| *** openstack has joined #openstack-telemetry | 00:53 | |
| pleia2 | well, momentarily you should | 00:53 |
|---|---|---|
| pleia2 | there we go! | 00:53 |
| *** pleia2 has left #openstack-telemetry | 00:53 | |
| *** poi has quit IRC | 01:15 | |
| *** diogogmt has joined #openstack-telemetry | 01:21 | |
| *** ljxiash has joined #openstack-telemetry | 01:24 | |
| *** ljxiash_ has joined #openstack-telemetry | 01:24 | |
| *** ljxiash__ has joined #openstack-telemetry | 01:25 | |
| *** ljxias___ has joined #openstack-telemetry | 01:25 | |
| *** ljxia____ has joined #openstack-telemetry | 01:26 | |
| *** ljxia____ has quit IRC | 01:27 | |
| *** ljxiash has quit IRC | 01:28 | |
| *** ljxiash has joined #openstack-telemetry | 01:28 | |
| *** ljxiash__ has quit IRC | 01:29 | |
| *** ljxiash_ has quit IRC | 01:29 | |
| *** ljxias___ has quit IRC | 01:30 | |
| *** _nadya_ has joined #openstack-telemetry | 01:32 | |
| *** _nadya_ has quit IRC | 01:37 | |
| *** ari___ has quit IRC | 01:40 | |
| *** caishan has joined #openstack-telemetry | 01:40 | |
| *** Liuqing has joined #openstack-telemetry | 01:48 | |
| *** Liuqing has quit IRC | 01:58 | |
| openstackgerrit | hardik proposed openstack/ceilometer: Delete python bytecode before every test run. https://review.openstack.org/255041 | 02:02 |
| *** leecalcote has joined #openstack-telemetry | 02:03 | |
| *** leecalcote has quit IRC | 02:12 | |
| openstackgerrit | yuntongjin proposed openstack/ceilometer: event to sample publisher https://review.openstack.org/227106 | 02:40 |
| *** thorst has joined #openstack-telemetry | 02:40 | |
| *** prashantD has quit IRC | 02:46 | |
| *** diogogmt has quit IRC | 02:53 | |
| *** sanjana has joined #openstack-telemetry | 03:22 | |
| *** links has joined #openstack-telemetry | 03:24 | |
| *** thorst has quit IRC | 03:25 | |
| *** thorst has joined #openstack-telemetry | 03:43 | |
| *** thorst has quit IRC | 03:44 | |
| *** thorst has joined #openstack-telemetry | 03:45 | |
| *** alejandrito has quit IRC | 03:52 | |
| *** fawadkhaliq has joined #openstack-telemetry | 03:53 | |
| *** thorst has quit IRC | 03:54 | |
| *** fawadkhaliq has quit IRC | 04:00 | |
| *** links has quit IRC | 04:03 | |
| openstackgerrit | Merged openstack/gnocchi: statsd have some required configuration options https://review.openstack.org/248621 | 04:30 |
| openstackgerrit | Merged openstack/gnocchi: devstack: add gnocchi-statsd https://review.openstack.org/252266 | 04:34 |
| *** links has joined #openstack-telemetry | 04:38 | |
| *** fawadkhaliq has joined #openstack-telemetry | 04:45 | |
| *** fawadkhaliq has quit IRC | 04:47 | |
| *** thorst has joined #openstack-telemetry | 04:52 | |
| *** sanjana has quit IRC | 04:52 | |
| *** thorst has quit IRC | 04:58 | |
| *** boris-42_ has quit IRC | 05:03 | |
| *** ljxiash has quit IRC | 05:14 | |
| *** ljxiash has joined #openstack-telemetry | 05:14 | |
| *** ljxiash has quit IRC | 05:16 | |
| *** ljxiash has joined #openstack-telemetry | 05:16 | |
| *** fawadkhaliq has joined #openstack-telemetry | 05:33 | |
| *** _nadya_ has joined #openstack-telemetry | 05:48 | |
| *** _nadya_ has quit IRC | 05:53 | |
| *** thorst has joined #openstack-telemetry | 05:56 | |
| *** thorst has quit IRC | 06:04 | |
| *** ljxiash has quit IRC | 06:05 | |
| *** Chaozhe has joined #openstack-telemetry | 06:05 | |
| *** chaozhechen_ has joined #openstack-telemetry | 06:05 | |
| *** ljxiash has joined #openstack-telemetry | 06:05 | |
| *** caishan_ has joined #openstack-telemetry | 06:06 | |
| *** caishan has quit IRC | 06:06 | |
| *** ljxiash has quit IRC | 06:08 | |
| *** ljxiash_ has joined #openstack-telemetry | 06:08 | |
| *** chaozhechen__ has joined #openstack-telemetry | 06:08 | |
| openstackgerrit | zhanghanqun proposed openstack/gnocchi: Allow the volume display_name field to be null https://review.openstack.org/254665 | 06:09 |
| *** chaozhechen_ has quit IRC | 06:10 | |
| *** Chaozhe has quit IRC | 06:10 | |
| *** Chaozhe has joined #openstack-telemetry | 06:11 | |
| *** logan- has quit IRC | 06:24 | |
| *** ljxiash_ has quit IRC | 06:29 | |
| *** ljxiash has joined #openstack-telemetry | 06:29 | |
| *** jaypipes has joined #openstack-telemetry | 06:31 | |
| *** yprokule has joined #openstack-telemetry | 06:32 | |
| *** ljxiash has quit IRC | 06:34 | |
| *** ljxiash has joined #openstack-telemetry | 06:37 | |
| *** _nadya_ has joined #openstack-telemetry | 06:42 | |
| *** _nadya_ has quit IRC | 06:47 | |
| *** PSS has joined #openstack-telemetry | 06:51 | |
| PSS | i have cloned the python-ceilometerclient repository from github. When i tru to run the tox command, i get the ERROR: py34: InterpreterNotFound: python3.4 , ERROR: pypy: InterpreterNotFound: pypy | 06:53 |
| PSS | while py27: commands succeeded and pep8: commands succeeded | 06:53 |
| PSS | how can i resolve the above error | 06:53 |
| openstackgerrit | zhanghanqun proposed openstack/gnocchi: Allow the volume display_name field to be null https://review.openstack.org/254665 | 07:01 |
| *** thorst has joined #openstack-telemetry | 07:01 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ceilometer: Updated from global requirements https://review.openstack.org/253890 | 07:04 |
| *** shardy has joined #openstack-telemetry | 07:06 | |
| *** thorst has quit IRC | 07:09 | |
| *** heha37 has joined #openstack-telemetry | 07:16 | |
| *** heha37 has quit IRC | 07:17 | |
| *** zqfan_travel has joined #openstack-telemetry | 07:20 | |
| *** rcernin has joined #openstack-telemetry | 07:20 | |
| *** _nadya_ has joined #openstack-telemetry | 07:21 | |
| *** zqfan_travel is now known as zqfan_AFK | 07:24 | |
| *** boris-42_ has joined #openstack-telemetry | 07:31 | |
| *** cdent has joined #openstack-telemetry | 07:50 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/ceilometer: gnocchi: fix cache hash logic https://review.openstack.org/254574 | 07:55 |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/ceilometer: gnocchi: use gnocchiclient instead of requests https://review.openstack.org/237538 | 07:55 |
| *** pece has joined #openstack-telemetry | 07:56 | |
| *** pece has quit IRC | 08:04 | |
| *** thorst has joined #openstack-telemetry | 08:06 | |
| *** Chaozhe has quit IRC | 08:09 | |
| *** chaozhechen_ has joined #openstack-telemetry | 08:10 | |
| *** chaozhechen__ has quit IRC | 08:10 | |
| *** Chaozhe has joined #openstack-telemetry | 08:13 | |
| *** thorst has quit IRC | 08:14 | |
| *** belmoreira has joined #openstack-telemetry | 08:15 | |
| *** ljxiash has quit IRC | 08:23 | |
| *** ljxiash has joined #openstack-telemetry | 08:23 | |
| *** _nadya_ has quit IRC | 08:28 | |
| *** _nadya_ has joined #openstack-telemetry | 08:29 | |
| *** _nadya_ has quit IRC | 08:29 | |
| *** fawadkhaliq has quit IRC | 08:32 | |
| *** r-mibu has joined #openstack-telemetry | 08:38 | |
| *** shardy has quit IRC | 08:42 | |
| *** shardy has joined #openstack-telemetry | 08:43 | |
| *** fawadkhaliq has joined #openstack-telemetry | 08:48 | |
| *** shardy has quit IRC | 08:49 | |
| *** fawadkhaliq has quit IRC | 08:49 | |
| *** shardy has joined #openstack-telemetry | 08:51 | |
| *** safchain has joined #openstack-telemetry | 08:57 | |
| *** yassine__ has joined #openstack-telemetry | 09:01 | |
| *** Chaozhe has quit IRC | 09:01 | |
| *** chaozhechen_ has quit IRC | 09:01 | |
| *** Chaozhe has joined #openstack-telemetry | 09:03 | |
| *** chaozhechen_ has joined #openstack-telemetry | 09:03 | |
| *** openstackstatus has quit IRC | 09:05 | |
| *** openstack has joined #openstack-telemetry | 09:08 | |
| *** thorst has joined #openstack-telemetry | 09:12 | |
| *** ljxiash has quit IRC | 09:12 | |
| *** ljxiash has joined #openstack-telemetry | 09:13 | |
| *** ljxiash has quit IRC | 09:17 | |
| *** thorst has quit IRC | 09:18 | |
| openstackgerrit | hardik proposed openstack/gnocchi: Delete python bytecode before every test run. https://review.openstack.org/255147 | 09:20 |
| *** ljxiash has joined #openstack-telemetry | 09:21 | |
| openstackgerrit | sanjana proposed openstack/python-ceilometerclient: Restructuring Comment in Python-Ceilometerclient(alarms.py) Related-Bug: https://bugs.launchpad.net/python-ceilometerclient/+bug/1523866 https://review.openstack.org/255149 | 09:23 |
| openstack | Launchpad bug 1523866 in python-ceilometerclient "Unclear comment in CeilometerClient module" [Low,Triaged] - Assigned to sanjana (sanjana) | 09:23 |
| *** openstackgerrit has quit IRC | 09:32 | |
| *** openstackgerrit has joined #openstack-telemetry | 09:33 | |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: storage: round back from_timestamp in get_measures() https://review.openstack.org/252525 | 09:34 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: storage/carbonara: simplify tooz locking https://review.openstack.org/250750 | 09:34 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: storage: fix expunge_metric https://review.openstack.org/254732 | 09:34 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: serialize time period in seconds https://review.openstack.org/251402 | 09:34 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: deprecate TimeSerieArchive https://review.openstack.org/240905 | 09:34 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: tests: block when acquiring processing lock https://review.openstack.org/249130 | 09:34 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: allow to create TimeSerie from existing ts https://review.openstack.org/255155 | 09:34 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: allow to split AggregatedTimeSerie https://review.openstack.org/255156 | 09:34 |
| *** fawadkhaliq has joined #openstack-telemetry | 09:34 | |
| *** ildikov has quit IRC | 09:42 | |
| *** ViswaV_ has quit IRC | 09:49 | |
| openstackgerrit | sanjana proposed openstack/python-ceilometerclient: Restructuring Comment in Python-Ceilometerclient(alarms.py) Related-Bug: https://bugs.launchpad.net/python-ceilometerclient/+bug/1523866 https://review.openstack.org/255171 | 09:58 |
| openstack | Launchpad bug 1523866 in python-ceilometerclient "Unclear comment in CeilometerClient module" [Low,Triaged] - Assigned to sanjana (sanjana) | 09:58 |
| *** zqfan_AFK has quit IRC | 10:01 | |
| *** nijaba has quit IRC | 10:02 | |
| *** openstackgerrit has quit IRC | 10:02 | |
| *** mattyw has joined #openstack-telemetry | 10:03 | |
| *** openstackgerrit has joined #openstack-telemetry | 10:03 | |
| *** ildikov has joined #openstack-telemetry | 10:09 | |
| *** ljxiash has quit IRC | 10:13 | |
| *** ljxiash has joined #openstack-telemetry | 10:14 | |
| *** sileht has quit IRC | 10:16 | |
| *** pece has joined #openstack-telemetry | 10:16 | |
| *** _nadya_ has joined #openstack-telemetry | 10:16 | |
| *** thorst has joined #openstack-telemetry | 10:18 | |
| *** ljxiash has quit IRC | 10:18 | |
| *** sileht has joined #openstack-telemetry | 10:19 | |
| *** sileht has quit IRC | 10:19 | |
| *** sileht has joined #openstack-telemetry | 10:20 | |
| *** thorst has quit IRC | 10:24 | |
| *** chaozhechen_ has quit IRC | 10:24 | |
| *** Chaozhe has quit IRC | 10:24 | |
| openstackgerrit | Ryota MIBU proposed openstack/aodh: tempest: copy alarming client from tempest tree https://review.openstack.org/255187 | 10:28 |
| openstackgerrit | Ryota MIBU proposed openstack/aodh: tempest: copy api tests from tempest tree https://review.openstack.org/255188 | 10:28 |
| openstackgerrit | Ryota MIBU proposed openstack/aodh: tempest: copy conf.py from tempest tree https://review.openstack.org/255189 | 10:28 |
| openstackgerrit | Ryota MIBU proposed openstack/aodh: tempest: add alarming client manager https://review.openstack.org/255190 | 10:28 |
| openstackgerrit | Ryota MIBU proposed openstack/aodh: tempest: add aodh tempest plugin https://review.openstack.org/255191 | 10:28 |
| openstackgerrit | Ryota MIBU proposed openstack/aodh: WIP: tempest: use aodh client manager instead of tempest one https://review.openstack.org/255192 | 10:28 |
| *** PSS has quit IRC | 10:29 | |
| *** caishan_ has quit IRC | 10:31 | |
| *** yassine__ has quit IRC | 10:56 | |
| *** fawadkhaliq has quit IRC | 10:57 | |
| *** fawadkhaliq has joined #openstack-telemetry | 10:57 | |
| *** fawadkhaliq has quit IRC | 11:03 | |
| *** thorst has joined #openstack-telemetry | 11:23 | |
| *** links has quit IRC | 11:26 | |
| *** thorst has quit IRC | 11:28 | |
| *** eglynn has joined #openstack-telemetry | 11:32 | |
| *** ljxiash has joined #openstack-telemetry | 11:32 | |
| *** yassine__ has joined #openstack-telemetry | 12:00 | |
| *** ljxiash has quit IRC | 12:01 | |
| *** links has joined #openstack-telemetry | 12:16 | |
| *** gordc has joined #openstack-telemetry | 12:24 | |
| *** thorst has joined #openstack-telemetry | 12:27 | |
| *** thorst has quit IRC | 12:33 | |
| *** thorst has joined #openstack-telemetry | 12:36 | |
| *** fawadkhaliq has joined #openstack-telemetry | 12:38 | |
| *** alejandrito has joined #openstack-telemetry | 12:40 | |
| *** zqfan_AFK has joined #openstack-telemetry | 12:42 | |
| *** boris-42_ has quit IRC | 12:43 | |
| *** _nadya_ has quit IRC | 12:43 | |
| *** thorst_ has joined #openstack-telemetry | 12:45 | |
| *** sergio_nubeliu has joined #openstack-telemetry | 12:47 | |
| *** shardy has quit IRC | 12:49 | |
| *** thorst has quit IRC | 12:49 | |
| *** claudiub has joined #openstack-telemetry | 12:50 | |
| *** shardy has joined #openstack-telemetry | 12:51 | |
| EmilienM | gordc: bien? | 12:53 |
| gordc | EmilienM: ca va. you? | 12:56 |
| EmilienM | gordc: cold. | 12:57 |
| EmilienM | I wonder why I moved here. | 12:57 |
| gordc | it's very warm here. :) | 12:58 |
| *** spredzy has joined #openstack-telemetry | 12:59 | |
| EmilienM | I'll move to toronto one day | 13:00 |
| *** julim has joined #openstack-telemetry | 13:01 | |
| gordc | there are many places better | 13:01 |
| gordc | we don't want you here.lol | 13:01 |
| *** Chaozhe has joined #openstack-telemetry | 13:02 | |
| *** chaozhechen_ has joined #openstack-telemetry | 13:02 | |
| *** Liuqing has joined #openstack-telemetry | 13:05 | |
| *** ildikov has quit IRC | 13:07 | |
| *** _nadya_ has joined #openstack-telemetry | 13:13 | |
| *** claudiub_ has joined #openstack-telemetry | 13:14 | |
| *** Chaozhe has quit IRC | 13:14 | |
| *** chaozhechen_ has quit IRC | 13:14 | |
| *** claudiub has quit IRC | 13:14 | |
| *** Chaozhe has joined #openstack-telemetry | 13:17 | |
| *** _nadya_ has quit IRC | 13:17 | |
| *** ildikov has joined #openstack-telemetry | 13:22 | |
| *** chaozhechen_ has joined #openstack-telemetry | 13:24 | |
| *** fawadkhaliq has quit IRC | 13:27 | |
| gordc | sileht: should i use some sort of LRU cache for per resource lock? | 13:31 |
| * cdent is feeling increasingly guilty | 13:38 | |
| *** diogogmt has joined #openstack-telemetry | 13:42 | |
| gordc | cdent: you should | 13:43 |
| *** ljxiash has joined #openstack-telemetry | 13:44 | |
| gordc | cdent: maybe i should create set of locks which they all fight for?... i wonder if i can tell if a lock is being used. | 13:45 |
| cdent | is there some way you can use the existence of the cache key (which only lives for a certain amount of time) to expire the locks? | 13:47 |
| cdent | or just make sileht figure it out ;) | 13:47 |
| gordc | cdent: so we'd have a background task clearing locks? | 13:47 |
| cdent | dunno, hadn't though of it that far | 13:48 |
| cdent | not clearing locks, clearing lock keys | 13:48 |
| gordc | hmm. i'm not sure per resource locks work... we could theorectically have hundred of thousands of resources active at any point so we'll never be able to expire it | 13:52 |
| gordc | i think we need to create a set number of locks and they they fight... but then resources also have to fight if they already have a lock... | 13:53 |
| gordc | which could block out other resources... | 13:53 |
| gordc | dammit | 13:53 |
| *** eglynn has quit IRC | 13:58 | |
| *** logan- has joined #openstack-telemetry | 14:01 | |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: storage: round back from_timestamp in get_measures() https://review.openstack.org/252525 | 14:01 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: serialize time period in seconds https://review.openstack.org/251402 | 14:01 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: deprecate TimeSerieArchive https://review.openstack.org/240905 | 14:01 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: allow to split AggregatedTimeSerie https://review.openstack.org/255156 | 14:01 |
| openstackgerrit | Julien Danjou proposed openstack/gnocchi: carbonara: allow to create TimeSerie from existing ts https://review.openstack.org/255155 | 14:01 |
| sileht | gordc, you can protect the dict with a lock :p | 14:04 |
| sileht | gordc, so the global locking time is just the time to do the dict manipulation | 14:05 |
| sileht | instead of the time of the gnocchi resource creation | 14:05 |
| openstackgerrit | Merged openstack/gnocchi: storage: fix expunge_metric https://review.openstack.org/254732 | 14:05 |
| gordc | sileht: what does the dict contain? | 14:08 |
| sileht | gordc, the dict{resource_id: lock} | 14:08 |
| gordc | sileht: won't the dict still have a key per resource? | 14:09 |
| sileht | if you want to cleanup a item dict you must be sure that the lock is not acquired | 14:09 |
| gordc | so have a cleanup tasks that runs periodically? | 14:09 |
| *** ljxiash has quit IRC | 14:10 | |
| *** chaozhechen_ has quit IRC | 14:10 | |
| sileht | gordc, or each time you release the lock | 14:10 |
| *** chaozhechen_ has joined #openstack-telemetry | 14:11 | |
| *** pece has quit IRC | 14:11 | |
| *** Chaozhe has quit IRC | 14:12 | |
| *** chaozhechen_ has quit IRC | 14:12 | |
| *** cdent has quit IRC | 14:12 | |
| gordc | sileht: ah... i see. | 14:12 |
| gordc | we can add that as a naive solution. i think at scale it'll be hundred of thousands of keys still but we'll worry about that when it happens. | 14:13 |
| *** ljxiash has joined #openstack-telemetry | 14:13 | |
| *** cdent has joined #openstack-telemetry | 14:18 | |
| *** diogogmt has quit IRC | 14:24 | |
| *** diogogmt has joined #openstack-telemetry | 14:26 | |
| alejandrito | sileht, what i dont get from the implementation is ... does HEAT through the gnocchi resource, uses AODH alarming? or uses the OLD ceilometer alarm api ? | 14:27 |
| alejandrito | sileht, i really want to understand that | 14:27 |
| *** ildikov has quit IRC | 14:28 | |
| sileht | alejandrito, from heat point that the same, it use python-ceilometerclient that take care to redirect to API to aodh if the aodh is present otherwise it send the request to ceilometer-api | 14:28 |
| sileht | alejandrito, rephrased: from heat point that the same, it use python-ceilometerclient that take care to redirect to API request to aodh API if the aodh endpoint is present otherwise it send the API request to ceilometer-api endpoint | 14:29 |
| *** ildikov has joined #openstack-telemetry | 14:30 | |
| alejandrito | sileht, ok, so i now need to understand more things ... im using aodh and gnocchi from stable ... but ceilometer and ceilometer client is kilo release, so ... HOW is the ceilometer client realizing that the AODH endpoint exists ? | 14:31 |
| sileht | alejandrito, Not sure this case works | 14:32 |
| alejandrito | sileht, if you want to drop me a doc i'll read it , specefically about the client redirecting to AODH when it exists and how to do it (keystone endpoint) | 14:32 |
| gordc | hm... i don't know if i can lock the dict. that just makes all the threads fight for same lock again | 14:32 |
| sileht | alejandrito, backwardcompat have been introduced in liberty not kilo | 14:33 |
| sileht | gordc, yes but for a shorter time | 14:34 |
| alejandrito | sileht, but since we've backported for example the gnocchi dispatcher on kilo, and works ok, maybe worth to know how its done in liberty to se if i can backport it, basically want to know how its done in liberty (what endpoint type to create for AODH, what config must be done for the client to realize that it needs to redirect to aodh, etc) | 14:35 |
| sileht | alejandrito, yes you can try the latest python-ceilometerclient lib | 14:35 |
| *** ildikov has quit IRC | 14:35 | |
| sileht | alejandrito, if your keystone have the aodh endpoint, the latest python-ceilometerclient will detect that automatically | 14:36 |
| *** Chaozhe has joined #openstack-telemetry | 14:36 | |
| alejandrito | sileht, perfect, and what doc should i read to know how to perfectly add the AODH endpoint to keystone for the client to work ? | 14:37 |
| *** chaozhechen_ has joined #openstack-telemetry | 14:37 | |
| *** diogogmt has quit IRC | 14:38 | |
| gordc | sileht: but i'd have to lock it twice. lock dict to get resource lock, and lock again to possibly delete lock | 14:39 |
| sileht | alejandrito, I don't think that documented | 14:40 |
| gordc | alejandrito: you asking how to install aodh? | 14:41 |
| *** pradk has joined #openstack-telemetry | 14:42 | |
| alejandrito | sileht, gordc nope, im just asking how (EG what service type) to add the AODH endpoint on keystone in the right way, so that the last python-ceilometerclient detects the aodh endpoint and redirects the alarms that heat creates to AODH | 14:42 |
| sileht | alejandrito, https://github.com/openstack/aodh/blob/master/devstack/plugin.sh#L128 | 14:43 |
| *** pradk has quit IRC | 14:43 | |
| sileht | alejandrito, the service name is aodh and the type is alarming | 14:43 |
| *** pradk has joined #openstack-telemetry | 14:44 | |
| alejandrito | sileht, perfect, i'll try that | 14:47 |
| krotscheck | Anyone here familiar with gabbi? I'm trying to write a test that asserts that an HTTP header is _NOT_ present, and doing a 'header_name: None' doesn't seem to work | 14:52 |
| sileht | krotscheck, https://github.com/cdent/gabbi/issues/108 | 14:52 |
| krotscheck | sileht: Oh, well then! | 14:53 |
| krotscheck | That's helpful | 14:53 |
| cdent | yeah, sorry about that krotscheck | 14:53 |
| cdent | in fact that bug was written when I was trying to write tests for the CORS stuff | 14:54 |
| krotscheck | cdent: No worries. Do you have any objections to me commenting those lines out for now until a syntax is developed? | 14:54 |
| krotscheck | Also, is it valid yaml to start a property name with a !? | 14:55 |
| krotscheck | As in "response_headers: !header_i_dont_want | 14:55 |
| cdent | I've not tested that, but that's one of the ideas that was suggested | 14:55 |
| cdent | which lines do you mean with regard to commenting? | 14:56 |
| krotscheck | Hrm, it's a tag thing, | 14:56 |
| *** Chaozhe has quit IRC | 14:57 | |
| krotscheck | in https://review.openstack.org/#/c/224464/12/ceilometer/tests/functional/gabbi/gabbits/middleware.yaml, lines 25/26, 42/43 | 14:57 |
| *** rbak has joined #openstack-telemetry | 14:58 | |
| cdent | yeah, that would be cool. leave a brief comment saying why | 14:58 |
| alejandrito | sileht, is there a plan to have a python-aodhclient ? or the alarming part will be still being used by the ceilometerclient and redirecting ? | 14:58 |
| sileht | alejandrito, we have planned to create python-aodhclient but nobody have taken the tasks yet | 14:59 |
| krotscheck | cdent: Does that bug read ... "today it's response_header_not_preset"... like that would actually work? | 15:00 |
| * krotscheck checks | 15:00 | |
| krotscheck | Nope, invalid key | 15:00 |
| *** ddieterly has joined #openstack-telemetry | 15:00 | |
| *** ljxiash_ has joined #openstack-telemetry | 15:02 | |
| gordc | sileht: i'll probably start next week if no one takes it. i'm just going to copy everything you did in gnocchiclient | 15:02 |
| sileht | gordc, that's a good plan | 15:02 |
| cdent | krotscheck: no that's jason saying "if we do this now, what will be the mess that comes later for additional cases" | 15:04 |
| *** claudiub_ has quit IRC | 15:04 | |
| *** ljxiash has quit IRC | 15:05 | |
| cdent | given that headers are "special" in gabbi it might be easiest to just go with that option: here's a list of headers that shouldn't be present | 15:05 |
| cdent | it's contained | 15:05 |
| cdent | sileht, krotscheck: if you guys can think of a good name, I could implement that functionality pretty easily and quickly | 15:06 |
| *** diazjf has joined #openstack-telemetry | 15:08 | |
| gordc | not_* | 15:09 |
| gordc | cdent: i'm not creative. | 15:09 |
| cdent | not_response_header: ? | 15:10 |
| gordc | sure | 15:10 |
| *** donghao has joined #openstack-telemetry | 15:10 | |
| cdent | does that have sufficient meaning (to you)? | 15:10 |
| *** donghao has quit IRC | 15:10 | |
| cdent | I guess it probably does, especially when the value is a list | 15:10 |
| gordc | i'm just basing it on the fact that i know what response_header is and i know what not usually does | 15:11 |
| sileht | forbidden_response_headers | 15:11 |
| sileht | prevent_response_headers | 15:11 |
| cdent | thou shalt not have these headers | 15:11 |
| gordc | sileht is more fancy. | 15:11 |
| cdent | as you would expect, really | 15:12 |
| sileht | unexpected_response_headers | 15:12 |
| sileht | :p | 15:12 |
| cdent | :) | 15:12 |
| gordc | do_not_want_these_* | 15:12 |
| *** chaozhechen_ has quit IRC | 15:12 | |
| gordc | http://www.thesaurus.com/browse/not | 15:13 |
| gordc | nay_* | 15:13 |
| cdent | shun | 15:14 |
| *** chaozhechen_ has joined #openstack-telemetry | 15:15 | |
| krotscheck | I'd opt for some kind of generic negation marker. | 15:16 |
| cdent | negation is problematic because it has two different meanings: this header does not have this value is not the same as this header is not present at all | 15:16 |
| krotscheck | My initial instinct was "header_i_dont_want: None" | 15:17 |
| cdent | what if the header is present with no value? | 15:17 |
| cdent | (which is physically possible) | 15:17 |
| krotscheck | True, but even so I'm in favor of intuitive things. | 15:18 |
| gordc | cdent: i think no value and not present are different things | 15:19 |
| cdent | gordc exactly | 15:19 |
| *** ildikov has joined #openstack-telemetry | 15:20 | |
| krotscheck | How about -header_i_dont_want | 15:24 |
| *** diogogmt has joined #openstack-telemetry | 15:25 | |
| gordc | krotscheck: i like how you gave gabbi an identity ;) | 15:25 |
| krotscheck | gordc: "Gabbi" is a german colloquial for "Gabriele" or "Gabriela". :) | 15:26 |
| * krotscheck has his origins showing | 15:26 | |
| gordc | :) | 15:27 |
| krotscheck | A minus prefix is not valid yaml :/ | 15:28 |
| cdent | a german friend of mine insist that gabbi is a common name for a german prostitute | 15:28 |
| krotscheck | That could be, I'm not up to date on today's wording. | 15:29 |
| krotscheck | s/wording/usage/ | 15:29 |
| gordc | ... is that why you choose the name for lib? | 15:29 |
| gordc | i'm starting to second guess whether we should've adopted it. | 15:30 |
| *** ljxiash_ has quit IRC | 15:30 | |
| *** Ephur has joined #openstack-telemetry | 15:30 | |
| cdent | gordc: no, I didn't know when I chose the name | 15:31 |
| gordc | cdent: sure sure. | 15:31 |
| cdent | basically I was going to call it gabby (for talkative), but I thought the 'i' would make it easier to make an acronym | 15:31 |
| gordc | you folks are very creative. i created ceilometermiddleware. it's middleware, ceilometer uses it. ceiloemtermiddleware. | 15:33 |
| *** thorst_ has quit IRC | 15:35 | |
| openstackgerrit | Pradeep Kilambi proposed openstack/ceilometer: Change to allow non-metric meters https://review.openstack.org/248241 | 15:37 |
| * krotscheck is usually in gordc's camp. | 15:38 | |
| *** belmoreira has quit IRC | 15:39 | |
| *** Liuqing has quit IRC | 15:47 | |
| *** nijaba has joined #openstack-telemetry | 15:47 | |
| *** nijaba has quit IRC | 15:47 | |
| *** nijaba has joined #openstack-telemetry | 15:47 | |
| *** itxaka has joined #openstack-telemetry | 15:49 | |
| *** diazjf has quit IRC | 15:50 | |
| *** ljxiash has joined #openstack-telemetry | 15:52 | |
| *** ljxiash has quit IRC | 15:56 | |
| openstackgerrit | Michael Krotscheck proposed openstack/ceilometer: Added CORS support to Ceilometer https://review.openstack.org/224464 | 15:56 |
| krotscheck | Ok, that _should_ work. | 15:57 |
| * krotscheck grumbles about oslo sideloading configs. | 15:58 | |
| *** smurke has joined #openstack-telemetry | 15:59 | |
| krotscheck | Does ceilometer use custom headers other than X-Auth-Token and X-Openstack-Request-Id? | 16:01 |
| *** pradk has quit IRC | 16:05 | |
| gordc | krotscheck: i don't think we use anything aside from typical headers from keystone | 16:06 |
| krotscheck | gordc: Sweet, then I won't have to update that patch then :) | 16:06 |
| krotscheck | (Assuming it passes this time) | 16:06 |
| krotscheck | Incidentally, I'm kindof liking gabbi. | 16:06 |
| gordc | krotscheck: don't pad cdent's ego. | 16:07 |
| cdent | krotscheck: excellent, glad to hear it. | 16:07 |
| * cdent slaps gordc down (again) | 16:07 | |
| gordc | dammit! | 16:07 |
| *** vishwanathj has quit IRC | 16:08 | |
| *** vishwanathj has joined #openstack-telemetry | 16:08 | |
| cdent | i've been wanting to get it in more projects but haven't had time | 16:08 |
| krotscheck | Well, it's especially pertinent because right now I'm writing a ton of API mocks in javascript test harnesses, and the ability to just describe the "If request then return" as yaml seems too easy to give up. | 16:09 |
| cdent | there's work in progress to get it useful with html and forms | 16:09 |
| cdent | have you seen the gabbi-run command line tool? | 16:09 |
| gordc | cdent: i thought it was because openstack community made you jaded... just me? cool. | 16:09 |
| cdent | gordc I want it used in more openstack projects but not be considered an openstack project (or be under infra) as that discourages some people from wanting to participate | 16:10 |
| krotscheck | cdent: So, random concept. | 16:10 |
| *** pradk has joined #openstack-telemetry | 16:10 | |
| krotscheck | cdent: Is it possible ot use gabbi markup as a... shared abstraction layer? As in, I have a shared YAML file, and when run against an API makes sure that the tests pass, but then I can also run it as a mock data server and run my javascript integration tests against it without having to spin up devstack. | 16:12 |
| krotscheck | i.e. the Gabbi yaml markup is "Here's the API contract" | 16:12 |
| cdent | something like swagger or raml is probably more appropriate for that, but I've found them both a bit noisy | 16:13 |
| cdent | I've thought about swagger2gabbi and gabbi2swagger but: time is lacking | 16:13 |
| krotscheck | Eeeeehh.... I don't know about swagger. That's more of a UI, no? | 16:13 |
| krotscheck | RAML, however.... | 16:13 |
| * krotscheck ponders. | 16:13 | |
| cdent | swagger's a yaml format that can produce an explorer UI | 16:13 |
| *** sergio_nubeliu has quit IRC | 16:17 | |
| *** ljxiash has joined #openstack-telemetry | 16:21 | |
| *** diogogmt has quit IRC | 16:22 | |
| *** sergio_nubeliu has joined #openstack-telemetry | 16:23 | |
| sileht | gordc, your still on the resource lock thing or can I propose something ? | 16:33 |
| sileht | your/your are | 16:33 |
| sileht | your/you are | 16:33 |
| gordc | sileht: i was thinking something like this: http://paste.openstack.org/show/481337/ | 16:35 |
| gordc | actually never mind, i don't hink that works. | 16:36 |
| gordc | sileht: go ahead | 16:36 |
| gordc | hmm.. or will it work... i haven't really tried it out | 16:37 |
| * cdent blinks | 16:37 | |
| *** yassine__ has quit IRC | 16:38 | |
| gordc | threading is too damn hard. /me goes back to drawing lines on paper | 16:38 |
| cdent | are we quite certain the caching is worth this effort? | 16:39 |
| cdent | When I posted the WIP it was basically a way of saying "what if we did this?" not "if we dont' do this the world will explode" | 16:39 |
| cdent | that is: do measurements support the effort | 16:40 |
| gordc | cdent: i think it's the update call we're trying to avoid? let's me check if i timed it. | 16:41 |
| cdent | when I did my initial explorations I was saving a _lot_ of calls to both create and update (but mostly update) | 16:42 |
| cdent | because it was wanting to try for every dispatch | 16:42 |
| cdent | so conceptually it ought to be a huge win | 16:42 |
| *** ctina_ has joined #openstack-telemetry | 16:43 | |
| gordc | it looks like it should really only save create calls whenver we first see a resource. | 16:43 |
| gordc | it shouldn't be calling it after that initial flood comes in right? | 16:43 |
| gordc | we should make gnocchi act nosql and handle it internally. problem solved. | 16:44 |
| gordc | let's all sleep | 16:44 |
| cdent | gnocchi _does_ handle it. What we're trying to avoid is asking gnocchi to handle it (thus avoiding the http request) | 16:44 |
| gordc | huh? i meant like here's a blob of data, figure it out. | 16:45 |
| gordc | right now we are explicitly saying this resource, this metric, blob of measuremetns | 16:46 |
| cdent | without the cache, there is no way, in the dispatcher, to know if the resource associated with the current list of samples is on gnocchi already | 16:46 |
| cdent | per list of samples we only need to check once | 16:46 |
| cdent | but we get samples often | 16:46 |
| gordc | yeah, so nosql approach would be not to care. it would create that resource model for you | 16:46 |
| gordc | i'm using nosql term very loosely here | 16:46 |
| cdent | seem so | 16:47 |
| gordc | it's just more that right now we need to define schema... where tyipically nosql is just dump | 16:47 |
| gordc | sileht: http://paste.openstack.org/show/481339/ | 16:50 |
| gordc | what about that? | 16:50 |
| gordc | basically it only deletes if it can... i have no idea if it'll ever be successful | 16:51 |
| sileht | gordc, that looks better | 16:51 |
| *** _nadya_ has joined #openstack-telemetry | 16:51 | |
| gordc | sileht: what was your idea? | 16:52 |
| sileht | gordc, my idea uses one lock and one condition to handle all resources locks: http://paste.openstack.org/show/481341/ | 16:54 |
| *** mattyw has quit IRC | 16:55 | |
| gordc | sileht: yeah, i thought about using something like that, but i don't knwo conditions well enough | 16:55 |
| gordc | sileht: feel free to post it. i'm going to read up on conditions | 16:55 |
| sileht | gordc, the bad thing we condition is that is wakeup all threads when only one must be unblocked | 16:55 |
| sileht | we/with | 16:56 |
| *** thorst has joined #openstack-telemetry | 16:56 | |
| sileht | gordc, yours use more memory but doesn't have this all threads wakeup | 16:56 |
| *** prashantD has joined #openstack-telemetry | 16:56 | |
| sileht | gordc, I prefer yours | 16:56 |
| gordc | sileht: ah i see, yeah, the whole communication between all the threads started to make my brain hurt | 16:57 |
| *** ljxiash has quit IRC | 16:57 | |
| sileht | and its smaller and easier to understand | 16:57 |
| gordc | which is why i stopped reading conditions | 16:57 |
| sileht | gordc, you should remove blocking=False line 11 | 16:58 |
| sileht | gordc, so the collections will be cleaned correclty | 16:58 |
| gordc | sileht: hmm.. i added it in if another thread was getting lock, delete would just skip | 17:00 |
| gordc | getting lock from dict* | 17:01 |
| sileht | gordc, the line 12 already protect that | 17:01 |
| *** links has quit IRC | 17:01 | |
| gordc | sileht: ah! yeah, i see what you're saying. | 17:02 |
| gordc | should i change line12 to call self.__getitem__ then? | 17:02 |
| gordc | or maybe i don't get what you're saying.lol dammit threads! | 17:03 |
| gordc | i think basic idea i'm trying is self.lock(blocking=false) locks the dict if it can... | 17:04 |
| gordc | and collections.defaultdict.__getitem__(self, key)(blocking=False) locks the resource lock if it can... and (hopefully) i can delete it safely then becasue no other thread can use resource lock or dict. | 17:06 |
| *** diogogmt has joined #openstack-telemetry | 17:08 | |
| *** rcernin has quit IRC | 17:11 | |
| *** diogogmt_ has joined #openstack-telemetry | 17:13 | |
| *** diogogmt has quit IRC | 17:13 | |
| *** diogogmt_ is now known as diogogmt | 17:13 | |
| *** Chaozhe has joined #openstack-telemetry | 17:15 | |
| openstackgerrit | gordon chung proposed openstack/ceilometer: add per resource lock https://review.openstack.org/253181 | 17:16 |
| *** jaypipes has quit IRC | 17:25 | |
| *** yprokule has quit IRC | 17:30 | |
| cdent | krotscheck: if you're still around: would 'response_forbidden_headers' be okay? If it has 'response' on the start the code is _much_ easier to integrate | 17:31 |
| cdent | sileht, gordc ^ | 17:31 |
| sileht | so go for it :) | 17:31 |
| * cdent is very particular about the names of the test keys | 17:31 | |
| cdent | I'm probably too particular | 17:31 |
| gordc | cdent: is this a fake question to make it seem like i have a say? | 17:32 |
| cdent | aw | 17:32 |
| gordc | cdent: i'm cool with it. response_headers_forbidden? | 17:33 |
| gordc | i'm joking. | 17:33 |
| gordc | i really just copy paste whatever exists. | 17:33 |
| krotscheck | cdent: Works for me! | 17:33 |
| cdent | ✔ | 17:34 |
| cdent | "AssertionError: Forbidden header x-gabbi-url found in response" | 17:42 |
| krotscheck | cdent: nice! | 17:43 |
| *** itxaka has quit IRC | 17:52 | |
| *** smurke has quit IRC | 17:57 | |
| *** _nadya_ has quit IRC | 18:01 | |
| *** Chaozhe has quit IRC | 18:02 | |
| *** chaozhechen_ has quit IRC | 18:02 | |
| *** Chaozhe has joined #openstack-telemetry | 18:05 | |
| *** chaozhechen_ has joined #openstack-telemetry | 18:05 | |
| *** safchain has quit IRC | 18:09 | |
| openstackgerrit | Merged openstack/gnocchi: tests: block when acquiring processing lock https://review.openstack.org/249130 | 18:10 |
| cdent | krotscheck: in case you're curious or want to review https://github.com/cdent/gabbi/pull/109 | 18:15 |
| *** _nadya_ has joined #openstack-telemetry | 18:16 | |
| krotscheck | cdent: Looking | 18:17 |
| krotscheck | cdent: Seems sane | 18:18 |
| cdent | I'll chase up jason or fnd for some additional commentary while I eat dinner and barring somebody saying eek, I'll release it tonight | 18:18 |
| krotscheck | cdent: Neat! | 18:19 |
| *** smurke has joined #openstack-telemetry | 18:21 | |
| openstackgerrit | Pradeep Kilambi proposed openstack/python-gnocchiclient: Fix delete metric on resource create/update https://review.openstack.org/255426 | 18:30 |
| *** ViswaV has joined #openstack-telemetry | 18:33 | |
| *** ViswaV has quit IRC | 18:37 | |
| *** ViswaV has joined #openstack-telemetry | 18:38 | |
| *** _nadya_ has quit IRC | 18:53 | |
| *** david-ly_ has joined #openstack-telemetry | 19:00 | |
| *** david-ly_ is now known as david-lyle_ | 19:03 | |
| *** david-lyle has quit IRC | 19:03 | |
| *** david-lyle_ is now known as david-lyle | 19:04 | |
| *** chaozhechen_ has quit IRC | 19:05 | |
| *** Chaozhe has quit IRC | 19:05 | |
| *** diazjf has joined #openstack-telemetry | 19:33 | |
| openstackgerrit | Lianhao Lu proposed openstack/ceilometer: verify gnocchi connection before processing https://review.openstack.org/253635 | 19:37 |
| *** mattyw has joined #openstack-telemetry | 19:38 | |
| *** _cjones_ has joined #openstack-telemetry | 19:45 | |
| cdent | krotscheck: new version released: https://pypi.python.org/pypi/gabbi | 19:50 |
| *** david-lyle has quit IRC | 19:58 | |
| *** david-lyle has joined #openstack-telemetry | 19:58 | |
| *** cdent has quit IRC | 20:08 | |
| *** belmoreira has joined #openstack-telemetry | 20:41 | |
| *** diazjf has quit IRC | 20:50 | |
| *** diazjf has joined #openstack-telemetry | 20:52 | |
| *** diazjf has quit IRC | 20:57 | |
| *** diazjf has joined #openstack-telemetry | 20:58 | |
| *** ctina_ has quit IRC | 21:09 | |
| *** thorst has quit IRC | 21:12 | |
| *** thorst has joined #openstack-telemetry | 21:13 | |
| *** cdent has joined #openstack-telemetry | 21:15 | |
| alejandrito | gordc, sileht, jd__ is there an official service type in keystone for the gnocchi endpoint ? | 21:18 |
| sileht | alejandrito, alarming | 21:19 |
| openstackgerrit | gordon chung proposed openstack/ceilometer: add per resource lock https://review.openstack.org/253181 | 21:19 |
| alejandrito | sileht, for GNOCCHI too ? | 21:19 |
| gordc | metric | 21:19 |
| gordc | sorry biab. stepping out | 21:19 |
| alejandrito | sileht, gordc, perfect service type METRIC | 21:20 |
| *** sergio_nubeliu has quit IRC | 21:21 | |
| *** thorst has quit IRC | 21:21 | |
| gordc | :) | 21:23 |
| * sileht should read the whole sentence before answer... | 21:23 | |
| gordc | lol | 21:26 |
| *** belmoreira has quit IRC | 21:26 | |
| openstackgerrit | Merged openstack/gnocchi: Adds aggregation across metrics tests https://review.openstack.org/252983 | 21:37 |
| openstackgerrit | Merged openstack/gnocchi: Checks percent_of_overlap when one boundary is set https://review.openstack.org/253035 | 21:37 |
| openstackgerrit | Merged openstack/gnocchi: Adds some docs about aggregation across metrics https://review.openstack.org/253036 | 21:37 |
| *** gordc has quit IRC | 21:39 | |
| *** gordc has joined #openstack-telemetry | 21:47 | |
| *** diazjf has quit IRC | 21:54 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ceilometer: Updated from global requirements https://review.openstack.org/253890 | 21:55 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/ceilometermiddleware: Updated from global requirements https://review.openstack.org/255549 | 21:55 |
| openstackgerrit | gordon chung proposed openstack/ceilometer: Revert "Revert "devstack config for dogpile cache"" https://review.openstack.org/255569 | 22:07 |
| *** smurke has quit IRC | 22:10 | |
| openstackgerrit | gordon chung proposed openstack/ceilometer: Revert "Revert "devstack config for dogpile cache"" https://review.openstack.org/255569 | 22:10 |
| *** zqfan_AFK has quit IRC | 22:11 | |
| *** rbak has quit IRC | 22:15 | |
| krotscheck | cdent: Sweeeeet, thanks :) | 22:19 |
| *** thorst has joined #openstack-telemetry | 22:19 | |
| cdent | thank you for giving the push that made it happen. Nothing gets done without that. | 22:19 |
| krotscheck | cdent: I'm a professional squeaky wheel :) | 22:22 |
| *** thorst has quit IRC | 22:22 | |
| *** thorst has joined #openstack-telemetry | 22:22 | |
| cdent | :) | 22:22 |
| *** thorst has quit IRC | 22:24 | |
| *** thorst has joined #openstack-telemetry | 22:24 | |
| *** alejandrito has quit IRC | 22:26 | |
| *** thorst has quit IRC | 22:29 | |
| *** david-lyle has quit IRC | 22:34 | |
| openstackgerrit | Michael Krotscheck proposed openstack/ceilometer: Added CORS support to Ceilometer https://review.openstack.org/224464 | 22:34 |
| *** larainema has quit IRC | 22:37 | |
| *** david-ly_ has joined #openstack-telemetry | 22:39 | |
| *** patrickeast has quit IRC | 22:40 | |
| *** shardy has quit IRC | 22:45 | |
| *** david-ly_ is now known as david-lyle | 22:45 | |
| *** mattyw has quit IRC | 22:46 | |
| *** harlowja has quit IRC | 22:51 | |
| *** harlowja has joined #openstack-telemetry | 22:51 | |
| *** boris-42_ has joined #openstack-telemetry | 22:59 | |
| *** rbak has joined #openstack-telemetry | 23:19 | |
| *** Ephur has quit IRC | 23:25 | |
| *** thorst has joined #openstack-telemetry | 23:27 | |
| *** thorst has quit IRC | 23:34 | |
| *** larainema has joined #openstack-telemetry | 23:34 | |
| *** rcernin has joined #openstack-telemetry | 23:38 | |
| *** pradk has quit IRC | 23:40 | |
| *** claudiub has joined #openstack-telemetry | 23:42 | |
| openstackgerrit | gordon chung proposed openstack/ceilometer: fix events rbac https://review.openstack.org/234823 | 23:42 |
| harlowja | gordc yt | 23:42 |
| harlowja | cdent | 23:42 |
| harlowja | i gotta question for u all | 23:43 |
| cdent | howdy | 23:43 |
| harlowja | yo yo | 23:43 |
| gordc | harlowja: i'm stepping out. | 23:43 |
| harlowja | kk | 23:43 |
| gordc | will wait a few min to see if it's easy | 23:43 |
| harlowja | its about scheduling and utilization aware scheduling | 23:43 |
| harlowja | not so easy probably | 23:43 |
| gordc | ...nope.lol | 23:43 |
| harlowja | ;) | 23:43 |
| gordc | all on cdent. | 23:44 |
| harlowja | a folk internally is starting to get involved in this (cause well it makes sense) | 23:44 |
| gordc | i'll read eavesdrop or check mail if needed. | 23:44 |
| harlowja | kk | 23:44 |
| gordc | latesr | 23:44 |
| *** gordc has quit IRC | 23:44 | |
| harlowja | annnnnd before it gets to deep into well we will extend nova to do more and more and more and more resource awareness and tracking and ... welllll maybe i'm thinking u guys have some known plan on this anyway | 23:44 |
| harlowja | and i was like welllll why not talk with u guys (and i'm here to do a little of that before i guess, ha) | 23:45 |
| cdent | sounds interesting but first a caveat: it's nearly midnight and I've had a long day so I'm pretty stupid at the moment | 23:45 |
| harlowja | i'm stupid all the time | 23:45 |
| harlowja | lol | 23:45 |
| cdent | stupids unite! | 23:45 |
| harlowja | ;) | 23:45 |
| openstackgerrit | Michael Krotscheck proposed openstack/ceilometer: Added CORS support to Ceilometer https://review.openstack.org/224464 | 23:45 |
| harlowja | derp | 23:45 |
| harlowja | so maybe can discuss other time, thats fine, i just wanted to see if people have been thinking about it | 23:45 |
| cdent | so, weirdly, I've sort of moved over to nova-land a bit and tracking resources, resource usage, and scheduler stuff is where I'm going to be poking | 23:45 |
| harlowja | aka, the feedback loop of nova -> nova-scheuduler -> nova-compute -> vms -> metrics -> cielometer(?) -> nova .... | 23:46 |
| harlowja | cdent hmmmm | 23:46 |
| cdent | within ceilometer there is a frequent debate about "what's all this data for" that has never been resolved well | 23:46 |
| harlowja | :) | 23:46 |
| harlowja | stupids will figure it out | 23:46 |
| harlowja | go stupid! | 23:47 |
| harlowja | lol | 23:47 |
| harlowja | ok, perhaps we can talk when less stupid is about | 23:47 |
| cdent | I overhead a conversation a couple days ago of someone saying "if we do come up with some kind of notification driven resource tracking it will at least have to support using ceilometer if not specifically use it" | 23:48 |
| harlowja | ya | 23:48 |
| cdent | so I reckon this is something which is kind of on people's minds | 23:48 |
| harlowja | https://review.openstack.org/#/c/254423/ is popping up (from a fellow coworker at y!) | 23:48 |
| * cdent marks that for later review | 23:49 | |
| harlowja | its similar, and got me thinking about this and the path forward here (not just the short-term path, of shove stuff into nova...) | 23:49 |
| harlowja | might as well use that data for something imho :) | 23:49 |
| cdent | my concern with a focus on ceilometer's style of handling such data is that ceilometer is usually concerned with the aggregation of the info over time | 23:49 |
| harlowja | use stupid scheduling until enough information has been gathered? | 23:49 |
| harlowja | then use less stupid one basedo n data | 23:50 |
| harlowja | ? | 23:50 |
| cdent | whereas resource tracking is more about "how much of this do we have _now_" driven by hearing information about consumption | 23:50 |
| cdent | "instance X just at 5GB RAM" | 23:50 |
| cdent | s/at/ate | 23:50 |
| harlowja | tasty | 23:50 |
| *** ljxiash has joined #openstack-telemetry | 23:50 | |
| cdent | I need to learn more about the existing metric driven filters to really know what I'm talking about | 23:51 |
| cdent | because thus far my intuitive reaction to the scheduler in general has been "weird" | 23:51 |
| harlowja | ya, let me see if i can get the coworker guy to get in here, when u are less tired | 23:51 |
| *** patrickeast has joined #openstack-telemetry | 23:51 | |
| harlowja | he should be able to be a good explainer of it (i would hope) | 23:51 |
| cdent | sounds like an excellent plan | 23:51 |
| harlowja | cool | 23:51 |
| cdent | afternoons/early evenings UTC are usually best | 23:51 |
| harlowja | kk | 23:52 |
| cdent | for me having a somewhat less stupid (but still stupid) brain | 23:52 |
| harlowja | u can never get all the way rid of the stupid | 23:52 |
| harlowja | i've tried | 23:52 |
| cdent | you might catch more fish with a mailing list post targeting both nova and ceilo | 23:52 |
| harlowja | right | 23:52 |
| harlowja | *eventually that i think | 23:52 |
| cdent | yeah, finding the ground might be good first | 23:52 |
| cdent | I reckon this is a fine opportunity for me to go to be | 23:53 |
| cdent | d | 23:53 |
| cdent | but yeah, look for me later | 23:53 |
| * cdent bows out | 23:54 | |
| cdent | g'night | 23:54 |
| harlowja | :) | 23:54 |
| harlowja | lata | 23:54 |
| *** cdent has quit IRC | 23:54 | |
| *** ljxiash has quit IRC | 23:55 | |
| * krotscheck wonders if cdent got around to doing a global requirements patch for gabbi | 23:56 | |
| * krotscheck goes to check | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!