Thursday, 2016-06-02

*** jmlowe has joined #openstack-telemetry00:01
*** diogogmt has joined #openstack-telemetry00:17
*** r-mibu has quit IRC00:36
*** r-mibu has joined #openstack-telemetry00:36
*** thorst has joined #openstack-telemetry00:46
*** cheneydc has joined #openstack-telemetry00:59
*** sheel has joined #openstack-telemetry01:28
*** elemoine has quit IRC01:51
*** elemoine has joined #openstack-telemetry01:56
*** jwcroppe has joined #openstack-telemetry02:02
*** vishwanathj has joined #openstack-telemetry02:03
openstackgerritZhenyu Zheng proposed openstack/python-aodhclient: Fix py34  https://review.openstack.org/32416702:08
*** ljxiash has joined #openstack-telemetry02:20
*** vinaypotluri has quit IRC02:20
*** liamji has joined #openstack-telemetry02:23
openstackgerritZhenyu Zheng proposed openstack/python-aodhclient: Fix py34  https://review.openstack.org/32416702:23
*** ljxiash has quit IRC02:26
*** ljxiash has joined #openstack-telemetry02:27
*** ljxiash has quit IRC02:31
openstackgerritliusheng proposed openstack/ceilometer: TEST: fix the py34 jenkins job  https://review.openstack.org/32417402:39
*** ljxiash has joined #openstack-telemetry02:57
*** rbak has quit IRC02:59
*** rbak has joined #openstack-telemetry02:59
*** vishwanathj has quit IRC03:01
*** julim has joined #openstack-telemetry03:05
*** diogogmt has quit IRC03:10
openstackgerritliusheng proposed openstack/python-aodhclient: Avoid AttributeError in exception handling  https://review.openstack.org/32418703:21
*** john5223 has quit IRC03:27
*** john5223 has joined #openstack-telemetry03:27
*** john5223 has quit IRC03:28
openstackgerritZhenyu Zheng proposed openstack/python-aodhclient: Fix py34  https://review.openstack.org/32416703:28
*** john5223 has joined #openstack-telemetry03:32
*** links has joined #openstack-telemetry03:35
*** julim has quit IRC03:36
*** ljxiash has quit IRC03:37
*** ljxiash has joined #openstack-telemetry03:37
*** ljxiash has quit IRC03:39
*** ljxiash has joined #openstack-telemetry03:39
zqfan"ImportError: No module named 'vine'" blocks everything03:52
*** liusheng has quit IRC03:52
zqfanit is in amqp package03:52
*** liusheng has joined #openstack-telemetry03:52
*** ljxiash has quit IRC03:58
*** thorst has quit IRC03:58
*** ljxiash has joined #openstack-telemetry03:58
*** thorst has joined #openstack-telemetry03:58
*** ljxiash has quit IRC04:03
*** thorst has quit IRC04:07
openstackgerritZhiQiang Fan proposed openstack/ceilometer: catch DriverLoadFailure for get_transport optional  https://review.openstack.org/32420104:20
*** smurke has quit IRC04:23
*** dave-mccowan has quit IRC04:36
openstackgerritliusheng proposed openstack/ceilometer: Fix the py34 jenkins job  https://review.openstack.org/32417404:40
openstackgerritliusheng proposed openstack/aodh: Support combination alarm to composite alarm conversion  https://review.openstack.org/32081204:48
openstackgerritZhenyu Zheng proposed openstack/python-aodhclient: Add releasenotes for Newton updates  https://review.openstack.org/32038304:51
openstackgerritliusheng proposed openstack/aodh: DriverLoadFailure for get_transport optional  https://review.openstack.org/32420704:55
*** ljxiash has joined #openstack-telemetry04:56
openstackgerritliusheng proposed openstack/aodh: Catch DriverLoadFailure for get_transport optional  https://review.openstack.org/32420704:56
*** thorst has joined #openstack-telemetry05:05
openstackgerritliusheng proposed openstack/aodh: Bump the oslo.messaging version  https://review.openstack.org/32421005:05
openstackgerritZhenyu Zheng proposed openstack/ceilometer: Bump to Nova v2.1  https://review.openstack.org/32421105:06
openstackgerritZhiQiang Fan proposed openstack/aodh: bump oslo.messaging to 5.2.0 to avoid broken amqp  https://review.openstack.org/32421205:07
*** thorst has quit IRC05:12
*** jwcroppe has quit IRC05:25
*** yprokule has joined #openstack-telemetry05:27
*** rcernin has joined #openstack-telemetry05:48
openstackgerritZhiQiang Fan proposed openstack/ceilometer: catch DriverLoadFailure for get_transport optional  https://review.openstack.org/32420105:52
openstackgerritZhenyu Zheng proposed openstack/ceilometer: Bump to Nova v2.1  https://review.openstack.org/32421106:07
*** thorst has joined #openstack-telemetry06:10
openstackgerritliusheng proposed openstack/aodh: Catch DriverLoadFailure for get_transport optional  https://review.openstack.org/32420706:12
*** thorst has quit IRC06:17
*** nadya has joined #openstack-telemetry06:28
openstackgerritMerged openstack/python-aodhclient: Add releasenotes for Newton updates  https://review.openstack.org/32038306:29
openstackgerritZhenyu Zheng proposed openstack/ceilometer: Bump to Nova v2.1  https://review.openstack.org/32421106:36
*** llu is now known as llu_06:40
*** llu has joined #openstack-telemetry06:41
*** nadya has quit IRC06:47
*** tesseract- has joined #openstack-telemetry06:53
*** cdent has joined #openstack-telemetry07:05
*** thorst has joined #openstack-telemetry07:15
*** vishwanathj has joined #openstack-telemetry07:15
*** vishwana_ has joined #openstack-telemetry07:15
openstackgerritMerged openstack/ceilometer: catch DriverLoadFailure for get_transport optional  https://review.openstack.org/32420107:17
*** vishwanathj has quit IRC07:19
*** thorst has quit IRC07:22
*** vishwana_ is now known as vishwanathj07:26
*** vishwanathj has quit IRC07:34
*** hoonetorg has quit IRC07:38
*** hoonetorg has joined #openstack-telemetry07:50
*** ljxiash_ has joined #openstack-telemetry07:50
*** ljxiash has quit IRC07:53
*** yprokule has quit IRC08:00
*** yprokule has joined #openstack-telemetry08:00
*** vishwanathj has joined #openstack-telemetry08:14
*** thorst has joined #openstack-telemetry08:20
openstackgerritMerged openstack/python-aodhclient: Only install hacking in pep8  https://review.openstack.org/31556808:24
*** jwcroppe has joined #openstack-telemetry08:25
cdentjd__:  no love on your gnocchi suggestion :(08:27
cdent(re monasca)08:27
*** thorst has quit IRC08:27
*** jwcroppe has quit IRC08:30
jd__cdent: I'm so sad08:31
jd__maybe that's not what they wanted to read08:31
cdentheh08:31
*** jmccarthy has quit IRC08:31
*** jmccarthy has joined #openstack-telemetry08:32
cdentEmilienM found another gabbi issue, that I've fixed, but not yet released. it's going to require a slight change to the gnocchi tempests tests. I've cc'd you, jd__, on reviewing the gabbi change08:32
cdentssl tests failing to behave08:32
cdenthttps://github.com/cdent/gabbi/pull/13908:33
*** yassine has joined #openstack-telemetry08:38
*** fwdit has joined #openstack-telemetry08:39
*** nijaba has quit IRC09:05
openstackgerritJulien Danjou proposed openstack/gnocchi: _carbonara: fix race condition in heartbeat stop condition  https://review.openstack.org/32431009:09
*** yprokule has quit IRC09:11
*** nijaba has joined #openstack-telemetry09:11
*** nijaba has quit IRC09:11
*** nijaba has joined #openstack-telemetry09:11
*** hughhalf has quit IRC09:12
jd__cdent: re: I'm so sad – it's actually worst than I thought now that I read the replies lol09:15
cdentyeah09:16
*** yprokule has joined #openstack-telemetry09:16
*** thorst has joined #openstack-telemetry09:25
*** thorst has quit IRC09:32
*** hughhalf has joined #openstack-telemetry09:48
*** vishwanathj has quit IRC09:57
openstackgerritChris Dent proposed openstack/gnocchi: Fix tempest tests that use SSL  https://review.openstack.org/32433710:02
cdentnot sure if that ^ is complete but I guess we'll see10:02
*** vishwanathj has joined #openstack-telemetry10:07
*** cheneydc has quit IRC10:07
*** sbadia has left #openstack-telemetry10:17
openstackgerritMerged openstack/gnocchi: _carbonara: fix race condition in heartbeat stop condition  https://review.openstack.org/32431010:20
*** vishwanathj has quit IRC10:21
*** nadya has joined #openstack-telemetry10:27
*** thorst has joined #openstack-telemetry10:30
*** thorst has quit IRC10:37
*** achatterjee_ has joined #openstack-telemetry10:41
*** achatterjee has quit IRC10:44
*** cdent has quit IRC10:55
zqfanping jd__: dead lock: https://review.openstack.org/323862 & https://review.openstack.org/32431610:55
zqfanany hint? combine the two?10:55
*** nadya has quit IRC11:05
*** ljxiash_ has quit IRC11:16
jd__zqfan: yeah you need to combine11:22
*** thorst has joined #openstack-telemetry11:26
*** julim has joined #openstack-telemetry11:31
zqfanjd__: is it ok to have two changeId in one commit?11:40
*** ljxiash has joined #openstack-telemetry11:40
jd__zqfan: no, but you can add the two cherry-pick ids11:45
openstackgerritMerged openstack/ceilometer: Bump to Nova v2.1  https://review.openstack.org/32421111:50
*** peristeri has joined #openstack-telemetry11:51
zqfanjd__: the combined backport patches: https://review.openstack.org/#/c/324316/11:52
*** cdent has joined #openstack-telemetry11:52
*** pradk has joined #openstack-telemetry11:52
openstackgerritliusheng proposed openstack/aodh: Add indexs of alarm.enabled and alarm.type  https://review.openstack.org/32439011:56
EmilienMcdent: cool, thanks for your help12:00
cdentEmilienM: a pleasure as always12:00
EmilienM:)12:00
*** ljxiash_ has joined #openstack-telemetry12:04
*** ljxiash has quit IRC12:05
*** ljxiash_ has quit IRC12:09
*** ljxiash has joined #openstack-telemetry12:09
*** hoonetorg has quit IRC12:12
*** ljxiash has quit IRC12:14
*** ljxiash has joined #openstack-telemetry12:14
*** ljxiash has quit IRC12:20
*** ljxiash has joined #openstack-telemetry12:21
*** hoonetorg has joined #openstack-telemetry12:27
*** dave-mccowan has joined #openstack-telemetry12:28
*** julim has quit IRC12:41
*** gordc has joined #openstack-telemetry12:46
*** julim has joined #openstack-telemetry13:14
*** nadya has joined #openstack-telemetry13:17
*** liamji has quit IRC13:20
*** diogogmt has joined #openstack-telemetry13:22
openstackgerritIgor Degtiarov proposed openstack/aodh: Add ca_bundle path in ssl request  https://review.openstack.org/32381613:30
*** rbak has quit IRC13:38
*** ljxiash has quit IRC13:39
*** ljxiash has joined #openstack-telemetry13:41
*** KrishR has joined #openstack-telemetry13:41
*** nadya_ has joined #openstack-telemetry13:43
*** nadya has quit IRC13:45
*** ljxiash has quit IRC13:45
*** ljxiash has joined #openstack-telemetry13:46
*** nadya has joined #openstack-telemetry13:47
*** nadya_ has quit IRC13:49
*** ljxiash has quit IRC13:50
*** ljxiash has joined #openstack-telemetry13:51
*** nadya has quit IRC13:51
*** nadya_ has joined #openstack-telemetry13:52
*** ametts has joined #openstack-telemetry13:53
*** rbak has joined #openstack-telemetry13:53
*** ljxiash_ has joined #openstack-telemetry13:54
*** ljxiash has quit IRC13:55
*** jwcroppe has joined #openstack-telemetry14:02
* gordc wonders if we should store all the aggregates together in gnocchi. the io performance seems to degrade pretty hard in ceph.14:04
*** jwcroppe has quit IRC14:04
*** jwcroppe has joined #openstack-telemetry14:04
jmlowegordc: I've turned my metrics way back, my ceph cluster, full metrics puts me some where between 8k -9k ceph ops14:05
jmlowegordc: cluster seems to handle it ok, but it makes me nervous that I'm going to starve my instances14:06
gordcjmlowe: yeah. i'm a ceph beginner... but i think i'm running out of iops.14:06
*** nadya_ has quit IRC14:07
jmlowegordc: any ideas about not getting any consumers now that I seem to have my zookeeper membership straightened out?14:07
gordci should add disclaimer, i only have 2 OSD nodes14:07
jmloweI have 120 osd's14:07
jmloweI've observed as many as 20k ceph ops14:07
gordci have 20... just trying to do some testing14:08
gordcjmlowe: how do you capture ceph ops?14:08
jmloweceph -w14:08
*** nadya has joined #openstack-telemetry14:08
jmlowealso should be in the /var/log/ceph/ceph.log on any of the mons14:09
gordckk. got it.14:09
gordcjmlowe: do you have debug turned on in notificaiton agent?14:09
jmloweI don't know how solid those numbers are but it should give you a decent idea of what's going on in your ceph cluster14:09
jmloweI do14:09
gordcdo you have a log somewhere that says "My subset "...14:11
*** links has quit IRC14:12
gordcor "Members of group"14:12
gordcthis is 5.0.2 correct? 5.0.3 has some group retry logic which should help14:13
jmlowe5.0.214:13
gordchow many agent services do you have? and how many workers do you have on each?14:14
gordcjd__: ^ you have any thoughts on minimising io in gnocchi?14:15
gordci guess it technically does limit how we parallelise stuff if we merge aggregates14:16
*** fwdit has quit IRC14:19
*** diogogmt has quit IRC14:19
*** nadya has quit IRC14:19
*** nadya has joined #openstack-telemetry14:20
jmloweI have workers = 12 for notification14:20
jmloweone agent14:20
*** diogogmt has joined #openstack-telemetry14:21
jmloweI'm just realizing that having fewer queues than workers might be problematic14:21
gordcoh. yeah, there was an issue i saw where workers couldn't join properly when done all at the (same) time.14:22
gordcdo you actually need 12 workers? that seems like a lot of workers unless you have a lot of messages14:22
*** ljxiash has joined #openstack-telemetry14:22
*** ljxiash_ has quit IRC14:23
gordcif you don't want to use 5.0.3, you can try manually deploying workers? ie, multiple agents, one worker. this i find allows you to join group successfully, more consistently14:23
*** liamji has joined #openstack-telemetry14:26
*** ljxiash has quit IRC14:27
openstackgerritgordon chung proposed openstack/python-gnocchiclient: add support for update archive policy  https://review.openstack.org/30736614:29
jmlowegordc: let me try 2 workers14:29
jmlowegordc: 12 does seem excessive14:30
gordcyou may still have issues, side effect of ceilometer using oslo.service unlike other telemetry projects. oslo.service doesn't allow much control over workers14:32
gordcbut yeah, it should be less of an issue?14:32
jmlowesanity check ceilometer-pipe-compute_source:meter_sink-0.sample would be serviced by the notification agent right?14:33
*** ljxiash has joined #openstack-telemetry14:33
jmlowehttp://paste.openstack.org/show/507305/14:35
gordcjmlowe: right14:37
jmlowewell, I still don't have zookeeper connections working it looks like14:39
jmlowehttp://paste.openstack.org/show/507307/14:41
gordcstrange. seems like a tooz issue. not that familiar with it myself14:42
*** david-lyle has joined #openstack-telemetry14:44
*** diogogmt has quit IRC14:44
*** thumpba has joined #openstack-telemetry14:49
jmlowehow long ago was 5.0.3 released, doesn't seem to be packaged yet14:50
jmlowe6.0.0 has retry logic?14:51
jmloweI'm mulling over moving from liberty to mitaka across the board14:53
gordcjmlowe: 6.x has the retry logic. you in theory can run notification agent 6.x with 5.x of other services14:54
gordcthis isn't tested though14:54
jmlowegordc: right, I have enough hardware to spread out, I'd update entire services not just agents14:56
*** thumpba has quit IRC15:00
*** nadya_ has joined #openstack-telemetry15:01
*** thumpba has joined #openstack-telemetry15:01
jmloweI guess in the mean time I should hunt down the tooz devs and see if they have a suggestion?15:01
*** nadya has quit IRC15:01
jmlowewhat happens if you reduce the number of queues, is there some logic to clean them out?15:02
gordcjmlowe: no. i couldn't find a way to have the service seek out queues by matching topic15:03
jmlowegordc: that seems like a difficult problem15:04
gordcif you shrink processing queues, the queues will sit idle until they are consumed or killed15:04
*** diogogmt has joined #openstack-telemetry15:04
*** thumpba has quit IRC15:05
gordci don't know if it's impossible. i think it's just a limitation currently. i think it should be possible to clean out if i add a new option to say where we can write to and read from.15:06
jmloweyou'd have to have a whole thing that examined the existing queues and if the count doesn't match the config drain then delete and then if you have more than one agent if the configs it gets ugly and uglier still if the configs for the agents don't match15:06
openstackgerritVictor Hugo de Araujo Monteiro proposed openstack/gnocchi: Added endpoint type on swift configuration. The swiftclient uses publicURL as default to connect on swift, but has the option to choose the endpoint type the user wants to use, for cases where you rather connect to adminURL or internalURL.  https://review.openstack.org/32460215:06
*** ildikov has quit IRC15:07
jmlowelet me try that again: you'd have to have a whole thing that examined the existing queues and if the count doesn't match the config drain then delete and then if you have more than one agent it gets ugly and uglier still if the configs for the agents don't match15:07
*** patrickeast has quit IRC15:08
*** ildikov has joined #openstack-telemetry15:09
*** patrickeast has joined #openstack-telemetry15:09
gordcyeah, i'd rather not add a new config option to ceilometer but it might be the easier way if a user wants to turn down the queues.15:10
gordcright now it behaves sort of like placement groups in ceph, you're going to lose data if you turn it down15:11
jmlowedistributed systems are fun!15:12
jmlowetooz is at 1.21, my gnocchi is running 1.34 and works15:13
*** nadya has joined #openstack-telemetry15:13
gordcjmlowe: lol "fun".15:14
*** nadya_ has quit IRC15:15
gordcgnocchi doesn't use oslo.service so we can control workers a bit easier/better in it.15:15
*** thumpba has joined #openstack-telemetry15:17
*** jmlowe has quit IRC15:23
*** vinaypotluri has joined #openstack-telemetry15:42
*** tesseract- has quit IRC15:59
*** peristeri has quit IRC16:01
cdentthere we go jd__ : "Has anyone talked with the gnocchi folks? It seems like a good time to. :)"16:01
*** ljxiash has quit IRC16:01
*** rcernin has quit IRC16:01
jd__cdent: finalllly16:06
*** nadya has quit IRC16:06
*** nadya has joined #openstack-telemetry16:10
openstackgerritMerged openstack/python-gnocchiclient: add support for update archive policy  https://review.openstack.org/30736616:10
gordcshould we go open core too?16:11
gordcit seems like the cool thing16:11
*** nadya has quit IRC16:22
*** david-lyle has quit IRC16:24
*** yprokule has quit IRC16:28
*** nadya has joined #openstack-telemetry16:31
*** nadya has quit IRC16:37
*** nadya has joined #openstack-telemetry16:37
*** jmlowe has joined #openstack-telemetry16:38
*** nadya has quit IRC16:41
*** nadya has joined #openstack-telemetry16:42
*** tesseract has joined #openstack-telemetry16:44
*** nadya_ has joined #openstack-telemetry16:48
*** nadya has quit IRC16:49
*** julim has quit IRC16:51
*** tesseract has quit IRC16:51
*** cdent has quit IRC16:55
*** nadya_ has quit IRC16:57
*** liamji has quit IRC17:20
openstackgerritVictor Hugo de Araujo Monteiro proposed openstack/gnocchi: Added endpoint type on swift configuration. The swiftclient uses publicURL as default to connect on swift, but has the option to choose the endpoint type the user wants to use, for cases where you rather connect to adminURL or internalURL.  https://review.openstack.org/32460217:40
*** vinaypotluri has left #openstack-telemetry17:44
*** thumpba has quit IRC17:48
*** thumpba has joined #openstack-telemetry17:54
*** pradk has quit IRC17:56
*** links has joined #openstack-telemetry17:57
*** diogogmt has quit IRC18:03
*** diogogmt has joined #openstack-telemetry18:04
*** david-lyle has joined #openstack-telemetry18:09
*** diogogmt has quit IRC18:10
*** diogogmt has joined #openstack-telemetry18:14
*** pradk has joined #openstack-telemetry18:16
*** julim has joined #openstack-telemetry18:23
openstackgerritgordon chung proposed openstack/gnocchi: cleanup metric delete in ceph  https://review.openstack.org/32475218:25
*** david-lyle has quit IRC18:35
*** peristeri has joined #openstack-telemetry18:37
*** links has quit IRC18:39
*** harlowja has quit IRC18:44
*** achatterjee_ has left #openstack-telemetry18:55
*** nadya has joined #openstack-telemetry18:59
*** yassine has quit IRC18:59
*** thumpba has quit IRC19:09
*** thumpba has joined #openstack-telemetry19:11
*** nadya has quit IRC19:11
*** nadya has joined #openstack-telemetry19:12
*** nadya has quit IRC19:12
*** nadya has joined #openstack-telemetry19:12
*** nadya has quit IRC19:21
*** nadya has joined #openstack-telemetry19:23
*** david-lyle has joined #openstack-telemetry19:24
*** nadya has quit IRC19:28
*** nadya has joined #openstack-telemetry19:32
*** nadya has quit IRC19:37
*** openstackstatus has quit IRC19:39
*** openstack has joined #openstack-telemetry19:43
*** nadya has joined #openstack-telemetry19:49
*** david-lyle has quit IRC19:57
*** thumpba has quit IRC20:12
*** rcernin has joined #openstack-telemetry20:21
*** harlowja has joined #openstack-telemetry20:24
*** sheel has quit IRC20:35
*** nadya has quit IRC20:35
openstackgerritgordon chung proposed openstack/gnocchi: cleanup metric delete in ceph  https://review.openstack.org/32475220:42
*** julim has quit IRC20:53
*** thorst has quit IRC21:03
*** thorst has joined #openstack-telemetry21:05
*** thorst_ has joined #openstack-telemetry21:08
*** thorst has quit IRC21:10
*** thorst_ has quit IRC21:12
*** thorst has joined #openstack-telemetry21:30
*** thorst has quit IRC21:35
*** rcernin_ has joined #openstack-telemetry21:38
*** peristeri has quit IRC21:45
openstackgerritgordon chung proposed openstack/ceilometer: WIP more refactoring  https://review.openstack.org/32489521:53
openstackgerritgordon chung proposed openstack/gnocchi: cleanup metric delete in ceph  https://review.openstack.org/32475221:54
*** claudiub has joined #openstack-telemetry21:59
*** gordc has quit IRC22:02
*** ametts has quit IRC22:03
*** rcernin_ has quit IRC22:06
*** openstackgerrit has quit IRC22:19
*** openstackgerrit has joined #openstack-telemetry22:20
*** dave-mccowan has quit IRC22:30
*** dave-mccowan has joined #openstack-telemetry22:48
*** rbak has quit IRC22:49
*** r-daneel has joined #openstack-telemetry23:07
*** rcernin has quit IRC23:10
*** agireud has quit IRC23:19
*** david-lyle has joined #openstack-telemetry23:20
*** agireud has joined #openstack-telemetry23:21
*** rbak has joined #openstack-telemetry23:27
*** KrishR has quit IRC23:34
*** vishwanathj has joined #openstack-telemetry23:49
openstackgerritgordon chung proposed openstack/ceilometer: refactor service to be less pipeline dependent  https://review.openstack.org/32361523:58

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