Thursday, 2014-05-08

*** matsuhashi has joined #openstack-ceilometer00:07
openstackgerritDina Belova proposed a change to openstack/ceilometer: TEST COMMIT TO FIND RPC PROBLEM  https://review.openstack.org/9248300:11
*** rwsu has joined #openstack-ceilometer00:20
*** matsuhashi has quit IRC00:28
*** matsuhas_ has joined #openstack-ceilometer00:31
*** prad_ has quit IRC00:39
*** matsuhas_ has quit IRC00:40
*** matsuhashi has joined #openstack-ceilometer00:41
*** sdake_ has quit IRC01:01
*** nacim has quit IRC01:05
*** shakamunyi has joined #openstack-ceilometer01:27
*** jaypipes has quit IRC01:28
*** _cjones_ has quit IRC01:40
*** _cjones_ has joined #openstack-ceilometer01:41
*** shakamunyi has quit IRC01:42
*** nosnos has joined #openstack-ceilometer01:44
*** _cjones_ has quit IRC01:45
*** shakamunyi has joined #openstack-ceilometer01:57
*** thomasem has joined #openstack-ceilometer01:59
*** thomasem has quit IRC01:59
*** mengxd has joined #openstack-ceilometer02:04
*** thomasem has joined #openstack-ceilometer02:06
*** thomasem has quit IRC02:06
*** thomasem has joined #openstack-ceilometer02:07
*** shakayumi has joined #openstack-ceilometer02:16
*** thomasem has quit IRC02:16
*** rdmcnair has joined #openstack-ceilometer02:19
*** shakamunyi has quit IRC02:19
*** gordc has joined #openstack-ceilometer02:40
*** shakayumi has quit IRC03:01
*** renlt has joined #openstack-ceilometer03:04
*** changbl has quit IRC03:21
*** matsuhashi has quit IRC03:24
*** shakayumi has joined #openstack-ceilometer03:31
*** gordc has quit IRC03:37
*** nosnos has quit IRC03:41
*** rdmcnair has quit IRC04:04
*** shakayumi has quit IRC04:08
openstackgerritliusheng proposed a change to openstack/ceilometer: Improves the query_to_kwargs method  https://review.openstack.org/9206604:13
*** mengxd has quit IRC04:28
*** nosnos has joined #openstack-ceilometer04:29
*** matsuhashi has joined #openstack-ceilometer04:37
*** matsuhashi has quit IRC05:19
*** matsuhashi has joined #openstack-ceilometer05:20
*** fnaval_ has quit IRC05:39
openstackgerritOpenStack Proposal Bot proposed a change to openstack/ceilometer: Imported Translations from Transifex  https://review.openstack.org/8850606:06
*** eglynn has joined #openstack-ceilometer06:25
eglynnsileht: good morning!06:26
eglynnsileht: ... I'm thinking we should revert that context-as-dict patch, as it seems to have broken the notification agent06:27
eglynnhttps://review.openstack.org/#/c/92720/06:27
eglynn... pending another crack at the fix, probably best to have the notification agent on master back in an operational state06:28
*** ildikov has quit IRC06:29
*** eglynn_ has joined #openstack-ceilometer06:33
*** eglynn has quit IRC06:35
*** eglynn_ has quit IRC07:02
openstackgerritliusheng proposed a change to openstack/ceilometer: Improve the timestamp validation of ceilometer API  https://review.openstack.org/9278607:03
*** _nadya_ has joined #openstack-ceilometer07:06
*** ildikov has joined #openstack-ceilometer07:08
*** eglynn has joined #openstack-ceilometer07:09
openstackgerritliusheng proposed a change to openstack/ceilometer: Improves the query_to_kwargs method  https://review.openstack.org/9278807:23
*** _nadya_ has quit IRC07:25
openstackgerritliusheng proposed a change to openstack/ceilometer: Improves the query_to_kwargs method  https://review.openstack.org/9278807:27
*** _nadya_ has joined #openstack-ceilometer07:29
*** matsuhashi has quit IRC07:54
*** matsuhashi has joined #openstack-ceilometer07:55
*** eglynn has quit IRC07:56
*** ildikov has quit IRC07:57
*** ildikov has joined #openstack-ceilometer07:57
*** renlt has quit IRC08:22
*** liusheng has joined #openstack-ceilometer08:30
openstackgerritA change was merged to openstack/ceilometer: Revert "oslo.messaging context must be a dict"  https://review.openstack.org/9272008:37
*** eglynn has joined #openstack-ceilometer08:47
*** _nadya_ has quit IRC09:15
*** aviau has quit IRC09:46
*** Alexei_987 has joined #openstack-ceilometer09:50
*** aviau has joined #openstack-ceilometer10:01
*** _nadya_ has joined #openstack-ceilometer10:23
*** matsuhashi has quit IRC10:27
*** matsuhashi has joined #openstack-ceilometer10:40
*** _nadya_ has quit IRC10:48
*** _nadya_ has joined #openstack-ceilometer10:48
*** ildikov is now known as ildikov_fan_repa10:57
*** ildikov_fan_repa has quit IRC10:57
*** _nadya_ has quit IRC11:00
*** matsuhas_ has joined #openstack-ceilometer11:22
*** matsuhashi has quit IRC11:23
*** Alexei_987 has quit IRC11:23
*** erecio has quit IRC11:23
*** isviridov has quit IRC11:23
openstackgerritA change was merged to openstack/ceilometer: Correcting formatting and adding period in measurement doc  https://review.openstack.org/9170811:26
*** Alexei_987 has joined #openstack-ceilometer11:41
eglynnhey folks!11:46
eglynnin advance of today's team meeting, please take a look at the draft BP template for Juno I've put together11:46
eglynnhttps://etherpad.openstack.org/p/ceilometer-juno-blueprint-template11:46
eglynn... it's heavily based on the nova BP template11:47
eglynn... but with a fair bit of customization for ceilometer specifics11:47
eglynn... TL;DR: the explicit diffs WRT the nova original are ... http://fpaste.org/100125/54945513/11:47
*** isviridov has joined #openstack-ceilometer11:48
*** erecio has joined #openstack-ceilometer11:48
eglynn... feel free to hack away or stick feedback/comments inline11:48
eglynn... thanks!11:48
eglynntyping makes /me hungry :)11:48
*** eglynn is now known as eglynn-lunch11:48
*** DrBacchus has joined #openstack-ceilometer11:56
*** fnaval has joined #openstack-ceilometer11:59
*** ondergetekende has quit IRC12:04
*** Shaan7 has quit IRC12:04
openstackgerritChangBo Guo(gcb) proposed a change to openstack/ceilometer: Clean up openstack-common.conf  https://review.openstack.org/9198812:14
*** ondergetekende has joined #openstack-ceilometer12:18
*** Shaan7 has joined #openstack-ceilometer12:18
*** eglynn-lunch is now known as eglynn12:19
*** jdob has joined #openstack-ceilometer12:26
*** julim has joined #openstack-ceilometer12:33
eglynnsileht: quick question about this eventlet monkey-patching change https://review.openstack.org/79628 if you got a sec?12:50
eglynnsileht: ... just wondering if there was a strong motiviation for consolidating all the console script entry points in cli.py ?12:51
*** gordc has joined #openstack-ceilometer12:51
eglynnsileht: ... or whether it was just done for neatness/consistency (context is https://bugs.launchpad.net/ceilometer/+bug/1317210 )12:51
*** aviau has quit IRC12:53
*** fnaval has quit IRC13:04
*** fnaval has joined #openstack-ceilometer13:05
*** changbl has joined #openstack-ceilometer13:18
*** aviau has joined #openstack-ceilometer13:21
openstackgerritBalazs Gibizer proposed a change to openstack/ceilometer: Introduce interface changes for project based data collection  https://review.openstack.org/9282213:25
openstackgerritBalazs Gibizer proposed a change to openstack/ceilometer: Adds project based filtering for compute pollsters  https://review.openstack.org/9282313:25
*** fnaval has quit IRC13:31
eglynnDrBacchus: in what version of ceilometer did you see that empty POST body for alarm notifications?13:34
eglynnDrBacchus: (... mentioned yesterday on IRC)13:34
eglynnDrBacchus: ... I just fired up a tiny python webserver that dumps incoming POSTs to a text file and I'm seeing the expected JSON body http://ur1.ca/ha2cb13:34
eglynnDrBacchus: ... ^^^ that's against master13:35
*** nosnos has quit IRC13:38
DrBacchuseglynn: I installed on CentOS via the RDO quick start right after the Icehouse release.13:39
DrBacchus[root@rdodemo ~(keystone_admin)]# ceilometer --version13:40
DrBacchus1.0.813:40
eglynnDrBacchus: $ sudo rpm -qa | grep ceilometer13:40
eglynnDrBacchus: (... that's just the CLI version given above)13:41
DrBacchuspython-ceilometerclient-1.0.8-1.el6.noarch13:41
DrBacchusopenstack-ceilometer-common-2014.1-1.el6.noarch13:41
DrBacchusopenstack-ceilometer-api-2014.1-1.el6.noarch13:41
DrBacchuspython-ceilometer-2014.1-1.el6.noarch13:41
DrBacchusopenstack-ceilometer-collector-2014.1-1.el6.noarch13:41
DrBacchusopenstack-ceilometer-notification-2014.1-1.el6.noarch13:41
DrBacchusopenstack-ceilometer-central-2014.1-1.el6.noarch13:41
DrBacchusopenstack-ceilometer-alarm-2014.1-1.el6.noarch13:41
eglynnDrBacchus: kk, I'll have a look at the specific RDO version13:41
DrBacchusLooks like there's updated versions of most of those.13:42
*** changbl has quit IRC13:42
DrBacchusShould I just update and try again? Perhaps this has been fixed already.13:43
*** changbl has joined #openstack-ceilometer13:43
eglynnDrBacchus: well I think you've got the latest build right there13:49
*** zul has joined #openstack-ceilometer13:49
eglynnDrBacchus: ... so I don't think upgrading is time well spent13:49
eglynnDrBacchus: ... I've got an f20 install of the same RDO version I can do a quick check against13:50
*** rdmcnair has joined #openstack-ceilometer13:52
*** changbl has quit IRC13:54
eglynnDrBacchus: ... yeap I'm seeing that POST body in that notification13:55
DrBacchusok ... bizarre. I wonder what I could be doing wrong.13:56
eglynnDrBacchus: ... here's what I used to create alarm:13:56
eglynnceilometer alarm-threshold-create --name non_zero_images --description 'got images' --meter-name image --threshold 0.01 --comparison-operator gt  --statistic avg --period 60 --evaluation-periods 1 --alarm-action 'http://localhost:8045'13:56
eglynnDrBacchus: ... and here's the POST request entity body13:57
DrBacchusLooks very similar to what I did.13:57
eglynn{"current": "alarm", "alarm_id": "778f5f95-e7a6-4cd4-b771-fc8b75e16c43", "reason": "Transition to alarm due to 1 samples outside threshold, most recent: 1.0", "reason_data": {"count": 1, "most_recent": 1.0, "type": "threshold", "disposition": "outside"}, "previous": "insufficient data"}13:57
eglynnDrBacchus: ... could it simply be an error in your micro-webserver?13:57
DrBacchusThe thing I was posting to is a php script running on Apache. Let me try something different.13:58
eglynnDrBacchus: ... here's the dumbest little python webserver in the west13:59
eglynnDrBacchus: ... http://fpaste.org/100166/99557517/13:59
*** thomasem has joined #openstack-ceilometer14:00
*** fnaval has joined #openstack-ceilometer14:01
*** prad_ has joined #openstack-ceilometer14:01
*** julim has quit IRC14:06
*** changbl has joined #openstack-ceilometer14:06
DrBacchusceilometer is being angry. One moment.14:07
*** gordc has quit IRC14:13
*** gordc has joined #openstack-ceilometer14:14
DrBacchusAh. It's complaining that amqp isn't running. Annoying14:16
*** prad__ has joined #openstack-ceilometer14:19
*** prad_ has quit IRC14:22
eglynnDrBacchus: ... annoying that ceilo complains, or annoying that the amqp provider has died?14:22
DrBacchusthe latter.14:23
eglynncool :)14:24
DrBacchusAh. Typoed port number. Geez.14:25
eglynnDrBacchus: ... d'ya mean, typo in the alarm_action URL?14:25
DrBacchusYes. Trying again.14:26
DrBacchusI don't know what's going on. The alarm isn't triggering at all now. Trying to figure out why.14:30
eglynnDrBacchus: is repeat_actions set to True?14:33
eglynnDrBacchus: ... otherwise notifications only occur on the actual state transition14:34
eglynnDrBacchus: ... for continuous notification, you'll need: $ ceilometer alarm-update -a $ALARM_ID --repeat-actions True14:35
DrBacchusok.14:35
eglynnDrBacchus: ... also note that the alarm-evaluator nudges the alarm-notifier over AMQP to trigger the alarm action14:36
*** gordc has quit IRC14:36
DrBacchusyeah. amqp is running now14:36
eglynnDrBacchus: ... so may be related to your AMQP outage referred to above14:36
eglynnDrBacchus: ... k14:36
*** _nadya_ has joined #openstack-ceilometer14:38
DrBacchusIn the ceilometer log I see14:39
DrBacchusINFO urllib3.connectionpool [-] Starting new HTTP connection (1): 192.168.0.17614:39
DrBacchusBut I don't see that request on the webserver.14:39
DrBacchusI can telnet to the webserver port and issue requests manually, so it's not a firewall problem.14:40
*** mengxd has joined #openstack-ceilometer14:41
DrBacchusI feel like I'm probably chasing a problem that's outside of ceilometer entirely, but I'm not sure.14:41
eglynnDrBacchus: is that log msg quited above that in the alarm-notifier.log?14:41
eglynnDrBacchus: ... I would expect something more like this14:42
eglynn2014-05-08 14:39:12.096 6553 INFO urllib3.connectionpool [-] Starting new HTTP connection (1): localhost14:42
eglynn2014-05-08 14:40:12.090 6553 INFO ceilometer.alarm.notifier.rest [req-02da581f-478f-4ed0-85f9-d6693ac1ef06 - - - - -] Notifying alarm 778f5f95-e7a6-4cd4-b771-fc8b75e16c43 from alarm to alarm with action SplitResult(scheme=u'http', netloc=u'localhost:8045', path=u'', query='', fragment='') because Remaining as alarm due to 1 samples outside threshold, most recent: 1.014:42
eglynnDrBacchus: ... but yeah, agreed, smelling less and less like a ceilometer bug14:42
DrBacchusNothing happening in alarm-notifier. That message was in api.log14:43
DrBacchusdeleting alarms and starting over. I must have screwed something up.14:44
DrBacchuseglynn: http://paste.openstack.org/show/79599/14:47
eglynnDrBacchus: ... haven't seen that before, that's complaining that it can't load the stevedore extension to dispatch data to the metering store14:49
DrBacchusNot sure what I've done but I think I might try a fresh VM and try this again.14:49
eglynnDrBacchus: ... that stevedore loading is just a simple mechanism based on the entry_points.txt14:49
eglynnDrBacchus: ... so yeah, smells like summat fundamental is askew with your VM14:50
eglynnDrBacchus: ... such as a corrupt disk or block dev14:50
DrBacchusok, thanks for the help. Sorry to waste your time. I'll let you know if I have any other problems.14:50
eglynnDrBacchus: ... yeah, we're talking cattle not pets, shoot it in the head and move on!14:51
eglynnDrBacchus: ... np!14:51
*** mengxd has quit IRC14:55
*** mengxd has joined #openstack-ceilometer14:55
*** mengxd has quit IRC14:55
*** mengxd has joined #openstack-ceilometer14:56
*** ildikov has joined #openstack-ceilometer14:58
*** gordc has joined #openstack-ceilometer15:00
*** julim has joined #openstack-ceilometer15:09
*** mengxd has quit IRC15:14
*** mengxd has joined #openstack-ceilometer15:15
*** fnaval has quit IRC15:20
*** fnaval has joined #openstack-ceilometer15:21
*** matsuhas_ has quit IRC15:25
*** matsuhashi has joined #openstack-ceilometer15:25
*** fnaval has quit IRC15:25
*** jaypipes has joined #openstack-ceilometer15:29
*** fnaval has joined #openstack-ceilometer15:29
*** matsuhashi has quit IRC15:31
*** _cjones_ has joined #openstack-ceilometer15:32
*** nekron99 has joined #openstack-ceilometer15:32
*** changbl has quit IRC15:33
*** _cjones_ has quit IRC15:48
*** changbl has joined #openstack-ceilometer15:48
*** _cjones_ has joined #openstack-ceilometer15:53
silehteglynn, o/15:59
silehteglynn, for the oslo.messaging I will try to write tests that doesn't mock anything of oslo.messaging, we have a fake driver that allows that.16:00
*** llu-laptop has joined #openstack-ceilometer16:01
*** Ruetobas has quit IRC16:01
eglynnsileht: hey!16:01
eglynnsileht: ... so I hope I wasn't too hasty in reverting that oslo-messaging patch?16:01
silehteglynn, for the context issue, I think the best option is to not unserialize the context for notification listener, because we cannot be sure of the dict content16:01
eglynnsileht: a-ha, yes, the content is completely out of our control in that case, right16:02
silehteglynn, it's ok for the revert, I'm off today and tomorrow, so I can't really work on that16:03
*** Ruetobas has joined #openstack-ceilometer16:03
silehteglynn, if you really need a fix for alarming you can propose my first patchset, it is a correct workaround16:03
eglynnsileht: yeah, cool, that's a reasonable holding pattern for now, pending a more complete fix16:04
*** llu-laptop has quit IRC16:04
eglynnsileht: ... I look at doing that so16:04
silehteglynn, also about: https://review.openstack.org/#/c/79628/, the issue I have encounter, was some imported modules was not correctly monkey patched because previously ceilometer monkey patch library at runtime and not at module load time16:06
silehteglynn, so any modules that have created an object with the unpatched module during module loading16:07
eglynnsileht: ... k, so importing all in the cli.py was a deliberate part of the fix for that?16:07
silehteglynn, was not monkey pathched16:07
eglynnsileht: ... i.e. moving the console_scripts entry points to the cli.py16:08
silehteglynn, I just want to be sure all things that bootstrap a part of ceilometer are correctly monkeypatched16:08
eglynnsileht: ... just wondering if the same effect re. monkey patching could be acheived without the dependencies all be consolidated like that16:08
*** Ruetobas has quit IRC16:08
eglynnsileht: ... as that gives our downstream packagers a sadface16:09
silehteglynn, it looks that redhat packages are more clean that the debian/ubuntu packages ;)16:10
silehteglynn, we can split the cli.py if needed, they are no special requirement16:12
eglynnsileht: ... cool, thanks, I'll talk to pbrady on what exactly we can live with in terms of consolidation16:12
eglynnsileht: ... *he can live with16:13
*** Ruetobas has joined #openstack-ceilometer16:14
*** Alexei_987 has quit IRC16:14
*** changbl has quit IRC16:16
*** _nadya_ has quit IRC16:18
*** nekron99 has quit IRC16:22
*** changbl has joined #openstack-ceilometer16:29
openstackgerritEugeniya Kudryashova proposed a change to openstack/python-ceilometerclient: Use HTTPClient from common Oslo code  https://review.openstack.org/6893916:33
*** eglynn has quit IRC16:34
*** ildikov has quit IRC16:43
*** georgem2 has joined #openstack-ceilometer16:50
georgem2is it possible to send the alert generated by ceilometer by email?17:17
*** blinky_ghost has quit IRC17:19
*** zul has quit IRC17:30
*** blinky_ghost has joined #openstack-ceilometer17:32
*** blinky_ghost has quit IRC17:43
*** _nadya_ has joined #openstack-ceilometer17:54
*** blinky_ghost has joined #openstack-ceilometer17:56
*** _nadya_ has quit IRC18:14
*** zul has joined #openstack-ceilometer18:14
*** _nadya_ has joined #openstack-ceilometer18:34
*** georgem2 has quit IRC18:52
*** nekron99 has joined #openstack-ceilometer18:53
*** promulo has joined #openstack-ceilometer18:55
*** xdmeng has joined #openstack-ceilometer18:57
*** blinky_ghost has quit IRC18:57
*** mengxd has quit IRC18:57
*** openstackgerrit has quit IRC19:04
*** openstackgerrit has joined #openstack-ceilometer19:05
*** _nadya_ has quit IRC19:18
*** _nadya_ has joined #openstack-ceilometer19:49
*** ildikov has joined #openstack-ceilometer19:50
*** _nadya_ has quit IRC19:53
*** thomasem has quit IRC19:58
*** thomasem has joined #openstack-ceilometer19:58
*** thomasem has joined #openstack-ceilometer19:59
*** reazem_ has joined #openstack-ceilometer20:01
*** aviau has quit IRC20:04
*** nekron99 has quit IRC20:06
*** jdob has quit IRC20:06
*** julim has quit IRC20:08
*** dperaza1 has joined #openstack-ceilometer20:58
*** thomasem has quit IRC20:59
*** dperaza has quit IRC20:59
*** thomasem has joined #openstack-ceilometer21:00
*** thomasem has quit IRC21:11
*** underyx is now known as Underyx|off21:20
*** xdmeng has quit IRC21:44
*** mengxd has joined #openstack-ceilometer21:44
*** zul has quit IRC22:05
*** fnaval has quit IRC22:27
*** prad__ has left #openstack-ceilometer22:27
*** gordc has quit IRC22:37
*** erecio has quit IRC22:42
*** fnaval has joined #openstack-ceilometer23:01
*** mengxd has quit IRC23:03

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!