upgrade https://review.openstack.org/371001 | 20:14 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: releasenotes: fix typo in storage-engine-v3 https://review.openstack.org/371091 | 20:15 |
| openstackgerrit | Merged openstack/aodh: Imported Translations from Zanata https://review.openstack.org/367140 | 20:18 |
| therve | jd__, Did you find the issue with sqlalchemy and panko? | 20:19 |
| jd__ | therve: I did not have time to dig into it | 20:20 |
| jd__ | I wrote a patch in Ceilometer to start having Panko enabled in the cross project test job we have | 20:20 |
| gordc | jd__: i will spend my evenings sleeping. | 20:20 |
| jd__ | but Panko is kinda low prio for me at the moment :( | 20:20 |
| jd__ | gordc: can you check https://review.openstack.org/#/c/335289/ it's the last patch for Aodh rc1 I think | 20:20 |
| therve | OK. That sounds kinda critical for the release though :) | 20:20 |
| jd__ | therve: agreed, though if nobody fixes it at some point, I don't know what to say lol | 20:21 |
| therve | Well, you know, you are never better served than by yourself :p | 20:23 |
| jd__ | therve: does that mean you're gonna fix it? ;-) | 20:24 |
| therve | You know you can guilt me into doing it | 20:25 |
| gordc | jd__: there's a requirement issue with that patch i feel. i can wait for flwang to see if he fixes it or just fix it ourselves | 20:26 |
| jd__ | gordc: because it's not in requirement list? I think we did not want to add it to not have a hard dep on it | 20:27 |
| jd__ | gordc: we should probably use extras | 20:27 |
| jd__ | gordc: for now I think even having it in requirements is OK | 20:27 |
| gordc | yeah. just not requirements.txt | 20:27 |
| *** fguillot has quit IRC | 20:28 | |
| jd__ | all packagers are going to make the dep on it anyway | 20:28 |
| gordc | ok. i think random people will say 'why do i need zaqar' but i'll let you approve | 20:28 |
| *** fguillot has joined #openstack-telemetry | 20:28 | |
| jd__ | nah I can add it in extras I think | 20:29 |
| jd__ | let me try | 20:29 |
| openstackgerrit | Julien Danjou proposed openstack/aodh: Add pre-signed support for zaqar notifier https://review.openstack.org/335289 | 20:30 |
| gordc | jd__: these are the only ones, i think might be useful for newton: https://review.openstack.org/#/c/368092/ https://review.openstack.org/#/c/367737/ | 20:31 |
| gordc | https://review.openstack.org/#/c/366022/ might be something we'll backport or something. | 20:31 |
| openstackgerrit | Merged openstack/ceilometer: Revert "[install] Create endpoint in one command" https://review.openstack.org/364659 | 20:32 |
| *** paw has quit IRC | 20:33 | |
| openstackgerrit | Doug Hellmann proposed openstack/panko: Update reno for stable/newton https://review.openstack.org/371143 | 20:35 |
| jd__ | ok | 20:36 |
| jd__ | cool thanks gordc | 20:39 |
| jd__ | I think the oslo.messaging prefetch needs more work so we'll see | 20:39 |
| gordc | yeah, agreed. | 20:41 |
| gordc | i'll track the approved patches to see if they fail. | 20:41 |
| *** ametts has quit IRC | 20:42 | |
| jd__ | thanks gordc | 20:43 |
| jd__ | you'll be a great PTL | 20:43 |
| gordc | hahah i know. my dream :P | 20:45 |
| gordc | or nightmare. | 20:45 |
| openstackgerrit | Julien Danjou proposed openstack/ceilometer: agentbase: remove flaky test https://review.openstack.org/371151 | 20:52 |
| EmilienM | what, gordc is going to be ptl? :P | 20:53 |
| gordc | EmilienM: i can't be ptl of telemetry, i'm already running for ptl of puppet | 20:55 |
| EmilienM | you'll do great | 20:56 |
| gordc | thanks for the support | 20:56 |
| gordc | i will put your endorsement in my self-nomination. | 20:56 |
| EmilienM | lol | 20:57 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: implement compressed format for BoundTimeSerie https://review.openstack.org/370849 | 20:58 |
| *** thumpba has quit IRC | 20:59 | |
| *** thumpba has joined #openstack-telemetry | 20:59 | |
| *** fguillot has quit IRC | 21:09 | |
| openstackgerrit | gordon chung proposed openstack/gnocchi: lower processor job polling and handle queuing https://review.openstack.org/370960 | 21:09 |
| *** fguillot has joined #openstack-telemetry | 21:10 | |
| *** thumpba has quit IRC | 21:10 | |
| openstackgerrit | Julien Danjou proposed openstack/panko: doc: remove talk about query/complex queries https://review.openstack.org/371159 | 21:15 |
| *** catintheroof has joined #openstack-telemetry | 21:15 | |
| *** fguillot has quit IRC | 21:21 | |
| zigo | jd__: http://logs.openstack.org/46/371146/2/check/gate-deb-aodh-dpkg-debian-jessie/24b5a16/console.html#_2016-09-15_21_19_29_321923 <--- The result of not approving pragmatism ... :( | 21:22 |
| zigo | jd__: I wonder how long we're going to carry this sillyness. | 21:22 |
| zigo | Otherwise, aodh built well, it seems. :P | 21:22 |
| zigo | (that's b3 ...) | 21:23 |
| jd__ | zigo: until someone fixes setuptools/pbr :( | 21:23 |
| zigo | jd__: Yeah... | 21:23 |
| zigo | jd__: I wonder where in setuptools we could fix it though. | 21:23 |
| jd__ | zigo: we're also trying to config-less in the future | 21:23 |
| jd__ | i.e. works by default with no config file | 21:24 |
| openstackgerrit | Merged openstack/ceilometer: add new meters about some perf events https://review.openstack.org/367737 | 21:24 |
| zigo | jd__: Did you guys removed completely the eventlet api thing? | 21:26 |
| zigo | I mean, the daemon ... | 21:26 |
| jd__ | zigo: hm yeah a while back | 21:27 |
| zigo | :( | 21:27 |
| jd__ | why? | 21:27 |
| zigo | I don't like at all things with apache. | 21:27 |
| zigo | I did it for keystone, and I'm not happy of the result. | 21:27 |
| jd__ | you can use uwsgi if you prefer | 21:27 |
| jd__ | it works well | 21:27 |
| zigo | I guess I should do some uwsgi stuff... | 21:27 |
| zigo | Yup. | 21:27 |
| jd__ | yeah it's easier than Apache too :) | 21:27 |
| *** fguillot has joined #openstack-telemetry | 21:27 | |
| zigo | I'm ":(" just because that's more work! :) | 21:28 |
| *** adriant has joined #openstack-telemetry | 21:28 | |
| zigo | The really bad thing about apache, is that when you restart it, it restarts EVERYTHING that's bound to it. | 21:28 |
| jd__ | indeed | 21:32 |
| *** rvasilets___ has joined #openstack-telemetry | 21:33 | |
| *** paw has joined #openstack-telemetry | 21:33 | |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: implement compressed format for BoundTimeSerie https://review.openstack.org/370849 | 21:42 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: implement compressed format for BoundTimeSerie https://review.openstack.org/370849 | 21:43 |
| *** AndyWojo has quit IRC | 21:45 | |
| *** AJaeger has quit IRC | 21:46 | |
| *** automagically has quit IRC | 21:46 | |
| *** swamireddy has quit IRC | 21:47 | |
| *** mstyne has quit IRC | 21:47 | |
| *** pradk has quit IRC | 21:48 | |
| openstackgerrit | Merged openstack/aodh: Add pre-signed support for zaqar notifier https://review.openstack.org/335289 | 21:50 |
| openstackgerrit | Merged openstack/ceilometer: improve notification processing https://review.openstack.org/275741 | 21:50 |
| *** automagically has joined #openstack-telemetry | 21:51 | |
| *** AndyWojo has joined #openstack-telemetry | 21:51 | |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: cli: set max_workers=1 to run tooz watchers https://review.openstack.org/371174 | 21:52 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: cli: do not run tooz watchers in parallel https://review.openstack.org/371174 | 21:54 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: cli: do not run tooz watchers in parallel https://review.openstack.org/371174 | 21:54 |
| *** mstyne has joined #openstack-telemetry | 21:55 | |
| *** swamireddy has joined #openstack-telemetry | 21:57 | |
| *** AJaeger has joined #openstack-telemetry | 22:00 | |
| openstackgerrit | Merged openstack/ceilometer: [Trivial] Refactor libvirt inspector connection & uri https://review.openstack.org/311081 | 22:04 |
| openstackgerrit | Merged openstack/gnocchi: releasenotes: fix typo in storage-engine-v3 https://review.openstack.org/371091 | 22:04 |
| *** fguillot has quit IRC | 22:16 | |
| *** jkraj has joined #openstack-telemetry | 22:19 | |
| openstackgerrit | Doug Hellmann proposed openstack/aodh: Update reno for stable/newton https://review.openstack.org/371195 | 22:27 |
| *** gordc has quit IRC | 22:29 | |
| *** thumpba has joined #openstack-telemetry | 22:41 | |
| *** rbak has quit IRC | 23:01 | |
| *** jkraj has quit IRC | 23:13 | |
| openstackgerrit | Merged openstack/panko: Update reno for stable/newton https://review.openstack.org/371143 | 23:13 |
| jd__ | sileht: any idea what this crap is http://logs.openstack.org/11/369011/5/check/gate-gnocchi-dsvm-functional-file-mysql-ubuntu-trusty/7289fa5/logs/devstacklog.txt.gz ? :( | 23:18 |
| *** catintheroof has quit IRC | 23:21 | |
| *** thumpba has quit IRC | 23:30 | |
| *** hoonetorg has quit IRC | 23:37 | |
| *** rvasilets___ has quit IRC | 23:37 | |
| openstackgerrit | Merged openstack/ceilometer: tox: Remove useless env https://review.openstack.org/369897 | 23:39 |
| openstackgerrit | Merged openstack/ceilometer: Set a correct number of threads for polling tasks https://review.openstack.org/368092 | 23:39 |
| *** fguillot has joined #openstack-telemetry | 23:54 | |
| *** hoonetorg has joined #openstack-telemetry | 23:55 | |
| *** thumpba has joined #openstack-telemetry | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!