| *** rbak has joined #openstack-telemetry | 00:00 | |
| *** diogogmt has quit IRC | 00:04 | |
| *** ddieterly has quit IRC | 00:13 | |
| *** rbak has quit IRC | 00:25 | |
| *** tongli has joined #openstack-telemetry | 00:30 | |
| *** Ephur has quit IRC | 00:31 | |
| *** mdih has quit IRC | 00:35 | |
| *** tongli has quit IRC | 00:40 | |
| *** tongli has joined #openstack-telemetry | 00:40 | |
| *** cheneydc has joined #openstack-telemetry | 01:00 | |
| *** tongli has quit IRC | 01:10 | |
| *** ekarlso has quit IRC | 01:10 | |
| *** tongli has joined #openstack-telemetry | 01:11 | |
| *** tongli has quit IRC | 01:11 | |
| *** tongli has joined #openstack-telemetry | 01:11 | |
| *** prashantD has quit IRC | 01:14 | |
| *** nicodemus_ has joined #openstack-telemetry | 01:17 | |
| nicodemus_ | hello | 01:21 |
|---|---|---|
| nicodemus_ | has anybody come across an issue with amqp messages not being processed in time in the metering.sample queue? | 01:22 |
| nicodemus_ | I have two compute nodes with about 50 instances with gnocchi as backend, and the metering.sample messages are building up | 01:23 |
| *** nicodemus_ has quit IRC | 01:33 | |
| *** fabian4 has quit IRC | 01:40 | |
| openstackgerrit | Michael Krotscheck proposed openstack/ceilometer: Added CORS support to Ceilometer https://review.openstack.org/224464 | 01:48 |
| *** fabian4 has joined #openstack-telemetry | 01:52 | |
| *** chlong has joined #openstack-telemetry | 01:53 | |
| *** chlong is now known as chlong-afk | 01:53 | |
| *** thumpba has joined #openstack-telemetry | 01:55 | |
| *** ljxiash has joined #openstack-telemetry | 01:55 | |
| *** thumpba_ has joined #openstack-telemetry | 01:58 | |
| *** thumpba has quit IRC | 01:58 | |
| *** thumpba_ has quit IRC | 02:01 | |
| *** thumpba has joined #openstack-telemetry | 02:01 | |
| *** diogogmt has joined #openstack-telemetry | 02:02 | |
| *** thumpba has quit IRC | 02:03 | |
| *** thorst_ has quit IRC | 02:03 | |
| *** thorst has joined #openstack-telemetry | 02:04 | |
| *** thumpba has joined #openstack-telemetry | 02:07 | |
| *** tongli has quit IRC | 02:09 | |
| *** tongli has joined #openstack-telemetry | 02:10 | |
| *** liamji has joined #openstack-telemetry | 02:11 | |
| *** thorst has quit IRC | 02:14 | |
| *** tongli has quit IRC | 02:15 | |
| *** mdih has joined #openstack-telemetry | 02:16 | |
| *** dave-mccowan has joined #openstack-telemetry | 02:30 | |
| *** thumpba has quit IRC | 02:31 | |
| *** thumpba has joined #openstack-telemetry | 02:31 | |
| *** thumpba has quit IRC | 02:32 | |
| *** thumpba has joined #openstack-telemetry | 02:38 | |
| *** thumpba has quit IRC | 02:40 | |
| *** thumpba has joined #openstack-telemetry | 02:42 | |
| openstackgerrit | Zi Lian Ji proposed openstack/python-gnocchiclient: Supported resources list cli support https://review.openstack.org/263623 | 02:43 |
| *** achatterjee has joined #openstack-telemetry | 02:51 | |
| *** oraclebill has joined #openstack-telemetry | 03:01 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ceilometer: Updated from global requirements https://review.openstack.org/266147 | 03:07 |
| *** thumpba has quit IRC | 03:08 | |
| *** thorst has joined #openstack-telemetry | 03:12 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-ceilometerclient: Updated from global requirements https://review.openstack.org/266162 | 03:12 |
| *** sudorandom has quit IRC | 03:14 | |
| *** thumpba has joined #openstack-telemetry | 03:16 | |
| *** sudorandom has joined #openstack-telemetry | 03:16 | |
| *** thorst has quit IRC | 03:19 | |
| *** ekarlso has joined #openstack-telemetry | 03:19 | |
| *** links has joined #openstack-telemetry | 03:23 | |
| *** jfluhmann has joined #openstack-telemetry | 03:43 | |
| achatterjee | Can someone please review https://review.openstack.org/#/c/260331/ | 03:43 |
| *** thumpba has quit IRC | 03:44 | |
| *** thumpba has joined #openstack-telemetry | 03:45 | |
| *** thumpba has quit IRC | 03:50 | |
| *** ljxiash has quit IRC | 03:56 | |
| *** ljxiash has joined #openstack-telemetry | 03:56 | |
| *** ljxiash has quit IRC | 04:00 | |
| *** nicodemus_ has joined #openstack-telemetry | 04:02 | |
| *** tongli has joined #openstack-telemetry | 04:10 | |
| *** nicodemus_ has quit IRC | 04:15 | |
| *** thorst has joined #openstack-telemetry | 04:17 | |
| *** tongli has quit IRC | 04:23 | |
| *** tongli has joined #openstack-telemetry | 04:24 | |
| *** thorst has quit IRC | 04:25 | |
| *** tongli has quit IRC | 04:26 | |
| *** tongli has joined #openstack-telemetry | 04:26 | |
| *** tongli has quit IRC | 04:30 | |
| *** tongli has joined #openstack-telemetry | 04:31 | |
| *** tongli has quit IRC | 04:35 | |
| *** tongli has joined #openstack-telemetry | 04:36 | |
| *** tongli has quit IRC | 04:40 | |
| *** dave-mccowan has quit IRC | 04:41 | |
| *** sudorandom has quit IRC | 04:45 | |
| *** sudorandom has joined #openstack-telemetry | 04:47 | |
| *** dave-mccowan has joined #openstack-telemetry | 04:47 | |
| *** ljxiash has joined #openstack-telemetry | 04:51 | |
| *** cheneydc has quit IRC | 04:55 | |
| *** thumpba has joined #openstack-telemetry | 04:55 | |
| *** sudorandom has quit IRC | 04:56 | |
| *** yprokule has joined #openstack-telemetry | 04:57 | |
| *** thumpba has quit IRC | 05:01 | |
| *** sudorandom has joined #openstack-telemetry | 05:02 | |
| *** boris-42 has quit IRC | 05:03 | |
| *** oraclebill has quit IRC | 05:17 | |
| *** cheneydc has joined #openstack-telemetry | 05:19 | |
| *** thorst has joined #openstack-telemetry | 05:21 | |
| *** thorst has quit IRC | 05:29 | |
| *** sanjana has quit IRC | 05:32 | |
| *** dave-mccowan has quit IRC | 05:36 | |
| *** jwcroppe has quit IRC | 05:40 | |
| *** _nadya_ has joined #openstack-telemetry | 05:48 | |
| *** fabian4 has quit IRC | 05:51 | |
| *** prashantD has joined #openstack-telemetry | 05:56 | |
| *** fabian4 has joined #openstack-telemetry | 06:04 | |
| *** _nadya_ has quit IRC | 06:14 | |
| *** fabian4 has quit IRC | 06:21 | |
| *** thorst has joined #openstack-telemetry | 06:26 | |
| *** fabian4 has joined #openstack-telemetry | 06:33 | |
| *** sanjana has joined #openstack-telemetry | 06:34 | |
| *** thorst has quit IRC | 06:34 | |
| *** prashantD has quit IRC | 06:34 | |
| *** mdih has quit IRC | 06:41 | |
| openstackgerrit | Zi Lian Ji proposed openstack/gnocchi: Replace the redundant code with the util function https://review.openstack.org/266725 | 06:49 |
| openstackgerrit | Zi Lian Ji proposed openstack/gnocchi: Replace the redundant code with the util function https://review.openstack.org/266725 | 06:55 |
| *** fabian4 has left #openstack-telemetry | 07:05 | |
| openstackgerrit | Ananya Chatterjee proposed openstack/ceilometer: Set None explicitly to filter options. https://review.openstack.org/260331 | 07:28 |
| *** thorst has joined #openstack-telemetry | 07:31 | |
| *** thorst has quit IRC | 07:39 | |
| *** _nadya_ has joined #openstack-telemetry | 07:39 | |
| *** safchain has joined #openstack-telemetry | 07:40 | |
| *** chenli has joined #openstack-telemetry | 07:58 | |
| *** ildikov has quit IRC | 08:00 | |
| *** shardy has joined #openstack-telemetry | 08:24 | |
| openstackgerrit | Zi Lian Ji proposed openstack/python-gnocchiclient: Supported resources list cli support https://review.openstack.org/263623 | 08:26 |
| *** ildikov has joined #openstack-telemetry | 08:29 | |
| *** liusheng has quit IRC | 08:35 | |
| *** liusheng has joined #openstack-telemetry | 08:35 | |
| *** thorst has joined #openstack-telemetry | 08:37 | |
| *** shardy has quit IRC | 08:40 | |
| *** mattyw has joined #openstack-telemetry | 08:41 | |
| openstackgerrit | Merged openstack/ceilometer: Add OSprofiler-specific events definitions https://review.openstack.org/245136 | 08:41 |
| *** shardy has joined #openstack-telemetry | 08:42 | |
| *** thorst has quit IRC | 08:43 | |
| openstackgerrit | Zi Lian Ji proposed openstack/python-gnocchiclient: Supported resources list cli support https://review.openstack.org/263623 | 08:47 |
| openstackgerrit | Xia Linjuan proposed openstack/python-ceilometerclient: change the dict output format to make consistency under py27 and py34 https://review.openstack.org/266187 | 09:00 |
| *** yassine has joined #openstack-telemetry | 09:23 | |
| *** chenli has quit IRC | 09:28 | |
| openstackgerrit | Xia Linjuan proposed openstack/python-ceilometerclient: change the dict output format to make consistency under py27 and py34 https://review.openstack.org/266187 | 09:32 |
| *** thorst has joined #openstack-telemetry | 09:41 | |
| *** liamji has quit IRC | 09:44 | |
| *** thorst has quit IRC | 09:48 | |
| *** rmart04 has joined #openstack-telemetry | 10:02 | |
| *** cheneydc has quit IRC | 10:02 | |
| rmart04 | Hi Guys, I asked this in ‘openstack’ but didnt get any hits unfortunately, I appreciate this is dev chat but I was hoping someone might have an idea, We are having a strange issue where we do not appear to gain any compute.instance.update event samples in ceilometer. I can confirm we have notify_on_state_change = vm_and_task_state configured on all of our nova compute (kvm) services. Ceilometer event-type-list does not include this event t | 10:04 |
| rmart04 | either. This suggests that it has never witnessed this type of event. Any Ideas? | 10:04 |
| *** cdent has joined #openstack-telemetry | 10:15 | |
| *** ljxiash has quit IRC | 10:23 | |
| *** ljxiash has joined #openstack-telemetry | 10:23 | |
| *** ljxiash has quit IRC | 10:28 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Bypass the auth when listing Gnocchi versions https://review.openstack.org/266301 | 10:33 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: Bypass the auth when listing Gnocchi versions https://review.openstack.org/266301 | 10:35 |
| *** thorst has joined #openstack-telemetry | 10:46 | |
| *** ljxiash has joined #openstack-telemetry | 10:48 | |
| *** zqfan has joined #openstack-telemetry | 10:48 | |
| *** thorst has quit IRC | 10:53 | |
| *** alextricity25 has quit IRC | 11:09 | |
| *** alextricity25 has joined #openstack-telemetry | 11:11 | |
| *** Aminasyan has joined #openstack-telemetry | 11:17 | |
| cdent | sileht: you see the jsonpath-rw-ext bug I posted last night? any ideas? | 11:18 |
| sileht | cdent, nop | 11:32 |
| *** Aminasyan has quit IRC | 11:32 | |
| *** Aminasyan has joined #openstack-telemetry | 11:32 | |
| cdent | sileht: https://github.com/sileht/python-jsonpath-rw-ext/issues/6 | 11:32 |
| sileht | cdent, I will take a look after my lunch | 11:34 |
| cdent | cool, thanks | 11:34 |
| *** sanjana has quit IRC | 11:40 | |
| *** flwang1 has quit IRC | 11:48 | |
| *** sergio_nubeliu has joined #openstack-telemetry | 11:48 | |
| *** flwang1 has joined #openstack-telemetry | 11:49 | |
| *** thorst has joined #openstack-telemetry | 11:51 | |
| *** thorst has quit IRC | 11:59 | |
| *** rbak has joined #openstack-telemetry | 12:07 | |
| *** Aminasyan has quit IRC | 12:10 | |
| *** rbak has quit IRC | 12:16 | |
| *** gordc has joined #openstack-telemetry | 12:30 | |
| *** julim has joined #openstack-telemetry | 12:33 | |
| _nadya_ | sileht: ping | 12:43 |
| *** thorst has joined #openstack-telemetry | 12:43 | |
| *** ityaptin has quit IRC | 12:46 | |
| *** ityaptin has joined #openstack-telemetry | 12:46 | |
| *** links has quit IRC | 13:04 | |
| *** rcernin has joined #openstack-telemetry | 13:09 | |
| *** nicodemus_ has joined #openstack-telemetry | 13:15 | |
| nicodemus_ | Hello gordc, do you have a moment? | 13:16 |
| nicodemus_ | I'm having an issue with the metering.sample queue in rabbitmq, the number of queued messages is growing | 13:17 |
| nicodemus_ | As if ceilometer-collector couldn't keep up with the amount of samples from the agents | 13:19 |
| *** pradk has joined #openstack-telemetry | 13:19 | |
| openstackgerrit | Xia Linjuan proposed openstack/python-ceilometerclient: change the dict output format to make consistency under py27 and py34 https://review.openstack.org/266187 | 13:19 |
| *** pradk has quit IRC | 13:23 | |
| gordc | nicodemus_: is collector consuming messages at all? | 13:24 |
| nicodemus_ | It is, but it seems as if more messages are arriving than those it can consume | 13:25 |
| nicodemus_ | I've increased the number of workers up to 32 | 13:26 |
| nicodemus_ | And this morning there were 0 consumers in the queue | 13:26 |
| gordc | ... you have 32 consumers of metering.sample queue? | 13:26 |
| nicodemus_ | As if they were disconnecting and not coming back | 13:27 |
| gordc | yeah, i was giong to say 32 sounds excessive... | 13:27 |
| gordc | try lowering it... how many messages are being sent per second? | 13:28 |
| nicodemus_ | When the agents run (every 60 seconds), Around 300-350 messages are sent | 13:29 |
| nicodemus_ | There are two computes with about 50 instances | 13:30 |
| gordc | ... taht shouldn't be that bad. i believe each worker can handle ~400 messages/s (i haven't tried this in a while so i might be totally off) | 13:31 |
| gordc | i would start off with 2 workers and see if they are active. | 13:31 |
| nicodemus_ | Would you rather that I open a bug report with my findings and the logs? | 13:31 |
| gordc | i've never tried 32 workers | 13:31 |
| gordc | nicodemus_: sure. if you have additional logs. | 13:32 |
| gordc | if you have 32 consumers i'd split that across multiple discrete processes rather than have a single process manage 32 workers. | 13:32 |
| gordc | that way, one process hanging doesn't take down 32 consumers. | 13:33 |
| openstackgerrit | Nadya Shakhat proposed openstack/ceilometer-specs: A proposal about distributed cache for notification agents https://review.openstack.org/250778 | 13:33 |
| nicodemus_ | I configured 32 due to the number of queued messages (55.000) | 13:33 |
| _nadya_ | nicodemus_: what db do you use? | 13:33 |
| nicodemus_ | Three nodes of Galera mysql | 13:34 |
| _nadya_ | check that it feels ok. do you use separate nodes? | 13:35 |
| gordc | nicodemus_: it depends on number of connections you configure for mysql but i think if you take all default values, i got around 40 writes/s per collector. | 13:36 |
| nicodemus_ | Yes, three separate hw nodes | 13:36 |
| gordc | nicodemus_: try 2 to 5 collectors first i think. and see if the connection to queue is maintained. | 13:37 |
| nicodemus_ | Those nodes dont have problems with cpu/memory | 13:37 |
| openstackgerrit | Merged openstack/ceilometer: Set None explicitly to filter options. https://review.openstack.org/260331 | 13:37 |
| nicodemus_ | Ok, will try that and submit the report in the meantime | 13:39 |
| nicodemus_ | Thanks! | 13:39 |
| gordc | sileht: new ceilometerclient is broken for heat. | 13:44 |
| sileht | gordc, I have send a patch ! | 13:44 |
| gordc | oh... i will stop debugging then :) | 13:44 |
| sileht | https://review.openstack.org/#/c/266825/ | 13:44 |
| gordc | sileht: hmm.. i think there's also another problem | 13:45 |
| gordc | https://github.com/openstack/python-ceilometerclient/blob/master/ceilometerclient/v2/client.py#L100 | 13:45 |
| gordc | that doesn't exist | 13:45 |
| sileht | gordc, we don't care | 13:45 |
| gordc | so i don't think client redirect works. | 13:45 |
| gordc | http://logs.openstack.org/41/266541/1/check/gate-ceilometer-dsvm-integration/1986475/logs/screen-h-eng.txt.gz#_2016-01-12_22_49_43_027 | 13:46 |
| nicodemus_ | _nadya_ when you ask about db, you mean ceilometer backend? If that was it, I'm using gnocchi | 13:46 |
| sileht | gordc,indead some fields are missing | 13:47 |
| nicodemus_ | two instances loadbalanced with haproxy | 13:47 |
| _nadya_ | nicodemus_: yep, I meant Ceilo backend. Gnocchi with Ceph? | 13:47 |
| gordc | nicodemus_: do you have ceilometer gnocchi dispatcher cache turned on? | 13:47 |
| nicodemus_ | Yes to ceph | 13:47 |
| nicodemus_ | No to cache | 13:48 |
| gordc | sileht: i think those fields are only checked if we don't have a valid endpoint+token | 13:48 |
| _nadya_ | sileht: Mehdi, hi! Do you know who is Kafka expert in oslo? | 13:48 |
| sileht | _nadya_, no idea | 13:49 |
| gordc | right now heat only gives token | 13:49 |
| sileht | gordc, I have removed this logic | 13:49 |
| _nadya_ | sileht: ok :) | 13:49 |
| sileht | gordc, the aodh should recreate another token | 13:49 |
| sileht | gordc, keystoneauth/keystoneclient doesn't support to be deepcopied | 13:50 |
| gordc | sileht: https://github.com/openstack/python-ceilometerclient/blob/master/ceilometerclient/client.py#L216 | 13:50 |
| *** cheneydc has joined #openstack-telemetry | 13:51 | |
| gordc | i think that's the line that we fail on... possibly | 13:51 |
| *** oraclebill has joined #openstack-telemetry | 13:51 | |
| gordc | nicodemus_: i think you need cache. ceilometer+gnocchi has a lot of writes without cache. | 13:51 |
| gordc | nicodemus_: also, how did you configure apache for gnocchi? ie how many processes/threads? | 13:52 |
| nicodemus_ | Ok, I'll look into that | 13:52 |
| nicodemus_ | Gnocchi has one API process and four metricd workers | 13:53 |
| gordc | nicodemus_: you don't have it configured with httpd/mod_wsgi? | 13:54 |
| nicodemus_ | No | 13:55 |
| nicodemus_ | Would it be better? | 13:56 |
| *** ddieterly has joined #openstack-telemetry | 13:56 | |
| gordc | nicodemus_: oh. yeah, that's the reason then. if writing from ceilometer->gnocchi, you need gnocchi configured with apache | 13:56 |
| gordc | nicodemus_: i think a single gnocchi api process can only make a handful of calls a second... and since ceilometer (before yesterday) writes a datapoint per call... | 13:57 |
| nicodemus_ | I see... Why is that? | 13:57 |
| gordc | you are only writing at most 5 messages per second. | 13:58 |
| gordc | why ceilometer writes one datapoint at a time? | 13:58 |
| nicodemus_ | Oh | 13:58 |
| gordc | nicodemus_: http://docs.openstack.org/developer/gnocchi/running.html#how-to-scale-out-the-gnocchi-http-rest-api-tier | 13:58 |
| nicodemus_ | No, I meant the reason for apache | 13:58 |
| nicodemus_ | It's clear now. | 13:59 |
| gordc | nicodemus_: oh. gnocchi on it's own can receive multiple datapoints per api call... it's just ceilometer isn't configured to do that (until yesterday) | 13:59 |
| gordc | but yeah. there's only so many calls you can make to a single api process per second... if you put it behind apache, you'll be able to serve a lot more requests. | 14:00 |
| nicodemus_ | I wasn't aware of that! I'll give it a try now | 14:02 |
| nicodemus_ | Thanks a lot! | 14:02 |
| gordc | nicodemus_: cool cool. you can use this if you need a template https://github.com/openstack/gnocchi/blob/master/devstack/apache-gnocchi.template | 14:02 |
| *** rmart04_ has joined #openstack-telemetry | 14:03 | |
| *** rmart04 has quit IRC | 14:04 | |
| *** rmart04_ is now known as rmart04 | 14:04 | |
| *** links has joined #openstack-telemetry | 14:04 | |
| gordc | sileht: i was thinking maybe we should try to get aodh endpoint if aodh_endpoint isn't provided (in _get_alarm_client) | 14:05 |
| *** ddieterly has quit IRC | 14:05 | |
| gordc | no idea if that's right | 14:05 |
| sileht | gordc, _do_authenticate already do that | 14:06 |
| *** nicodemus_ has quit IRC | 14:06 | |
| *** nicodemus_ has joined #openstack-telemetry | 14:07 | |
| gordc | sileht: hmm... does it fail because we remove the endpoint here? https://github.com/openstack/python-ceilometerclient/blob/master/ceilometerclient/v2/client.py#L109 | 14:07 |
| gordc | so it doesn't know aodh or ceilometer endpoints? | 14:07 |
| sileht | gordc, if the endpoint is None, it should looks to the keystone endpoint to find an endpoint with service_type="alarming" | 14:08 |
| *** links has quit IRC | 14:09 | |
| gordc | ah i see... | 14:09 |
| gordc | i have no idea then.lol | 14:09 |
| *** links has joined #openstack-telemetry | 14:09 | |
| gordc | https://github.com/openstack/python-ceilometerclient/blob/master/ceilometerclient/openstack/common/apiclient/auth.py#L203 it seems maybe it fails before it can even try to find aodh endpoint | 14:11 |
| *** thumpba has joined #openstack-telemetry | 14:15 | |
| *** ildikov has quit IRC | 14:15 | |
| *** liamji has joined #openstack-telemetry | 14:16 | |
| *** ljxiash_ has joined #openstack-telemetry | 14:19 | |
| liamji | Hello Gordon | 14:19 |
| *** ljxiash has quit IRC | 14:20 | |
| *** links has quit IRC | 14:25 | |
| *** links has joined #openstack-telemetry | 14:25 | |
| *** diogogmt has quit IRC | 14:26 | |
| *** links has quit IRC | 14:29 | |
| *** links has joined #openstack-telemetry | 14:30 | |
| *** thumpba has quit IRC | 14:33 | |
| sileht | gordc, the issue occurs when you put a token, then discovery of aodh endpoint is skipped :( | 14:33 |
| gordc | liamji: hi (i assume you mean me) | 14:34 |
| gordc | sileht: :( where's that? | 14:35 |
| sileht | gordc, when you have a token you don't need the endpoint, if you have the service_type, but here we enforce to have an endpoint: https://github.com/openstack/python-ceilometerclient/blob/master/ceilometerclient/client.py#L216 | 14:36 |
| gordc | sileht: ah. got it. | 14:39 |
| liamji | gordc: yes. thanks for responding me. LBaaS v1 is deprecated in the M release. For v2 is different with v1, I think the ceilometer will lost the function to collect the LBaaS data if v2 is not supported | 14:39 |
| *** thumpba has joined #openstack-telemetry | 14:39 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/python-ceilometerclient: Only token or creds are required. https://review.openstack.org/266949 | 14:39 |
| sileht | gordc, that should fix the issue https://review.openstack.org/266949, I will test it with heat, nwo | 14:39 |
| sileht | now | 14:39 |
| liamji | gordc: some users are using this function and report the bug from Oct last year | 14:40 |
| gordc | sileht: looks good. | 14:40 |
| liamji | http://developer.openstack.org/api-ref-networking-v2-ext.html | 14:40 |
| gordc | liamji: i see. thanks for bringing this up. | 14:41 |
| gordc | are you proposing a change? | 14:41 |
| *** ddieterly has joined #openstack-telemetry | 14:41 | |
| liamji | yes. I have submit a proposal https://review.openstack.org/#/c/244139/ | 14:41 |
| *** cdent has quit IRC | 14:42 | |
| liamji | gordc: I think your suggesions will be uesful. thanks | 14:43 |
| gordc | liamji: have you started coding? | 14:43 |
| liamji | yes, I have built the environment and started to write some code. | 14:44 |
| *** ddieterly has quit IRC | 14:46 | |
| gordc | liamji: cool cool | 14:46 |
| gordc | liamji: you are keeping the same meters ? are you adding or removing any? | 14:47 |
| openstackgerrit | Zi Lian Ji proposed openstack/gnocchi: Replace the redundant code with the utility function https://review.openstack.org/266725 | 14:48 |
| *** pradk has joined #openstack-telemetry | 14:49 | |
| liamji | gordc: I am not planning to add more meters supported by v2. but I am afraid that some existing meters will be not supported for they are not supported by v2 | 14:55 |
| gordc | liamji: kk. can you add a list of differences to your spec. | 14:56 |
| liamji | gordc: but it needs to support more meters. I am glad to do it. | 14:56 |
| gordc | liamji: once you do that, i'll fix some english and we can merge it | 14:56 |
| gordc | liamji: if you have new meters, feel free to add. not really a requirement. | 14:57 |
| *** davidlenwell has quit IRC | 14:57 | |
| liamji | okey, I will do it. After it's done, I will send it for review. thanks | 14:58 |
| gordc | liamji: sounds good. | 14:59 |
| *** davidlenwell has joined #openstack-telemetry | 14:59 | |
| *** ddieterly has joined #openstack-telemetry | 14:59 | |
| liamji | gordc: thanks for your suggestions. :) | 14:59 |
| *** links has quit IRC | 15:07 | |
| *** dave-mccowan has joined #openstack-telemetry | 15:09 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/python-ceilometerclient: Only token or creds are required. https://review.openstack.org/266949 | 15:11 |
| sileht | gordc, ^ that one works with heat | 15:11 |
| *** dave-mccowan has quit IRC | 15:13 | |
| gordc | sileht: cool cool. want to link it to this bug just for reference. https://bugs.launchpad.net/python-ceilometerclient/+bug/1533712 | 15:14 |
| openstack | Launchpad bug 1533712 in python-ceilometerclient "ceilometerclient cannot resolve aodh redirect" [High,Triaged] - Assigned to Mehdi Abaakouk (sileht) | 15:14 |
| *** cheneydc has quit IRC | 15:15 | |
| *** cdent has joined #openstack-telemetry | 15:17 | |
| *** boris-42 has joined #openstack-telemetry | 15:20 | |
| *** liusheng has quit IRC | 15:28 | |
| *** liamji has quit IRC | 15:31 | |
| pradk | hey gordc, i have the gnocchi alarm commands mostly in place | 15:31 |
| pradk | gordc, i'm trying to see if i can add test scenarios for these commands but since the api call in turn calls gnocchi it fails with auth/endpoint errors | 15:31 |
| pradk | gordc, i dint see tests for gnocchi commands in ceilo client either..? looks like this will be tricky | 15:33 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/python-ceilometerclient: Only token or creds are required. https://review.openstack.org/266949 | 15:35 |
| ljxiash_ | pradk: Hi I already add some, https://review.openstack.org/#/c/250131/ | 15:36 |
| pradk | ljxiash_, ah dint know, i'll check it out, thx | 15:37 |
| gordc | pradk: ? why would it need gnocchi? | 15:41 |
| gordc | pradk: the tests don't evaluate anything. they just create the definition | 15:42 |
| pradk | gordc, yea thoughts what i thought.. for some reason i see it hitting this line and failing when accessing validating the aggregation methods | 15:45 |
| pradk | gordc, hence wanted to understand how you had the base class configured.. i see it calling the self.app.client with params and then i see an auth failure getting keystone client in aodh | 15:46 |
| pradk | and throws a 500 | 15:46 |
| gordc | what line? | 15:46 |
| pradk | gordc, https://github.com/openstack/aodh/blob/master/aodh/api/controllers/v2/alarm_rules/gnocchi.py#L63 | 15:47 |
| pradk | gordc, i assumed we would have mocked this part somewhere.. but apparently not | 15:47 |
| gordc | pradk: word... yeah i didn't do anything related to gnocchi so it's not mocked anywhere. | 15:49 |
| pradk | gordc, where is the event and threshold stuff mocked? | 15:49 |
| gordc | it's not? | 15:49 |
| *** mragupat has joined #openstack-telemetry | 15:50 | |
| gordc | it doesn't need since aodh only queries ceilometer when evaluating alarm. | 15:50 |
| *** jfluhmann has quit IRC | 15:50 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ceilometer: Updated from global requirements https://review.openstack.org/266147 | 15:50 |
| *** alextricity25 has quit IRC | 15:51 | |
| *** alextricity25 has joined #openstack-telemetry | 15:52 | |
| *** alextricity25 has quit IRC | 15:54 | |
| *** alextricity25 has joined #openstack-telemetry | 15:54 | |
| pradk | gordc, k then shouldnt that be the case for gnocchi too? | 15:54 |
| pradk | gordc, only hit gnocchi when evaluating vs at creation time? jus trying to understand the flow here' | 15:55 |
| *** diogogmt has joined #openstack-telemetry | 15:55 | |
| gordc | pradk: um... possibly? are you suggesting to change aodh to not validate? | 16:02 |
| gordc | _nadya_: added a comment to your docs. | 16:08 |
| *** ddaskal has joined #openstack-telemetry | 16:08 | |
| *** rmart04 has quit IRC | 16:10 | |
| *** ljxiash_ has quit IRC | 16:11 | |
| *** ljxiash has joined #openstack-telemetry | 16:13 | |
| pradk | gordc, not necessarily change in aodh (although it would make it a bit more simpler, if it isnt talking to another service just to validate aggr method) | 16:14 |
| pradk | gordc, i'll try to mock the server calls out and see where i get with it | 16:15 |
| *** mattyw has quit IRC | 16:24 | |
| *** ildikov has joined #openstack-telemetry | 16:28 | |
| *** oheydorn has joined #openstack-telemetry | 16:29 | |
| *** oheydorn has quit IRC | 16:45 | |
| *** rcernin has quit IRC | 16:51 | |
| *** lexholden has joined #openstack-telemetry | 16:53 | |
| *** prashantD has joined #openstack-telemetry | 16:54 | |
| *** jfluhmann has joined #openstack-telemetry | 16:57 | |
| *** lexholden has quit IRC | 17:01 | |
| *** zqfan has quit IRC | 17:01 | |
| *** lexholden has joined #openstack-telemetry | 17:05 | |
| *** lexholden has quit IRC | 17:06 | |
| *** mattyw has joined #openstack-telemetry | 17:06 | |
| *** mragupat has quit IRC | 17:20 | |
| *** mragupat has joined #openstack-telemetry | 17:20 | |
| *** thumpba has quit IRC | 17:22 | |
| *** thumpba has joined #openstack-telemetry | 17:22 | |
| *** mragupat_ has joined #openstack-telemetry | 17:23 | |
| gordc | harlowja: jd__: were there any notable changes to tooz this cycle? | 17:25 |
| gordc | specificall redis/memcache drivers? | 17:25 |
| *** mragupat has quit IRC | 17:26 | |
| *** cdent has quit IRC | 17:27 | |
| jd__ | gordc: not from the top of my head | 17:28 |
| *** _nadya_ has quit IRC | 17:28 | |
| jd__ | a few improvements and bug removed, that's all | 17:28 |
| gordc | jd__: hmm... for some reason i can't consistently get back group members when i enable coordination in ceilometer. | 17:29 |
| gordc | i'll debug | 17:29 |
| jd__ | in aodh? | 17:29 |
| gordc | ceilometer. | 17:30 |
| gordc | ie polling and notification agents | 17:30 |
| krotscheck | Urgh, why is devstack against ceilometer so volatile? | 17:34 |
| jd__ | maybe try with aodh then to see if it's a bug in tooz or ceilo? just an idea | 17:34 |
| jd__ | krotscheck: define volatile? | 17:34 |
| krotscheck | jd__: +2A, makes it through check, fails in gate. I rebase, fails in check. Recheck, it passes. | 17:35 |
| krotscheck | Lather, rinse, repeat. | 17:35 |
| * krotscheck is just ranting | 17:35 | |
| jd__ | I feel you | 17:36 |
| *** prashantD has quit IRC | 17:36 | |
| jd__ | I hate this test | 17:36 |
| gordc | krotscheck: patch against stable? | 17:36 |
| jd__ | there are too many projects involved, and some of these projects like Heat don't even run it | 17:36 |
| jd__ | so they can break us without noticing | 17:36 |
| krotscheck | gordc: Nope. https://review.openstack.org/#/c/224464/ | 17:37 |
| *** prashantD has joined #openstack-telemetry | 17:37 | |
| gordc | oh... yeah i have no idea. | 17:37 |
| * krotscheck grumps some more. | 17:37 | |
| gordc | i've actually never understood why unrelated tests get run for all projects | 17:38 |
| krotscheck | gordc: Maybe ask on infra? ¯\_(ツ)_/¯ | 17:38 |
| gordc | krotscheck: i leave that to others with better connections. | 17:39 |
| gordc | if you're not part of the crew, questions usually fall on deaf ears. -- quote from me | 17:39 |
| krotscheck | gordc: I would love to know how many patches fail in devstack, only to pass with a recheck. | 17:40 |
| krotscheck | And how many hours of waiting that involves. | 17:40 |
| gordc | :) i think it will depress you more. just live on false hope. | 17:40 |
| krotscheck | Well, I have a talk proposal for the summit that's going to address those questions. | 17:41 |
| gordc | false hope? :) | 17:41 |
| * krotscheck is collecting a list of thigns that slow down openstack. | 17:41 | |
| gordc | hahah so it's a multi-day talk? got it. | 17:42 |
| krotscheck | gordc: Well, I was thinking a summit parallel with the Ops track. | 17:42 |
| gordc | :) i'll attend. | 17:43 |
| *** rmart04 has joined #openstack-telemetry | 17:43 | |
| openstackgerrit | Rohit Jaiswal proposed openstack/ceilometer: Enhances get_meters to return unique meters https://review.openstack.org/259626 | 17:49 |
| *** shardy has quit IRC | 17:55 | |
| *** rmart04 has left #openstack-telemetry | 17:56 | |
| *** cdent has joined #openstack-telemetry | 17:58 | |
| nicodemus_ | If I request status to gnocchi (via gnocchi client and/or curl to /v1/status) and there's no response/output, would it be correct to assume that gnocchi is way behind measures processing? | 17:59 |
| *** mattyw has quit IRC | 18:00 | |
| nicodemus_ | (taking into account that gnocchi-metricd child processes are segfaulting every now and then)? | 18:01 |
| *** boris-42 has quit IRC | 18:08 | |
| *** larainema has quit IRC | 18:08 | |
| *** elemoine_ has quit IRC | 18:08 | |
| *** patrickeast has quit IRC | 18:08 | |
| *** ramishra_ has quit IRC | 18:08 | |
| *** tpeoples has quit IRC | 18:08 | |
| *** serverascode has quit IRC | 18:08 | |
| *** serverascode has joined #openstack-telemetry | 18:14 | |
| *** larainema has joined #openstack-telemetry | 18:18 | |
| *** ramishra_ has joined #openstack-telemetry | 18:18 | |
| openstackgerrit | Ilya Tyaptin proposed openstack/ceilometer: [WIP] [PoC] Caching metadata of compute pollster https://review.openstack.org/267078 | 18:19 |
| *** yprokule has quit IRC | 18:21 | |
| *** patrickeast has joined #openstack-telemetry | 18:23 | |
| *** tpeoples has joined #openstack-telemetry | 18:24 | |
| *** _nadya_ has joined #openstack-telemetry | 18:29 | |
| *** ddaskal has quit IRC | 18:30 | |
| *** _nadya_ has quit IRC | 18:30 | |
| *** _nadya_ has joined #openstack-telemetry | 18:30 | |
| *** mgagne has quit IRC | 18:30 | |
| *** ddaskal has joined #openstack-telemetry | 18:30 | |
| *** elemoine_ has joined #openstack-telemetry | 18:30 | |
| *** rcernin has joined #openstack-telemetry | 18:31 | |
| gordc | nicodemus_: do you still have the one api process? it's entirely possible it's just getting blocked by competing api requests from ceilometer | 18:31 |
| *** Guest82803 has joined #openstack-telemetry | 18:31 | |
| gordc | segfaults are concerning though. be good to see a bug on that. | 18:31 |
| *** Guest82803 is now known as mgagne | 18:32 | |
| *** mgagne has quit IRC | 18:33 | |
| *** mgagne has joined #openstack-telemetry | 18:33 | |
| nicodemus_ | I've successfully put it behind apache with 3 workers | 18:33 |
| nicodemus_ | the API now is running smoothly | 18:33 |
| *** ddaskal has quit IRC | 18:34 | |
| nicodemus_ | ok, I'll submit a bug report on metricd segfault | 18:36 |
| gordc | nicodemus_: cool cool. thanks | 18:38 |
| *** lexholden has joined #openstack-telemetry | 18:42 | |
| *** ig0r_ has joined #openstack-telemetry | 18:43 | |
| *** _nadya_ has quit IRC | 18:46 | |
| *** lexholde_ has joined #openstack-telemetry | 18:47 | |
| nicodemus_ | gordc, I've submitted the bug report (#1533793) | 18:48 |
| nicodemus_ | please let me know if I can provide any further information | 18:48 |
| gordc | nicodemus_: sure. i'm sure one of devs will comment soon. (i'm looking at another bug right now so probably won't have time today/tomorrow) | 18:50 |
| *** lexholden has quit IRC | 18:50 | |
| nicodemus_ | great, many thanks for your time :) | 18:51 |
| *** lexholden has joined #openstack-telemetry | 18:59 | |
| openstackgerrit | akanksha proposed openstack/python-ceilometerclient: Fix to disable meter-links from CLI https://review.openstack.org/267102 | 19:00 |
| *** lexholde_ has quit IRC | 19:03 | |
| *** lexholde_ has joined #openstack-telemetry | 19:04 | |
| cdent | gordc: WTF on the https://bugs.launchpad.net/ceilometer/+bug/1533787 | 19:04 |
| openstack | Launchpad bug 1533787 in Ceilometer "workload_partitioning inconsistently reports group members" [Undecided,New] | 19:04 |
| cdent | I hope that's new and not just been that way since the beginning? | 19:04 |
| gordc | cdent: first i've seen it. | 19:05 |
| gordc | cdent: i'm trying with aodh | 19:05 |
| gordc | cdent: right now it seems like tooz is acting wonky | 19:06 |
| *** lexholden has quit IRC | 19:07 | |
| *** lexhold__ has joined #openstack-telemetry | 19:07 | |
| *** lexholde_ has quit IRC | 19:10 | |
| *** lexholden has joined #openstack-telemetry | 19:14 | |
| *** lexhold__ has quit IRC | 19:16 | |
| *** lexholden has quit IRC | 19:17 | |
| *** ddaskal has joined #openstack-telemetry | 19:24 | |
| *** _nadya_ has joined #openstack-telemetry | 19:28 | |
| *** pauloewerton has joined #openstack-telemetry | 19:34 | |
| *** boris-42 has joined #openstack-telemetry | 19:40 | |
| openstackgerrit | George Peristerakis proposed openstack/ceilometer: Load a directory of YAML event config files https://review.openstack.org/247177 | 19:51 |
| *** mattyw has joined #openstack-telemetry | 19:54 | |
| *** davidlenwell has quit IRC | 19:57 | |
| *** cdent has quit IRC | 19:57 | |
| *** davidlenwell has joined #openstack-telemetry | 19:58 | |
| *** ddieterly has quit IRC | 20:09 | |
| oraclebill | looking in my logs I’m seeing messages that look like apache server logs e.g. - ‘Jan 13 20:10:22 host123 ceilometer-api[24819]: 10.146.129.15 - - [13/Jan/2016 20:10:22] "GET /v2/alarms?q.field=enabled&q.op=&q.type=&q.value=True HTTP/1.1" 200 2’ | 20:14 |
| oraclebill | these messages only appear in /var/log/messages - not the ceilometer/api.log file.. | 20:15 |
| oraclebill | any idea where they might be generated from? | 20:15 |
| *** terriyu has joined #openstack-telemetry | 20:17 | |
| gordc | your ceilometer logs are generating in /var/log/messages? | 20:18 |
| *** rcernin has quit IRC | 20:20 | |
| openstackgerrit | Pradeep Kilambi proposed openstack/aodh: Zaqar notifier for alarms https://review.openstack.org/264309 | 20:22 |
| gordc | oraclebill: i would check where you configured your apache logs to write to. see: /etc/httpd/conf.d/ceilometer.conf | 20:23 |
| *** _nadya_ has quit IRC | 20:24 | |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: indexer: remove with_resource argument in get_metrics() https://review.openstack.org/267148 | 20:27 |
| *** ddieterly has joined #openstack-telemetry | 20:35 | |
| openstackgerrit | Merged openstack/python-ceilometerclient: Only token or creds are required. https://review.openstack.org/266949 | 20:40 |
| *** ddieterly has quit IRC | 20:41 | |
| *** ddaskal has quit IRC | 20:48 | |
| *** ddaskal has joined #openstack-telemetry | 20:49 | |
| *** rcernin has joined #openstack-telemetry | 20:50 | |
| *** ddaskal has quit IRC | 20:53 | |
| *** belmoreira has joined #openstack-telemetry | 20:56 | |
| *** ekarlso has quit IRC | 21:01 | |
| *** ekarlso has joined #openstack-telemetry | 21:01 | |
| *** rcernin has quit IRC | 21:04 | |
| *** lexholden has joined #openstack-telemetry | 21:04 | |
| *** lexholde_ has joined #openstack-telemetry | 21:05 | |
| *** ddieterly has joined #openstack-telemetry | 21:08 | |
| *** lexholden has quit IRC | 21:09 | |
| oraclebill | gordc: only some of them, it seems like some of the ceilometer components are logging to stderr, which is directed to /var/log/messages | 21:09 |
| gordc | oraclebill: i'm not sure. could be a packaging thing. i'd probably include distro, and logs if you want to open bug | 21:11 |
| oraclebill | gordc: httpd isn’t running locally | 21:12 |
| oraclebill | ok. i’ll keep looking | 21:13 |
| oraclebill | i was guessing pecan.. | 21:13 |
| gordc | possibly? i don't have a /var/log/messages location so i'm not sure. | 21:14 |
| *** ig0r_ has quit IRC | 21:14 | |
| *** pauloewerton has quit IRC | 21:16 | |
| *** ig0r_ has joined #openstack-telemetry | 21:23 | |
| *** ig0r_ has quit IRC | 21:31 | |
| *** lexholde_ has quit IRC | 21:32 | |
| *** sergio_nubeliu has quit IRC | 21:41 | |
| *** thumpba has quit IRC | 21:44 | |
| *** ig0r_ has joined #openstack-telemetry | 21:45 | |
| *** ig0r_ has quit IRC | 21:45 | |
| openstackgerrit | gordon chung proposed openstack/ceilometer: test https://review.openstack.org/266541 | 21:51 |
| *** rcernin has joined #openstack-telemetry | 21:55 | |
| *** ddaskal has joined #openstack-telemetry | 21:58 | |
| *** belmoreira has quit IRC | 22:05 | |
| *** rbak has joined #openstack-telemetry | 22:09 | |
| *** rcernin has quit IRC | 22:10 | |
| *** ddieterly has quit IRC | 22:16 | |
| *** boris-42 has quit IRC | 22:23 | |
| *** thorst has quit IRC | 22:43 | |
| *** thorst has joined #openstack-telemetry | 22:44 | |
| *** thorst has quit IRC | 22:48 | |
| *** thorst has joined #openstack-telemetry | 23:04 | |
| *** thorst has quit IRC | 23:08 | |
| *** thorst has joined #openstack-telemetry | 23:17 | |
| *** safchain has quit IRC | 23:22 | |
| *** nicodemus_ has quit IRC | 23:23 | |
| *** gordc has quit IRC | 23:29 | |
| openstackgerrit | Pradeep Kilambi proposed openstack/python-aodhclient: WIP: Add gnocchi alarm rules https://review.openstack.org/266073 | 23:31 |
| *** ddaskal has quit IRC | 23:31 | |
| *** ddaskal has joined #openstack-telemetry | 23:32 | |
| *** ddaskal has quit IRC | 23:36 | |
| *** chlong-afk has quit IRC | 23:40 | |
| *** rbak has quit IRC | 23:41 | |
| *** mragupat_ has quit IRC | 23:41 | |
| *** yassine has quit IRC | 23:43 | |
| *** mattyw has quit IRC | 23:46 | |
| *** diogogmt has quit IRC | 23:49 | |
| *** pradk has quit IRC | 23:53 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!