*** catintheroof has joined #openstack-telemetry | 00:03 | |
*** iceyao has joined #openstack-telemetry | 00:07 | |
*** iceyao has quit IRC | 00:09 | |
*** catintheroof has quit IRC | 00:16 | |
openstackgerrit | gordon chung proposed openstack/gnocchi master: fix redis storage delete https://review.openstack.org/446794 | 00:45 |
---|---|---|
openstackgerrit | gordon chung proposed openstack/gnocchi master: test that storage is deleted https://review.openstack.org/446767 | 00:46 |
openstackgerrit | liusheng proposed openstack/python-pankoclient master: Modify the doc descriptions of pankoclient https://review.openstack.org/441848 | 01:00 |
*** catintheroof has joined #openstack-telemetry | 01:02 | |
*** catintheroof has quit IRC | 01:02 | |
*** catintheroof has joined #openstack-telemetry | 01:02 | |
*** donghao has joined #openstack-telemetry | 01:15 | |
*** leitan has joined #openstack-telemetry | 01:15 | |
*** donghao has quit IRC | 01:20 | |
*** vint_bra has joined #openstack-telemetry | 01:27 | |
*** lhx__ has joined #openstack-telemetry | 01:39 | |
*** ddyer has quit IRC | 02:04 | |
*** leitan has quit IRC | 02:14 | |
*** leitan has joined #openstack-telemetry | 02:16 | |
*** donghao has joined #openstack-telemetry | 02:18 | |
*** vint_bra has quit IRC | 02:18 | |
*** ddyer has joined #openstack-telemetry | 02:20 | |
openstackgerrit | gordon chung proposed openstack/gnocchi master: fix redis storage delete https://review.openstack.org/446794 | 02:21 |
openstackgerrit | gordon chung proposed openstack/gnocchi master: test that storage is deleted https://review.openstack.org/446767 | 02:21 |
*** donghao has quit IRC | 02:24 | |
*** masber has joined #openstack-telemetry | 02:26 | |
*** lhx__ has quit IRC | 02:31 | |
*** lhx__ has joined #openstack-telemetry | 02:31 | |
*** catintheroof has quit IRC | 02:36 | |
*** lhx__ has quit IRC | 02:52 | |
*** lhx__ has joined #openstack-telemetry | 02:53 | |
*** donghao has joined #openstack-telemetry | 03:22 | |
*** rwsu has joined #openstack-telemetry | 03:23 | |
*** chlong has quit IRC | 03:25 | |
*** donghao has quit IRC | 03:27 | |
*** leitan has quit IRC | 03:29 | |
*** leitan has joined #openstack-telemetry | 03:29 | |
*** leitan has quit IRC | 03:30 | |
*** adriant has joined #openstack-telemetry | 03:34 | |
*** flwang1 has joined #openstack-telemetry | 03:38 | |
*** lhx__ has quit IRC | 03:42 | |
*** flwang1 has quit IRC | 03:56 | |
*** dschultz has joined #openstack-telemetry | 04:01 | |
*** donghao has joined #openstack-telemetry | 04:25 | |
*** gongysh has joined #openstack-telemetry | 04:27 | |
*** donghao has quit IRC | 04:31 | |
*** Gautam has joined #openstack-telemetry | 05:20 | |
*** Jack_Iv has joined #openstack-telemetry | 05:24 | |
*** donghao has joined #openstack-telemetry | 05:30 | |
*** donghao has quit IRC | 05:34 | |
*** nadya has joined #openstack-telemetry | 05:41 | |
*** lhx__ has joined #openstack-telemetry | 05:43 | |
*** lhx__ has quit IRC | 05:43 | |
*** lhx__ has joined #openstack-telemetry | 05:44 | |
*** lhx__ has quit IRC | 05:45 | |
*** lhx__ has joined #openstack-telemetry | 05:45 | |
*** nadya has quit IRC | 05:46 | |
*** lhx__ has quit IRC | 05:50 | |
*** yprokule has joined #openstack-telemetry | 05:58 | |
openstackgerrit | fengchaoyang proposed openstack/gnocchi master: Modify outdated parameters comment https://review.openstack.org/446852 | 06:08 |
*** zhurong has joined #openstack-telemetry | 06:21 | |
openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/ceilometer master: DNM: project-config testing https://review.openstack.org/444391 | 06:33 |
*** adriant has quit IRC | 06:34 | |
*** gongysh has quit IRC | 07:07 | |
*** gongysh has joined #openstack-telemetry | 07:10 | |
*** lhx__ has joined #openstack-telemetry | 07:14 | |
*** basilAB_ has quit IRC | 07:18 | |
Gautam | Hi sileht: I wanted to check with you if translations for Panko has been set up. There's no locale folder in panko source. There's not even english .po file present. I see some progress at zanata, but not sure if translations for Panko is in pipeline. | 07:19 |
*** nadya has joined #openstack-telemetry | 07:21 | |
*** vaishali has joined #openstack-telemetry | 07:23 | |
*** basilAB has quit IRC | 07:23 | |
*** basilAB has joined #openstack-telemetry | 07:26 | |
*** nadya_ has joined #openstack-telemetry | 07:27 | |
*** nadya has quit IRC | 07:29 | |
*** leitan has joined #openstack-telemetry | 07:30 | |
*** donghao has joined #openstack-telemetry | 07:34 | |
*** tesseract has joined #openstack-telemetry | 07:34 | |
*** leitan has quit IRC | 07:35 | |
*** donghao has quit IRC | 07:41 | |
*** pcaruana has joined #openstack-telemetry | 07:43 | |
*** Jack_Iv has quit IRC | 07:52 | |
*** Jack_Iv has joined #openstack-telemetry | 07:52 | |
openstackgerrit | Merged openstack/ceilometer master: tests: simplify broken test https://review.openstack.org/446335 | 07:54 |
*** Jack_Iv has quit IRC | 07:56 | |
*** Jack_Iv has joined #openstack-telemetry | 08:01 | |
*** Jack_Iv has quit IRC | 08:22 | |
*** Jack_Iv has joined #openstack-telemetry | 08:23 | |
*** Jack_Iv has quit IRC | 08:27 | |
*** zhurong has quit IRC | 08:27 | |
*** lhx__ has quit IRC | 08:29 | |
*** lhx__ has joined #openstack-telemetry | 08:29 | |
sileht | Gautam, I have never looked at how translations works, I just merge the zanata reviews. | 08:30 |
*** lhx__ has quit IRC | 08:31 | |
Gautam | sileht: Thanks. Probably gordc will know.. will ping him later. | 08:31 |
openstackgerrit | lan proposed openstack/ceilometer master: use HostAddressOpt for opts that accept IP and hostnames https://review.openstack.org/446886 | 08:35 |
*** amoralej|off is now known as amoralej | 08:38 | |
*** xianghui has joined #openstack-telemetry | 08:42 | |
xianghui | Hi guys, anyone have used 'POST /v2/meters/(meter_name)' to post Samples to Telemetry? | 08:45 |
xianghui | I have done this step, and can see ceilometer api received this post | 08:45 |
xianghui | but can't find in ceilometer sample-list and meter-list | 08:45 |
xianghui | is there any additional steps needed to filter them out? | 08:45 |
openstackgerrit | lan proposed openstack/ceilometer master: use HostAddressOpt for opts that accept IP and hostnames. https://review.openstack.org/446886 | 08:51 |
*** shardy has joined #openstack-telemetry | 09:01 | |
openstackgerrit | lan proposed openstack/ceilometer master: 1.use HostAddressOpt for opts that accept IP and hostnames. 2.add default value. https://review.openstack.org/446886 | 09:04 |
*** nijaba has quit IRC | 09:05 | |
*** nijaba has joined #openstack-telemetry | 09:12 | |
*** jahsis has joined #openstack-telemetry | 09:19 | |
*** Gautam has quit IRC | 09:20 | |
*** zhangguoqing has joined #openstack-telemetry | 09:24 | |
*** Gautam has joined #openstack-telemetry | 09:24 | |
*** zhangguoqing has quit IRC | 09:28 | |
jahsis | hello all. In panko are there API call where admin user can receive all events for all projects? | 09:28 |
*** donghao has joined #openstack-telemetry | 09:43 | |
*** zhangguoqing has joined #openstack-telemetry | 09:47 | |
*** donghao has quit IRC | 09:47 | |
*** Gautam has quit IRC | 09:50 | |
*** Gautam has joined #openstack-telemetry | 09:51 | |
*** Gautam has quit IRC | 09:55 | |
*** nadya_ has quit IRC | 09:56 | |
openstackgerrit | Merged openstack/gnocchi master: fix redis storage delete https://review.openstack.org/446794 | 10:01 |
openstackgerrit | Merged openstack/gnocchi master: test that storage is deleted https://review.openstack.org/446767 | 10:01 |
openstackgerrit | Merged openstack/gnocchi master: Modify outdated parameters comment https://review.openstack.org/446852 | 10:01 |
*** nadya has joined #openstack-telemetry | 10:03 | |
*** rcernin has joined #openstack-telemetry | 10:04 | |
openstackgerrit | Merged openstack/ceilometer master: libvirt: rewrite the error handling https://review.openstack.org/398297 | 10:14 |
openstackgerrit | Julien Danjou proposed openstack/ceilometer master: coordination: create coordinator at init time https://review.openstack.org/443269 | 10:24 |
openstackgerrit | Julien Danjou proposed openstack/ceilometer master: Replace Ceilometer coordination layer by tooz partition system https://review.openstack.org/446016 | 10:24 |
openstackgerrit | Julien Danjou proposed openstack/ceilometer master: coordination: remove group_id check https://review.openstack.org/443271 | 10:24 |
openstackgerrit | Julien Danjou proposed openstack/ceilometer master: coordination: stop checking for _coordinator to be None https://review.openstack.org/443270 | 10:24 |
openstackgerrit | Julien Danjou proposed openstack/ceilometer master: coordination: remove started check https://review.openstack.org/443277 | 10:24 |
openstackgerrit | Julien Danjou proposed openstack/ceilometer master: tests: stop hammering CPU while waiting for sample to wait https://review.openstack.org/446015 | 10:24 |
openstackgerrit | Julien Danjou proposed openstack/ceilometer master: coordination: make group_id to never be None https://review.openstack.org/443273 | 10:24 |
*** cdent has joined #openstack-telemetry | 10:38 | |
*** zhangguoqing has quit IRC | 10:38 | |
*** zhangguoqing has joined #openstack-telemetry | 10:44 | |
*** rcernin has quit IRC | 10:46 | |
*** donghao has joined #openstack-telemetry | 10:46 | |
*** donghao has quit IRC | 10:50 | |
*** gongysh has quit IRC | 10:56 | |
*** xianghui has quit IRC | 11:06 | |
*** zhurong has joined #openstack-telemetry | 11:13 | |
*** zhangguo_ has joined #openstack-telemetry | 11:22 | |
*** zhangguoqing has quit IRC | 11:26 | |
*** Jack_Iv has joined #openstack-telemetry | 11:40 | |
*** Jack_Iv has quit IRC | 11:40 | |
*** Jack_Iv has joined #openstack-telemetry | 11:40 | |
*** zhangguoqing has joined #openstack-telemetry | 11:47 | |
*** zhangguo_ has quit IRC | 11:51 | |
*** Gautam has joined #openstack-telemetry | 11:52 | |
jahsis | i try to use gnocchi publisher, but all the time receive error | 11:59 |
jahsis | NoMatches: No 'ceilometer.publisher' driver found, looking for 'gnocchi' | 11:59 |
*** shardy is now known as shardy_lunch | 12:04 | |
*** nadya has quit IRC | 12:04 | |
*** leitan has joined #openstack-telemetry | 12:10 | |
*** Jack_Iv_ has joined #openstack-telemetry | 12:15 | |
*** Jack_Iv has quit IRC | 12:18 | |
*** donghao has joined #openstack-telemetry | 12:27 | |
*** nadya has joined #openstack-telemetry | 12:31 | |
*** gordc has joined #openstack-telemetry | 12:31 | |
*** Jack_Iv_ has quit IRC | 12:33 | |
*** nadya has quit IRC | 12:47 | |
*** Jack_Iv has joined #openstack-telemetry | 12:52 | |
*** xiaofandh12 has joined #openstack-telemetry | 12:54 | |
*** donghao has quit IRC | 12:55 | |
*** zhangguo_ has joined #openstack-telemetry | 12:57 | |
*** zhangguoqing has quit IRC | 12:57 | |
*** catintheroof has joined #openstack-telemetry | 13:10 | |
*** catintheroof has quit IRC | 13:10 | |
*** catintheroof has joined #openstack-telemetry | 13:10 | |
*** nadya has joined #openstack-telemetry | 13:11 | |
*** nadya has quit IRC | 13:20 | |
*** nadya has joined #openstack-telemetry | 13:25 | |
*** nadya_ has joined #openstack-telemetry | 13:27 | |
*** nadya has quit IRC | 13:28 | |
*** dave-mccowan has joined #openstack-telemetry | 13:30 | |
*** chlong has joined #openstack-telemetry | 13:32 | |
*** nadya_ has quit IRC | 13:33 | |
*** nadya has joined #openstack-telemetry | 13:34 | |
*** zhangguo_ has quit IRC | 13:45 | |
*** zhangguoqing has joined #openstack-telemetry | 13:47 | |
*** lhx_ has joined #openstack-telemetry | 13:50 | |
*** shardy_lunch is now known as shardy | 13:56 | |
*** zhurong has quit IRC | 13:56 | |
*** gongysh has joined #openstack-telemetry | 14:06 | |
Gautam | Hi gordc, I wanted to check with you if translations has been set for Panko. There is no locale folder in panko. There isn't an english .po file. I could see something at zanata for panko master, but not sure if translations for panko are in pipeline. | 14:12 |
gordc | Gautam: i don't think so? we don't have anyone actively developing it | 14:16 |
gordc | you can add it if you want... but it's not important to those who contribute to this project currently | 14:16 |
Gautam | oh ok. Thanks for the info. We are trying to use Panko and realized about the translations. We will see what we can do. Thanks! | 14:19 |
*** nadya has quit IRC | 14:19 | |
gordc | cool cool. | 14:21 |
*** chlong has quit IRC | 14:23 | |
*** jmlowe has joined #openstack-telemetry | 14:24 | |
jmlowe | Quick question, are the dispatchers mutually exclusive? Can I have panko and http? | 14:24 |
gordc | yep | 14:25 |
gordc | well you can multiple publishers (in ocata+ context) | 14:26 |
*** zhangguo_ has joined #openstack-telemetry | 14:27 | |
*** zhangguoqing has quit IRC | 14:27 | |
jmlowe | ok yeah, that's what I was looking for, sending off metrics and events to a 3rd party while keeping gnocchi and panko for in house use | 14:29 |
jmlowe | that's new for ocata? | 14:30 |
gordc | jmlowe: yeah, you can just define multiple publishers for the pipeline. (or multiple dispatchers if you are using collector but it has coarser control) | 14:30 |
gordc | no, you should be able to do it in all versions | 14:31 |
gordc | or at least past few years. | 14:31 |
jmlowe | ok, perfect, on newton now | 14:31 |
*** amoralej is now known as amoralej|lunch | 14:34 | |
*** zhangguoqing has joined #openstack-telemetry | 14:36 | |
gordc | jmlowe: you might need to either backport http publisher to newton or use dispatcher | 14:37 |
*** zhangguo_ has quit IRC | 14:37 | |
gordc | i vaguely remember the http publisher being weird previously. | 14:37 |
*** gongysh has quit IRC | 14:39 | |
jmlowe | ok, looking at the timeline for xdmod (http://open.xdmod.org/) I think it lines up with my existing plans for ocata upgrade right after the Boston summit | 14:40 |
*** xiaofandh12 has quit IRC | 14:40 | |
gordc | hmm... never tried that before | 14:42 |
*** donghao has joined #openstack-telemetry | 14:45 | |
*** leitan has quit IRC | 14:46 | |
openstackgerrit | gordon chung proposed openstack/gnocchi master: redis: use hashes for aggregated storage https://review.openstack.org/447019 | 14:46 |
*** leitan has joined #openstack-telemetry | 14:46 | |
jmlowe | they have, let's call it, a challenging task before them, adapting their traditional hpc batch system metrics to cloud and they are starting with eucalyptus | 14:47 |
*** Gautam has quit IRC | 14:50 | |
openstackgerrit | Eric Berglund proposed openstack/ceilometer master: Tenacity defined twice in requirements https://review.openstack.org/447022 | 14:51 |
openstackgerrit | Eric Berglund proposed openstack/ceilometer master: Remove second tenacity in requirements https://review.openstack.org/447022 | 14:52 |
jmlowe | they said their ingestion rate was 300/min currently, I think they have a lot of work to do before they can take metrics from our 600-1200 vm's | 14:54 |
gordc | jmlowe: is this generating metrics or storing metrics? | 14:55 |
*** Gautam has joined #openstack-telemetry | 14:55 | |
gordc | i need to start working on hpc workload stuff and i know nothing about it.lol | 14:55 |
jmlowe | they want to ingest and store | 14:55 |
*** leitan has quit IRC | 14:57 | |
gordc | hmm.. i see. i guess i don't really need that part then. i guess something for us to look at to see if we can be 'inspired' aka copy. | 14:57 |
jmlowe | hpc workloads are all about driving the system usage up past 95%, that's the threshold the funding agencies set for batch systems they purchase | 14:57 |
jmlowe | single tenancy per compute node and anti affinity in nova terms are the rule | 14:58 |
gordc | you folks use Condor for scheduling? | 15:01 |
*** zhangguoqing has quit IRC | 15:01 | |
gordc | (only one i actually know something about) | 15:01 |
jmlowe | then you use a scheduler to place the larger jobs in 2 dimensions, nodes and time, use small jobs to backfill the holes, push the big jobs around a little bit to get the tightest packing, fairshare to degrade the priority of users who have been using lots of system time so nobody gets starved | 15:02 |
jmlowe | condor is kind of adjacent, it's a cycle scavenger, like seti or folding at home | 15:02 |
jmlowe | the hpc world is changing a bit, used to be all about the interconnect | 15:03 |
*** donghao has quit IRC | 15:03 | |
gordc | jmlowe: guess it's good i don't neeed to worry about forgeting the old stuff... dont know it.lol | 15:04 |
jmlowe | now we are in the gpu era, the $10M 20 rack 1024 node power5 system that was our flagship system when I started delivered 40TFOPS and hit #23 on the top 500 supercomputer list, you can now get a single gpu instance on AWS that will deliver 47TFLOPS | 15:05 |
gordc | ... | 15:06 |
* gordc wonders how big jmlowe's budget is. :) | 15:06 | |
jmlowe | for my current project, $6.5M | 15:06 |
jmlowe | 2 clouds of 320 nodes and 960 TB of ceph each | 15:07 |
gordc | hahah, wasn't expecting a real answer but i guess that data is public? | 15:07 |
*** dewanee_ is now known as dewanee | 15:07 | |
jmlowe | yep, nearly everything we do is public | 15:08 |
jmlowe | configs, salaries, emails, reports | 15:08 |
*** leitan has joined #openstack-telemetry | 15:10 | |
* gordc looks in docs for "party fund": $1 million. | 15:12 | |
*** amoralej|lunch is now known as amoralej | 15:13 | |
*** dave-mccowan has quit IRC | 15:15 | |
jd_ | sileht: https://bugs.launchpad.net/bugs/1673798 | 15:17 |
openstack | Launchpad bug 1673798 in Ceilometer "Tenacity defined twice in requirements.txt" [Undecided,In progress] - Assigned to Eric Berglund (esberglu) | 15:17 |
*** Gautam has quit IRC | 15:18 | |
*** Gautam has joined #openstack-telemetry | 15:18 | |
*** Gautam has quit IRC | 15:19 | |
*** Gautam has joined #openstack-telemetry | 15:19 | |
*** dewanee has quit IRC | 15:20 | |
*** Jack_Iv has quit IRC | 15:20 | |
*** dewanee has joined #openstack-telemetry | 15:20 | |
*** Jack_Iv has joined #openstack-telemetry | 15:21 | |
*** dewanee is now known as mattia | 15:22 | |
*** mattia is now known as Guest38487 | 15:22 | |
*** yprokule has quit IRC | 15:22 | |
*** Guest38487 has quit IRC | 15:23 | |
*** dewanee has joined #openstack-telemetry | 15:23 | |
*** dewanee has quit IRC | 15:24 | |
*** dewanee has joined #openstack-telemetry | 15:24 | |
*** dave-mccowan has joined #openstack-telemetry | 15:28 | |
openstackgerrit | gordon chung proposed openstack/gnocchi master: default to v3 format https://review.openstack.org/447032 | 15:30 |
*** vint_bra has joined #openstack-telemetry | 15:37 | |
openstackgerrit | gordon chung proposed openstack/gnocchi master: redis: use hashes for aggregated storage https://review.openstack.org/447019 | 15:37 |
openstackgerrit | gordon chung proposed openstack/gnocchi master: default to v3 format https://review.openstack.org/447032 | 15:38 |
*** leitan has quit IRC | 15:39 | |
jahsis | When using gnocchi publisher, all pipeline configuration should be in ceilometer-agent-notification node are also on compute nodes? | 15:46 |
jahsis | or* | 15:46 |
*** Jack_Iv has quit IRC | 15:49 | |
gordc | jahsis: i don't know what you just asked. if you're not on master, you can use the same pipeline.yaml file for polling agents and notifications agents | 15:55 |
gordc | in master (and ocata), you should be using polling.yaml in polling agents and pipeline.yaml in notification agents | 15:55 |
jahsis | gordc, ohh | 15:56 |
jahsis | right now I am very confused | 15:56 |
jahsis | Looks like in each release telemetry service architecture changes rapidly :D | 15:56 |
*** nicodemus_ has joined #openstack-telemetry | 15:56 | |
nicodemus_ | hello | 15:57 |
gordc | we're trying :) | 15:57 |
gordc | nicodemus_: hey | 15:57 |
jahsis | gordc, I have one node where are running ceilometer-agent-central and ceilometer-agent-notification services, as I understand from documentation, they should be on separate node | 15:57 |
nicodemus_ | gordc, you know that if I logged here, I'm going to ask you something right? | 15:58 |
gordc | the central agent and notification can be on same node. any 'controller' node. | 15:58 |
jahsis | gordc, also I have ceilometer-agent-compute on each of my compute nodes | 15:58 |
gordc | compute agent needs to be on each compute node | 15:58 |
gordc | right | 15:58 |
jahsis | gordc, yes, I understand, so, on compute nodes this agent are polling agent, right? | 15:59 |
gordc | nicodemus_: you know because i'm here, it doesn't mean i know answer right? :P | 15:59 |
gordc | jahsis: yes, central and compute agents are 'polling' agents | 15:59 |
nicodemus_ | I'm having a little trouble with the disk pollster from ceilometer ocata... specifically here: https://github.com/openstack/ceilometer/blob/stable/ocata/ceilometer/compute/pollsters/disk.py#L133 | 15:59 |
gordc | just diff namespace | 15:59 |
jahsis | gordc, ok, thanks, so this is clear. | 16:00 |
jahsis | Than pipeline file should be only on notification node? | 16:00 |
gordc | jahsis: yes. (as of ocata) | 16:00 |
gordc | there was no polling.yaml before that | 16:00 |
nicodemus_ | I'm trying to create an instance_disk resource in gnocchi, and in netwon I could add a 'name' metadata using 'name: resource_metadata.disk_name', but with Ocata that no longer works :( | 16:01 |
nicodemus_ | the name used to be <resource_id>-vda for example | 16:01 |
nicodemus_ | I've noticed that the base class for this pollster has changed compared to newton | 16:01 |
jahsis | gordc, ok, right now I have some kind of "mixture". I was really tired with gnocchi in mitaka version, so I upgraded telemetry service to Ocata, but left compute nodes to mitaka ... | 16:01 |
gordc | nicodemus_: yeah? you have the patch? which changes it? | 16:02 |
gordc | nicodemus_: havne't looked at pollster stuff in a while | 16:02 |
gordc | in truth, gnocchi (mitaka) probably sucks... newton or ocata might be where you want to start. | 16:03 |
gordc | jahsis: that should be fine | 16:03 |
*** vishwanathj has joined #openstack-telemetry | 16:03 | |
gordc | i don't think anything changed between mitaka->ocata in terms of how agents communicate. | 16:04 |
*** donghao has joined #openstack-telemetry | 16:04 | |
jahsis | gordc, I see.. So basicaly architecture should look like that all polling agent's via messaging service push data to notification agent, and after that notification agent push to gnocchi/panko, right? | 16:04 |
nicodemus_ | gordc, I think this is the one: https://github.com/openstack/ceilometer/commit/a8ea36090a107da1111eb23611f1f79dfc7f289d but I'm not 100% sure if the 'disk_name' additional metadata stopped being there with that commit | 16:07 |
*** jahsis has quit IRC | 16:09 | |
gordc | jahsis: yes | 16:09 |
*** donghao has quit IRC | 16:09 | |
*** nadya has joined #openstack-telemetry | 16:10 | |
gordc | nicodemus_: hmm. seems like it should work. | 16:10 |
gordc | nicodemus_: i guess there's bug it created. | 16:10 |
nicodemus_ | I'll try replacing the disk pollster with the one from newton and see what happens | 16:11 |
gordc | nicodemus_: kk | 16:11 |
gordc | we can always revert that if needed. | 16:11 |
*** leitan has joined #openstack-telemetry | 16:12 | |
*** vishwanathj has quit IRC | 16:17 | |
nicodemus_ | gordc, something else seems to be amiss... even with the mitaka version of that pollster, I cannot get the disk_name from the samples. I'll dig some more. | 16:41 |
*** Gautam has quit IRC | 16:43 | |
*** leitan has quit IRC | 16:43 | |
nicodemus_ | jd_, when you have a moment... We're seeing tooz errors with tooz v1.47 (that in theory were fixed according to https://bugs.launchpad.net/gnocchi/+bug/1557593) | 16:48 |
openstack | Launchpad bug 1557593 in tooz "Unable to extend Redis lock in certain cases" [Medium,Fix released] - Assigned to Julien Danjou (jdanjou) | 16:48 |
*** Jack_Iv has joined #openstack-telemetry | 16:50 | |
nicodemus_ | the error log looks exactly like the one we were getting with tooz v1.34 | 16:51 |
nicodemus_ | http://paste.openstack.org/show/603164/ | 16:51 |
*** Jack_Iv has quit IRC | 16:53 | |
*** leitan has joined #openstack-telemetry | 16:54 | |
jd_ | nicodemus_: how often? | 16:56 |
nicodemus_ | in the last 500 lines of log, it looks like every four minutes | 16:57 |
nicodemus_ | give or take | 16:57 |
nicodemus_ | we're trying now with tooz v1.51 | 16:58 |
jd_ | nicodemus_: ok, I think I've an idea that could improve that | 16:59 |
nicodemus_ | awesome! :) | 17:00 |
openstackgerrit | gordon chung proposed openstack/gnocchi master: check unprocessed measures cleared on delete https://review.openstack.org/447069 | 17:01 |
*** nijaba has quit IRC | 17:05 | |
*** nijaba has joined #openstack-telemetry | 17:06 | |
*** nijaba has quit IRC | 17:06 | |
*** nijaba has joined #openstack-telemetry | 17:06 | |
openstackgerrit | Merged openstack/ceilometer master: Remove second tenacity in requirements https://review.openstack.org/447022 | 17:12 |
jd_ | nicodemus_: https://review.openstack.org/447074 if you wanna try | 17:18 |
*** pcaruana has quit IRC | 17:18 | |
nicodemus_ | jd_, excellent! trying it | 17:19 |
*** nadya has quit IRC | 17:20 | |
*** leitan has quit IRC | 17:23 | |
*** nijaba has quit IRC | 17:27 | |
*** nijaba has joined #openstack-telemetry | 17:33 | |
*** nijaba has quit IRC | 17:33 | |
*** nijaba has joined #openstack-telemetry | 17:33 | |
*** Pneumatic_ has joined #openstack-telemetry | 17:39 | |
*** Jack_Iv has joined #openstack-telemetry | 17:42 | |
*** leitan has joined #openstack-telemetry | 17:42 | |
*** lhx_ has quit IRC | 17:58 | |
*** nadya has joined #openstack-telemetry | 18:12 | |
*** nadya has quit IRC | 18:17 | |
*** nicodemus_ has quit IRC | 18:24 | |
*** nicodemus_ has joined #openstack-telemetry | 18:39 | |
*** tesseract has quit IRC | 18:50 | |
*** donghao has joined #openstack-telemetry | 18:53 | |
*** catintheroof has quit IRC | 18:59 | |
*** donghao has quit IRC | 19:01 | |
*** catintheroof has joined #openstack-telemetry | 19:06 | |
*** amoralej is now known as amoralej|off | 19:09 | |
*** leitan has quit IRC | 19:10 | |
*** leitan has joined #openstack-telemetry | 19:10 | |
*** Jack_Iv has quit IRC | 19:19 | |
*** Jack_Iv has joined #openstack-telemetry | 19:23 | |
nicodemus_ | jd_, so far tooz didn't throw another 'unlocked lock' error | 19:25 |
nicodemus_ | it's been running about 3 hours | 19:25 |
*** shardy has quit IRC | 19:44 | |
*** flwang1 has joined #openstack-telemetry | 19:45 | |
*** Pneumatic_ has quit IRC | 19:57 | |
*** leitan has quit IRC | 20:00 | |
openstackgerrit | gordon chung proposed openstack/gnocchi master: redis: use hashes for aggregated storage https://review.openstack.org/447019 | 20:01 |
*** donghao has joined #openstack-telemetry | 20:02 | |
openstackgerrit | gordon chung proposed openstack/gnocchi master: default to v3 format https://review.openstack.org/447032 | 20:04 |
nicodemus_ | jd_, I guess I spoke too early... the 'unlocked lock' error happened again despite the patch | 20:07 |
*** dave-mccowan has quit IRC | 20:07 | |
*** cdent has quit IRC | 20:09 | |
*** donghao has quit IRC | 20:10 | |
*** dschultz has quit IRC | 20:16 | |
*** sergio has joined #openstack-telemetry | 20:17 | |
*** sergio is now known as Guest76216 | 20:17 | |
*** flwang1 has left #openstack-telemetry | 20:23 | |
nicodemus_ | gordc, just an FYI with the disk_name thing... I finally got it to work mapping 'name' to 'resource_id' in gnocchi_resources.yaml | 20:23 |
gordc | nicodemus_: hm? not sure how that looks. | 20:24 |
gordc | something that needs to be fixed upstream? | 20:24 |
nicodemus_ | for some reason, a sample digested by a collector/dispatcher mitaka get POSTed to gnocchi with the name ad resoure_metadata.disk_name | 20:24 |
nicodemus_ | but the same sample digested by a collector/dispatcher ocata, resoure_metadata.display_name is not there :( | 20:25 |
*** leitan has joined #openstack-telemetry | 20:26 | |
nicodemus_ | I can't make sense of it | 20:26 |
gordc | don't know either. as long as it works for you | 20:27 |
*** Jack_Iv has quit IRC | 20:34 | |
*** dschultz has joined #openstack-telemetry | 20:38 | |
*** Jack_Iv has joined #openstack-telemetry | 20:47 | |
openstackgerrit | gordon chung proposed openstack/gnocchi master: push incoming into different sacks https://review.openstack.org/441389 | 20:50 |
openstackgerrit | gordon chung proposed openstack/gnocchi master: use tooz partitioner https://review.openstack.org/443980 | 20:50 |
*** Guest76216 has quit IRC | 20:54 | |
jd_ | nicodemus_: same backtrace? | 20:57 |
nicodemus_ | jd_, yes... http://paste.openstack.org/show/603189/ | 21:00 |
jd_ | hm | 21:01 |
jd_ | nicodemus_: no other warning? | 21:01 |
jd_ | nicodemus_: like "Heartbeating took too long to execute" ? | 21:02 |
nicodemus_ | no, the only line before the trace reads "Unable to heartbeat lock '<tooz.drivers.redis.RedisLock object at 0x7f207d453b90>'" | 21:02 |
nicodemus_ | it was working fine for a little more than 3 hours, and all of the sudden these traces reappeared | 21:03 |
leitan | Hi guys, jd_ im looking for configurable timeout parameters on the gnocchiclient, i dont clearly see it like for example the timeout attribute when creating a client in python-novaclient, is this configurable ? we are experiencing issues in low responsive environment with this | 21:04 |
leitan | like this for example https://github.com/openstack/python-novaclient/blob/41f66d15aa83004d5b26f7d510b19cd8c4b39e8b/novaclient/v2/client.py#L95 | 21:04 |
jd_ | nicodemus_: hum that's weird, and pretty hard to reproduce | 21:04 |
jd_ | nicodemus_: I'd give a try increasing the lock timeout just in case | 21:04 |
jd_ | nicodemus_: I'll try to take a look at the code to see if I see anything suspicious | 21:04 |
nicodemus_ | jd_, If I can help with testing, let me know | 21:05 |
jd_ | leitan: you can pass requests session parameters in gnocchiclient IIRC | 21:05 |
jd_ | nicodemus_: sure we'll do | 21:05 |
leitan | jd_: great so i should pass timeout value to the session object and should work | 21:08 |
jd_ | leitan: yes | 21:08 |
leitan | jd_: hm, ill try that out, thanks | 21:08 |
*** catintheroof has quit IRC | 21:16 | |
*** dschultz has quit IRC | 21:16 | |
*** dschultz has joined #openstack-telemetry | 21:22 | |
*** leitan has quit IRC | 21:26 | |
*** pradk has quit IRC | 21:26 | |
*** dschultz has quit IRC | 21:27 | |
*** dschultz has joined #openstack-telemetry | 21:28 | |
*** gordc has quit IRC | 21:30 | |
*** dschultz has quit IRC | 21:33 | |
*** Jack_Iv has quit IRC | 21:34 | |
*** dschultz has joined #openstack-telemetry | 21:34 | |
*** Jack_Iv has joined #openstack-telemetry | 21:35 | |
*** dschultz has quit IRC | 21:39 | |
*** Jack_Iv has quit IRC | 21:39 | |
openstackgerrit | gordon chung proposed openstack/gnocchi master: push incoming into different sacks https://review.openstack.org/441389 | 21:45 |
*** dschultz has joined #openstack-telemetry | 21:54 | |
*** leitan has joined #openstack-telemetry | 22:00 | |
*** nicodemus_ has quit IRC | 22:02 | |
*** leitan has quit IRC | 22:13 | |
*** gongysh has joined #openstack-telemetry | 22:40 | |
*** donghao has joined #openstack-telemetry | 23:29 | |
*** donghao has quit IRC | 23:30 | |
*** donghao has joined #openstack-telemetry | 23:44 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!