| *** thorst has joined #openstack-telemetry | 00:00 | |
| *** thorst has quit IRC | 00:02 | |
| *** thorst has joined #openstack-telemetry | 00:05 | |
| *** jmlowe has joined #openstack-telemetry | 00:06 | |
| *** thorst has quit IRC | 00:23 | |
| *** thorst has joined #openstack-telemetry | 00:23 | |
| *** harlowja has quit IRC | 00:25 | |
| *** thorst has quit IRC | 00:27 | |
| *** vint_bra has quit IRC | 00:34 | |
| *** thorst has joined #openstack-telemetry | 00:38 | |
| *** jmlowe has quit IRC | 01:00 | |
| *** catintheroof has joined #openstack-telemetry | 01:02 | |
| *** zhurong has joined #openstack-telemetry | 01:11 | |
| *** catintheroof has quit IRC | 01:11 | |
| openstackgerrit | yanghuichan proposed openstack/ceilometer master: Replace six.iteritems() with dict.items() in ceilometer https://review.openstack.org/491391 | 01:11 |
|---|---|---|
| *** hieulq has quit IRC | 01:23 | |
| *** thorst has quit IRC | 01:24 | |
| *** jmlowe has joined #openstack-telemetry | 01:24 | |
| *** rwsu has quit IRC | 01:32 | |
| *** rwsu has joined #openstack-telemetry | 01:44 | |
| *** lhx__ has joined #openstack-telemetry | 02:01 | |
| *** Marx314 has quit IRC | 02:16 | |
| *** aagate has quit IRC | 02:18 | |
| *** tovin07_ has joined #openstack-telemetry | 02:20 | |
| *** ddyer has quit IRC | 02:25 | |
| *** aagate has joined #openstack-telemetry | 02:26 | |
| *** dixiaoli has joined #openstack-telemetry | 03:39 | |
| *** dave-mccowan has quit IRC | 03:47 | |
| *** links has joined #openstack-telemetry | 03:53 | |
| *** thorst has joined #openstack-telemetry | 04:25 | |
| *** thorst has quit IRC | 04:30 | |
| *** harlowja has joined #openstack-telemetry | 04:35 | |
| *** harlowja has quit IRC | 05:14 | |
| *** sanchitmalhotra has quit IRC | 05:26 | |
| *** sanchitmalhotra has joined #openstack-telemetry | 05:27 | |
| *** Gautam has joined #openstack-telemetry | 05:27 | |
| *** psachin has joined #openstack-telemetry | 05:34 | |
| *** lhx__ has quit IRC | 05:36 | |
| *** yprokule has joined #openstack-telemetry | 05:44 | |
| openstackgerrit | yanghuichan proposed openstack/ceilometer master: Replace six.iteritems() with dict.items() in ceilometer https://review.openstack.org/491391 | 05:45 |
| *** lhx__ has joined #openstack-telemetry | 05:57 | |
| *** thorst has joined #openstack-telemetry | 05:58 | |
| *** thorst has quit IRC | 06:03 | |
| openstackgerrit | JeffreyGuan proposed openstack/aodh master: Typo input in "aodh alarm create" for param "--name". Update the typo "instance_on_but_in err_state" to "instance_on_but_in_err_state". Closes-Bug: #1709241 https://review.openstack.org/491674 | 06:07 |
| openstack | bug 1709241 in Aodh "typo input in "aodh alarm create" for param "--name"" [Undecided,In progress] https://launchpad.net/bugs/1709241 - Assigned to Jeffrey Guan (double12gzh) | 06:07 |
| *** Gautam has quit IRC | 06:07 | |
| *** Gautam has joined #openstack-telemetry | 06:15 | |
| *** rcernin has joined #openstack-telemetry | 06:21 | |
| *** Gautam_ has joined #openstack-telemetry | 06:28 | |
| *** Gautam has quit IRC | 06:29 | |
| *** dixiaoli has quit IRC | 06:39 | |
| *** flg_ has joined #openstack-telemetry | 06:46 | |
| *** flg has joined #openstack-telemetry | 06:48 | |
| *** zhurong has quit IRC | 06:59 | |
| *** pcaruana has joined #openstack-telemetry | 07:00 | |
| openstackgerrit | wenbin luo proposed openstack/ceilometer master: Fix typos in ceilometer https://review.openstack.org/491689 | 07:01 |
| *** tesseract has joined #openstack-telemetry | 07:21 | |
| *** thorst has joined #openstack-telemetry | 07:59 | |
| *** thorst has quit IRC | 08:04 | |
| *** shardy has joined #openstack-telemetry | 08:06 | |
| *** flg has quit IRC | 08:14 | |
| *** flg_ has quit IRC | 08:16 | |
| *** flg_ has joined #openstack-telemetry | 08:34 | |
| *** flg has joined #openstack-telemetry | 08:34 | |
| *** snecklifter has joined #openstack-telemetry | 08:41 | |
| *** flg has quit IRC | 09:09 | |
| *** flg_ has quit IRC | 09:09 | |
| *** Gautam_ has quit IRC | 09:19 | |
| *** Gautam has joined #openstack-telemetry | 09:24 | |
| *** Gautam_ has joined #openstack-telemetry | 09:26 | |
| *** Gautam has quit IRC | 09:29 | |
| *** shardy has quit IRC | 09:43 | |
| *** yassine has quit IRC | 09:48 | |
| *** Gautam has joined #openstack-telemetry | 09:53 | |
| *** Gautam_ has quit IRC | 09:55 | |
| *** shardy has joined #openstack-telemetry | 09:56 | |
| *** flg_ has joined #openstack-telemetry | 09:57 | |
| *** flg has joined #openstack-telemetry | 09:58 | |
| *** thorst has joined #openstack-telemetry | 10:00 | |
| *** tovin07_ has quit IRC | 10:02 | |
| *** thorst has quit IRC | 10:05 | |
| *** dr_gogeta86 has quit IRC | 10:09 | |
| *** dr_gogeta86 has joined #openstack-telemetry | 10:12 | |
| *** dr_gogeta86 has quit IRC | 10:12 | |
| *** dr_gogeta86 has joined #openstack-telemetry | 10:12 | |
| *** Gautam_ has joined #openstack-telemetry | 10:19 | |
| *** Gautam has quit IRC | 10:22 | |
| *** zhurong has joined #openstack-telemetry | 10:24 | |
| *** lhx__ has quit IRC | 10:28 | |
| *** lhx__ has joined #openstack-telemetry | 10:28 | |
| openstackgerrit | liuwei proposed openstack/ceilometer master: Add meters for libvirt: disk.device.read.latency, disk.device.write.latency https://review.openstack.org/489910 | 10:34 |
| *** lhx__ has quit IRC | 10:40 | |
| *** thorst has joined #openstack-telemetry | 10:42 | |
| *** thorst has quit IRC | 10:54 | |
| *** thorst has joined #openstack-telemetry | 10:54 | |
| *** thorst has quit IRC | 10:59 | |
| *** lhx_ has joined #openstack-telemetry | 11:16 | |
| *** gordc has joined #openstack-telemetry | 11:37 | |
| *** dave-mccowan has joined #openstack-telemetry | 11:40 | |
| *** thorst has joined #openstack-telemetry | 11:51 | |
| *** yassine has joined #openstack-telemetry | 11:54 | |
| *** psachin has quit IRC | 11:58 | |
| *** psachin has joined #openstack-telemetry | 12:00 | |
| *** rwsu has quit IRC | 12:28 | |
| *** rwsu has joined #openstack-telemetry | 12:29 | |
| *** zhurong has quit IRC | 12:30 | |
| *** flg_ has quit IRC | 12:42 | |
| *** flg has quit IRC | 12:42 | |
| *** leitan has joined #openstack-telemetry | 12:45 | |
| *** lhx_ has quit IRC | 12:53 | |
| *** pradk has joined #openstack-telemetry | 12:54 | |
| *** flg has joined #openstack-telemetry | 12:58 | |
| *** flg_ has joined #openstack-telemetry | 12:58 | |
| *** catintheroof has joined #openstack-telemetry | 13:00 | |
| openstackgerrit | Samuel Pilla proposed openstack/ceilometer master: Add bandit to ceilometer tests https://review.openstack.org/490621 | 13:03 |
| *** pradk has quit IRC | 13:06 | |
| *** links has quit IRC | 13:11 | |
| *** Liuqing has joined #openstack-telemetry | 13:17 | |
| *** lhx_ has joined #openstack-telemetry | 13:19 | |
| *** jmlowe has quit IRC | 13:28 | |
| *** vint_bra has joined #openstack-telemetry | 13:36 | |
| *** Gautam_ has quit IRC | 13:38 | |
| *** flg has quit IRC | 13:55 | |
| *** flg_ has quit IRC | 13:55 | |
| openstackgerrit | gordon chung proposed openstack/aodh master: Typo input in "aodh alarm create" for param "--name" https://review.openstack.org/491674 | 13:57 |
| *** Liuqing has quit IRC | 13:58 | |
| *** p01nt3r75 has joined #openstack-telemetry | 13:59 | |
| *** jmlowe has joined #openstack-telemetry | 14:11 | |
| *** flg_ has joined #openstack-telemetry | 14:11 | |
| *** flg has joined #openstack-telemetry | 14:12 | |
| *** flg_ has quit IRC | 14:46 | |
| *** flg has quit IRC | 14:48 | |
| *** links has joined #openstack-telemetry | 14:48 | |
| *** links has quit IRC | 14:49 | |
| *** snecklifter has quit IRC | 14:55 | |
| *** psachin has quit IRC | 15:01 | |
| *** yprokule has quit IRC | 15:03 | |
| *** pradk has joined #openstack-telemetry | 15:05 | |
| *** pradk has quit IRC | 15:07 | |
| *** benoit has joined #openstack-telemetry | 15:22 | |
| benoit | hello | 15:25 |
| benoit | i have a large number of measures to proceed with gnocchi (300k env), how can i delete all pending measures to proceed ? | 15:26 |
| gordc | benoit: what version of gnocchi you have? | 15:28 |
| gordc | and what backend storage driver? | 15:29 |
| benoit | gordc: 3.0.11 / swift | 15:29 |
| gordc | benoit: 300K env means you have 300K measures to be processed? | 15:30 |
| benoit | gordc: yep (gnocchi status output) | 15:30 |
| gordc | benoit: there isn't really a way to clear pending (becasue that's essentially throwing away data) | 15:30 |
| benoit | gordc: gnocchi-metricd process were freezed for a while | 15:31 |
| gordc | i would probably suggest you make your processing_delay shorter... and/or hack the code https://github.com/gnocchixyz/gnocchi/blob/stable/3.0/gnocchi/cli.py#L142 to use a higher TASKS_PER_WORKER | 15:32 |
| *** jmlowe has quit IRC | 15:32 | |
| gordc | what do you mean freeze? | 15:32 |
| benoit | gordc: is that a issue throwing them away ? (if do not care of this metrics for now ?) | 15:32 |
| *** jmlowe has joined #openstack-telemetry | 15:32 | |
| gordc | benoit: i guess not.. you'll need to delete some containers in swift | 15:33 |
| gordc | probably easier to just disable gnocchi-api and let metricd clean it for you. | 15:33 |
| benoit | gordc: i have 3x16 workers, and processing delay to 20, tweaking them, is just to much IO intensive | 15:34 |
| benoit | gordc: are measures to proceed already in swift ? i thought metrics to process was to store them in swift (metricd job) ? did i missunderstood the process ? | 15:35 |
| gordc | benoit: how many metrics does it say you have? (not measures)? | 15:35 |
| benoit | 800 metrics / 261 000 measures | 15:36 |
| gordc | benoit: in v3, pending measures are stored in same place as processed measures. | 15:36 |
| benoit | (sorry for using metrics instead of measures) | 15:36 |
| gordc | benoit: there's a 'measure' container in your swift that you'll need to delete | 15:36 |
| gordc | benoit: i think if you change your TASKS_PER_WORKER to something larger, it might help (the scheduling isn't good in v3.x) | 15:38 |
| gordc | but yeah, swift driver has a lot of IO... only ceph/redis drivers have solution that minimises IO | 15:39 |
| benoit | gordc: ok for TASKS_PER_WORKER, it's actually set to 16 - how much can I raise it ? | 15:40 |
| benoit | checking also how to find "measure" container to delete | 15:40 |
| gordc | benoit: depends how many active metrics you're capturing and how many workers you have. see: scheduling in v3.x is subpar :( | 15:41 |
| gordc | benoit: if it's 800 metrics, i would raise it 2x (if you keep 3x16 metricd) | 15:42 |
| benoit | gordc: ok, i will check that. i have disable some metrics i do not care (principally only capturing instance ones) | 15:43 |
| benoit | btw i find the measure container :p - there is also a lot of "gnocchi.XXX" containers | 15:44 |
| gordc | kk, yeah, play around with that. (or upgrade to v4 to avoid the hacking) | 15:44 |
| gordc | the gnocchi.XXX is the containers which store the actual values you get when you GET measures | 15:45 |
| benoit | do i just need to do "swift -U service:gnocchi delete measure" ? | 15:45 |
| gordc | measure container is an intermediate step | 15:45 |
| benoit | ok | 15:45 |
| gordc | ... i don't actually use swift driver so i don't know swift command | 15:46 |
| benoit | ok lets try it :p | 15:46 |
| gordc | is your measure container full of measures with the same metric_id? | 15:46 |
| benoit | i just launch the delete command | 15:47 |
| gordc | in that case, i think increasing TASKS_PER_WORKER will clear up the 300K measrues really fast (since it processes them in bulk) | 15:47 |
| gordc | lol nm | 15:47 |
| gordc | gnocchi should recreate that container if you restart. | 15:48 |
| benoit | ok, delete is still ongoing | 15:52 |
| *** p01nt3r75 has quit IRC | 16:06 | |
| *** yassine has quit IRC | 16:10 | |
| benoit | gordc: still deleting - i dont know how long it will take - thanks for your help | 16:14 |
| gordc | wow, it takes that long to delete?lol | 16:16 |
| gordc | should try bulk delete next time :P | 16:16 |
| benoit | lol, yes i m surprise it take so long ... | 16:17 |
| *** dewanee has joined #openstack-telemetry | 16:20 | |
| *** pcaruana has quit IRC | 16:27 | |
| *** rcernin has quit IRC | 16:31 | |
| *** tesseract has quit IRC | 16:31 | |
| *** shardy has quit IRC | 16:46 | |
| *** rwsu has quit IRC | 17:04 | |
| *** aagate has quit IRC | 17:35 | |
| *** aagate has joined #openstack-telemetry | 17:37 | |
| *** aagate has quit IRC | 17:48 | |
| *** pradk has joined #openstack-telemetry | 17:48 | |
| *** aagate has joined #openstack-telemetry | 17:49 | |
| *** pradk has quit IRC | 17:52 | |
| *** lhx_ has quit IRC | 18:02 | |
| *** ddyer has joined #openstack-telemetry | 18:19 | |
| *** vishwanathj has quit IRC | 18:36 | |
| *** vishwanathj has joined #openstack-telemetry | 18:37 | |
| *** flg_ has joined #openstack-telemetry | 19:11 | |
| *** flg has joined #openstack-telemetry | 19:11 | |
| *** jmlowe has quit IRC | 19:54 | |
| gordc | NostawRm: if you didn't figure it out, you should search on instance_id http://gnocchi.xyz/rest.html#searching-for-resources | 20:10 |
| NostawRm | wow, really not a lot of activity here :P I'm still doing what I was because I don't believe that instance_network_interfaces doesn't have just the direct instance_id anywhere, trying to double check now but i recall looking for it | 20:12 |
| NostawRm | gordc, i mean, I guess it is there, suppose I could do a 'like' to do just the instance id in the original_resource_id but I've already got it to get the entire original_resource_id | 20:14 |
| gordc | canadian holiday... and yeah. there are very few devs... and not in north america. | 20:14 |
| NostawRm | lol I was just surprised that my question wasn't pushed out of my scroll back | 20:14 |
| NostawRm | a lot of the other channels would've done it on joins/parts alone | 20:15 |
| gordc | NostawRm: i was looking at logs. :) | 20:15 |
| gordc | i vaguely recall it should map to original_resource_id but i'm not 100%. i could just be making that up | 20:16 |
| NostawRm | its a shame there's not more activity, the idea behind it all is great, addressing a problem I didn't even consider. I wish my plans were to use it to its fullest but right now we're only using it to track monthly bandwidth usage :/ | 20:17 |
| NostawRm | original_resource_id for an interface ends up a real mess, like instance-000000d2-12831df5-442e-45da-a42c-9aabfffe1490-tapc2631880-1d | 20:17 |
| NostawRm | which, iirc, is '<libvirt instance name>-<openstackuuid>-tap<first 11 characters of port uuid> | 20:18 |
| gordc | yeah, that's because of ceilometer... there is no unique id for network iirc | 20:18 |
| gordc | there's https://bugs.launchpad.net/ceilometer/+bug/1383123 | 20:19 |
| openstack | Launchpad bug 1383123 in Ceilometer "vnic's resource_id contains redundant information" [Medium,Triaged] | 20:19 |
| gordc | but we never figured out how to handle it better without breaking old code... or what we have is our attempt | 20:19 |
| NostawRm | no complaints really, I'm using Rackspaces 'php-opencloud' library and I was trying to see if there was an option other than searching to get the resource by original_resource_id since it there should only ever be one that matches | 20:21 |
| NostawRm | but that was a minor complaint and "fixing" it in gnochhi would just move where the same logic is done | 20:22 |
| gordc | cool cool. if you have a workaround then that's good. i don't really see it being fixed on ceilometer pov. maybe gnocchi if you think it can be done there. | 20:23 |
| NostawRm | nah, no reason to, I get tunnel vision and make mountains out of mole hills | 20:23 |
| gordc | :) | 20:24 |
| *** jmlowe has joined #openstack-telemetry | 20:28 | |
| openstackgerrit | Merged openstack/ceilometer master: fix cpu_util precision is too precise https://review.openstack.org/490278 | 20:29 |
| *** vint_bra has quit IRC | 20:59 | |
| *** vint_bra has joined #openstack-telemetry | 20:59 | |
| *** thorst has quit IRC | 21:37 | |
| *** gordc has quit IRC | 22:37 | |
| *** thorst has joined #openstack-telemetry | 22:38 | |
| *** leitan has quit IRC | 22:40 | |
| *** thorst has quit IRC | 22:43 | |
| *** persia has joined #openstack-telemetry | 23:22 | |
| persia | Good day. The PTL election cycle nomination period ends tomorrow, and there is nobody who is standing as PTL for Telemetry. Does anyone know who might be a good candidate? | 23:22 |
| *** catintheroof has quit IRC | 23:29 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!