| *** rbak has quit IRC | 00:00 | |
| *** rbak has joined #openstack-ceilometer | 00:09 | |
| *** rbak has quit IRC | 00:09 | |
| *** rbak has joined #openstack-ceilometer | 00:16 | |
| *** rbak has quit IRC | 00:20 | |
| *** rbak has joined #openstack-ceilometer | 00:21 | |
| *** rbak has quit IRC | 00:21 | |
| *** _cjones_ has quit IRC | 00:39 | |
| *** thorst has quit IRC | 01:02 | |
| *** rbak has joined #openstack-ceilometer | 01:07 | |
| *** ankita_w_ has quit IRC | 01:12 | |
| *** julim has joined #openstack-ceilometer | 01:24 | |
| *** rbak has quit IRC | 01:26 | |
| *** ankita_wagh has joined #openstack-ceilometer | 01:48 | |
| *** liusheng_ has joined #openstack-ceilometer | 01:49 | |
| *** rjaiswal has quit IRC | 02:10 | |
| *** liusheng_ has quit IRC | 02:15 | |
| *** ankita_wagh has quit IRC | 02:22 | |
| *** tiny-hands has quit IRC | 02:59 | |
| *** tiny-hands has joined #openstack-ceilometer | 03:16 | |
| *** rodrigods has quit IRC | 03:16 | |
| *** rodrigods has joined #openstack-ceilometer | 03:19 | |
| *** annasort has quit IRC | 03:24 | |
| *** dikonoor has joined #openstack-ceilometer | 03:26 | |
| *** raginbajin has quit IRC | 03:27 | |
| *** raginbajin has joined #openstack-ceilometer | 03:29 | |
| *** logan2 has quit IRC | 03:31 | |
| *** ankita_wagh has joined #openstack-ceilometer | 03:45 | |
| *** haigang has joined #openstack-ceilometer | 03:47 | |
| *** shardy_ has joined #openstack-ceilometer | 03:47 | |
| *** shardy has quit IRC | 03:49 | |
| *** shardy_ has quit IRC | 03:52 | |
| *** shardy has joined #openstack-ceilometer | 03:53 | |
| *** ankita_wagh has quit IRC | 03:55 | |
| *** ankita_wagh has joined #openstack-ceilometer | 03:56 | |
| *** julim has quit IRC | 03:59 | |
| *** raginbajin has quit IRC | 04:10 | |
| *** dikonoor has quit IRC | 04:24 | |
| *** raginbajin has joined #openstack-ceilometer | 04:28 | |
| *** rbak has joined #openstack-ceilometer | 04:49 | |
| *** yprokule has joined #openstack-ceilometer | 04:54 | |
| *** rjaiswal has joined #openstack-ceilometer | 05:00 | |
| *** rbak has quit IRC | 05:00 | |
| *** boris-42 has quit IRC | 05:20 | |
| *** nadya has joined #openstack-ceilometer | 05:24 | |
| *** nadya has quit IRC | 05:24 | |
| *** ankita_w_ has joined #openstack-ceilometer | 05:41 | |
| *** ankita_wagh has quit IRC | 05:45 | |
| *** afazkas has joined #openstack-ceilometer | 06:06 | |
| *** lsmola has joined #openstack-ceilometer | 06:21 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/aodh: Imported Translations from Transifex https://review.openstack.org/214027 | 06:21 |
|---|---|---|
| *** Ruetobas has joined #openstack-ceilometer | 06:27 | |
| *** Qlawy has joined #openstack-ceilometer | 06:31 | |
| *** Ruetobas has quit IRC | 06:32 | |
| *** Ruetobas has joined #openstack-ceilometer | 06:37 | |
| *** Ruetobas has quit IRC | 06:45 | |
| *** nadya has joined #openstack-ceilometer | 06:45 | |
| *** Ruetobas has joined #openstack-ceilometer | 06:50 | |
| *** Ruetobas has quit IRC | 06:55 | |
| *** Ruetobas has joined #openstack-ceilometer | 06:55 | |
| *** sileht has joined #openstack-ceilometer | 07:15 | |
| openstackgerrit | Ren Qiaowei proposed openstack/ceilometer: xenapi: support the session to "unix://local" https://review.openstack.org/214483 | 07:16 |
| *** fawadkhaliq has joined #openstack-ceilometer | 07:23 | |
| *** belmoreira has joined #openstack-ceilometer | 07:28 | |
| *** ankita_w_ has quit IRC | 07:41 | |
| openstackgerrit | liusheng proposed openstack/ceilometer: Use start status of coodinator in tooz https://review.openstack.org/201443 | 07:50 |
| *** boris-42 has joined #openstack-ceilometer | 07:58 | |
| *** yassine__ has joined #openstack-ceilometer | 08:09 | |
| *** timbyr_ has joined #openstack-ceilometer | 08:21 | |
| *** exploreshaifali has joined #openstack-ceilometer | 08:24 | |
| *** fawadkhaliq has quit IRC | 08:36 | |
| *** rjaiswal has quit IRC | 08:50 | |
| *** haigang has quit IRC | 08:52 | |
| *** haigang has joined #openstack-ceilometer | 08:52 | |
| *** haigang has quit IRC | 08:52 | |
| *** haigang has joined #openstack-ceilometer | 08:53 | |
| *** haigang has quit IRC | 08:53 | |
| *** haigang has joined #openstack-ceilometer | 08:53 | |
| *** nijaba has quit IRC | 09:00 | |
| *** nadya has quit IRC | 09:27 | |
| *** fawadkhaliq has joined #openstack-ceilometer | 09:28 | |
| *** cdent has joined #openstack-ceilometer | 09:43 | |
| *** giroro_ has joined #openstack-ceilometer | 09:53 | |
| *** Ruetobas has quit IRC | 09:55 | |
| *** exploreshaifali has quit IRC | 09:57 | |
| *** giroro_ has quit IRC | 09:57 | |
| *** Ruetobas has joined #openstack-ceilometer | 09:58 | |
| *** exploreshaifali has joined #openstack-ceilometer | 10:02 | |
| *** belmoreira has quit IRC | 10:08 | |
| *** belmoreira has joined #openstack-ceilometer | 10:09 | |
| *** Ruetobas has quit IRC | 10:14 | |
| *** belmoreira has quit IRC | 10:14 | |
| *** belmoreira has joined #openstack-ceilometer | 10:15 | |
| *** Ruetobas has joined #openstack-ceilometer | 10:21 | |
| *** Ruetobas has quit IRC | 10:25 | |
| *** Ruetobas has joined #openstack-ceilometer | 10:25 | |
| openstackgerrit | Merged openstack/aodh: tox: generate sample config file on default target https://review.openstack.org/210842 | 10:28 |
| *** cdent has quit IRC | 10:32 | |
| _gryf | hi | 10:45 |
| _gryf | I've just started my brand new devstack setup with ceilometer enabled, and noticed that from dashboard i cannot access resource usage | 10:46 |
| _gryf | is it a known issue? | 10:47 |
| *** belmoreira has quit IRC | 10:48 | |
| _gryf | note, that from commandline I can see statistics and meters… | 10:49 |
| *** fawadkhaliq has quit IRC | 10:53 | |
| *** thorst has joined #openstack-ceilometer | 11:09 | |
| *** boris-42 has quit IRC | 11:20 | |
| *** tiny-hands has quit IRC | 11:29 | |
| *** gordc has joined #openstack-ceilometer | 11:33 | |
| *** fawadkhaliq has joined #openstack-ceilometer | 11:35 | |
| *** nadya_ has joined #openstack-ceilometer | 12:03 | |
| *** afazkas has quit IRC | 12:04 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/gnocchi: TEST: gate tests https://review.openstack.org/214580 | 12:07 |
| *** Ruetobas has quit IRC | 12:08 | |
| openstackgerrit | Merged openstack/aodh: Imported Translations from Transifex https://review.openstack.org/214027 | 12:09 |
| *** julim has joined #openstack-ceilometer | 12:10 | |
| *** Ruetobas has joined #openstack-ceilometer | 12:11 | |
| *** Ruetobas has quit IRC | 12:15 | |
| *** Ruetobas has joined #openstack-ceilometer | 12:15 | |
| *** belmoreira has joined #openstack-ceilometer | 12:19 | |
| *** belmoreira has quit IRC | 12:19 | |
| *** cdent has joined #openstack-ceilometer | 12:23 | |
| *** edmondsw has joined #openstack-ceilometer | 12:27 | |
| *** Ruetobas has quit IRC | 12:33 | |
| *** Ruetobas has joined #openstack-ceilometer | 12:36 | |
| *** logan2 has joined #openstack-ceilometer | 12:39 | |
| *** haigang has quit IRC | 12:39 | |
| *** Ruetobas has quit IRC | 12:40 | |
| *** julim has quit IRC | 12:41 | |
| *** julim has joined #openstack-ceilometer | 12:42 | |
| *** Ruetobas has joined #openstack-ceilometer | 12:45 | |
| *** fawadkhaliq has quit IRC | 12:46 | |
| *** exploreshaifali has quit IRC | 12:54 | |
| *** tiny-hands has joined #openstack-ceilometer | 13:01 | |
| openstackgerrit | Ryota MIBU proposed openstack/aodh: Add evaluator for event alarm https://review.openstack.org/200972 | 13:08 |
| *** prad_ has joined #openstack-ceilometer | 13:32 | |
| openstackgerrit | Zhai, Edwin proposed openstack/ceilometer: Add hypervisor inspector sanity check https://review.openstack.org/214612 | 13:42 |
| openstackgerrit | Ryota MIBU proposed openstack/aodh: Add evaluator for event alarm https://review.openstack.org/200972 | 13:43 |
| *** prad_ is now known as pradk | 13:56 | |
| *** rbak has joined #openstack-ceilometer | 14:02 | |
| openstackgerrit | Mehdi Abaakouk (sileht) proposed openstack/ceilometer: integration tests: stop on first failed test https://review.openstack.org/214627 | 14:03 |
| *** kbyrne has quit IRC | 14:07 | |
| *** timbyr_ has quit IRC | 14:07 | |
| *** kbyrne has joined #openstack-ceilometer | 14:11 | |
| *** thumpba has joined #openstack-ceilometer | 14:15 | |
| *** boris-42 has joined #openstack-ceilometer | 14:15 | |
| *** edmondsw has quit IRC | 14:22 | |
| pradk | gordc, so instead of target what if we call it attribute and let it be defined as a list or a string.. that way we can leverage the existing attribute logic from resource id | 14:27 |
| *** edmondsw has joined #openstack-ceilometer | 14:37 | |
| openstackgerrit | Ryota MIBU proposed openstack/aodh: Add listener service for event alarm evaluation https://review.openstack.org/200974 | 14:37 |
| *** arcimboldo has joined #openstack-ceilometer | 14:41 | |
| arcimboldo | hi all, I have a question | 14:41 |
| arcimboldo | the "cpu" metric, what's actually counting? | 14:42 |
| arcimboldo | the cpu time used by the vm or just the wallclock time? | 14:42 |
| arcimboldo | is this chan active? | 14:45 |
| openstackgerrit | Pradeep Kilambi proposed openstack/ceilometer: Add support to compute cpu meters in yaml https://review.openstack.org/210619 | 14:45 |
| cdent | arcimboldo: it's active but people tend to only hear when they are either looking directly at it or are pinged | 14:46 |
| cdent | the 'cpu' metric in nanoseconds of cpu time | 14:46 |
| cdent | it's _not_ wall clock | 14:47 |
| cdent | brb | 14:47 |
| openstackgerrit | Pradeep Kilambi proposed openstack/ceilometer: Add support to compute cpu meters in yaml https://review.openstack.org/210619 | 14:47 |
| arcimboldo | cdent, thnx | 14:47 |
| arcimboldo | cdent, the cpu metric doesn't seems right, then | 14:47 |
| cdent | there's slightly more info here: http://docs.openstack.org/admin-guide-cloud/content/section_telemetry-compute-meters.html | 14:47 |
| gordc | arcimboldo: because it resets? | 14:49 |
| arcimboldo | I have a vm started ~20 hours ago, cpu_util reports an average 10% usage, it's using 8 cores, and cpu says ~130hours | 14:49 |
| arcimboldo | 140 actually | 14:49 |
| arcimboldo | I use cpu.sum/1e+9/60/60 | 14:50 |
| arcimboldo | to convert from ns to hours | 14:50 |
| arcimboldo | it should be ~16 hours instead | 14:50 |
| gordc | pradk: that's kinda what i wanted to avoid. i know you're reusing the attribute code but giving it a list doesn't make sense. | 14:50 |
| arcimboldo | gordc, I don't know if it resets, but it gives results way higher than expected | 14:51 |
| pradk | gordc, i'm not giving it a list | 14:51 |
| pradk | volume: | 14:51 |
| pradk | attribute: payload.metrics[?(@.name='cpu.kernel.percent')].value | 14:51 |
| pradk | scale: 100 | 14:51 |
| pradk | is what i mean | 14:51 |
| pradk | for resource id its the same attribute key with a list | 14:51 |
| pradk | just want to keep the key as attribute | 14:51 |
| pradk | so we dont expose too many different keys to end user and confuse | 14:52 |
| arcimboldo | cdent, what does ceilometer use to get the stats for the cpu? libvirt? | 14:52 |
| pradk | gordc, https://gist.github.com/pkilambi/938aa6e93de9f4386c14 is what i mean | 14:52 |
| pradk | i dont want to call it target in one place and attribute in another | 14:53 |
| gordc | ah i see you coded in new support | 14:53 |
| gordc | pradk: technically they aren't really related but sure. attribute is fine. | 14:54 |
| openstackgerrit | Pradeep Kilambi proposed openstack/ceilometer: Add support to compute cpu meters in yaml https://review.openstack.org/210619 | 14:54 |
| gordc | arcimboldo: yes.. by default libvirt... there's xenapi, hyper-v, etc... if you don't use libvirt | 14:54 |
| arcimboldo | gordc, I'm using libvirt, and virsh cpu-stats <domain> gives a totally different value | 14:55 |
| arcimboldo | it's in seconds, for starter | 14:55 |
| arcimboldo | could it be that ceilometer is getting the *total* value reported by libvirt, and it's then moltiplied by the number of vcpus? | 14:56 |
| gordc | arcimboldo: what version are you running on? | 14:56 |
| arcimboldo | 2015.1.0-0ubuntu1~cloud0 | 14:56 |
| arcimboldo | I'm on ubuntu trusty, using the cloud archive repositories | 14:56 |
| *** logan2 has quit IRC | 15:01 | |
| gordc | arcimboldo: i'd need to debug but basically we're using cputime returned by https://libvirt.org/python.html | 15:01 |
| *** logan2 has joined #openstack-ceilometer | 15:05 | |
| *** Ruetobas has quit IRC | 15:08 | |
| gordc | arcimboldo: it's cputime across all vcpus. so yeah, you'll need to divide cputime by num of cpu | 15:09 |
| *** bdossant has joined #openstack-ceilometer | 15:10 | |
| arcimboldo | somewhere it's multiplied by the number of cpus | 15:10 |
| arcimboldo | the value I get from dom0.info() is 8 times lower than the number I get from statistics.list('cpu') | 15:11 |
| arcimboldo | but https://libvirt.org/html/libvirt-libvirt-domain.html#virDomainInfo is not clear if the cputime is the average or is total | 15:11 |
| arcimboldo | I guess is total | 15:12 |
| *** Ruetobas has joined #openstack-ceilometer | 15:12 | |
| arcimboldo | if I run virsh cpu-stats, the "total cpu_time" is actually what's reported by dom0.info() | 15:13 |
| arcimboldo | then, ceilometer reports a value which is ~8 times bigger, so maybe there is a bug in the code? | 15:13 |
| gordc | arcimboldo: trying to find somewhere it explicitly says it's total time across all vcpus. | 15:15 |
| gordc | arcimboldo: i don't see anywhere we multiply by vcpu... we actually divide cputime by vcpu to calculate cpu_util so it seems like we're expecting it. | 15:15 |
| arcimboldo | gordc, just run virsh cpu-stats and then use the python code, you'll see the number from dom0.info() is the value reported as total cpu time | 15:15 |
| arcimboldo | which makes sense, btw | 15:16 |
| arcimboldo | but wait.... this value is cumulative, right? So I should get the max, not the sum, right? | 15:18 |
| *** Ruetobas has quit IRC | 15:19 | |
| *** rbak has quit IRC | 15:22 | |
| *** rbak has joined #openstack-ceilometer | 15:22 | |
| *** rbak has quit IRC | 15:22 | |
| *** rbak has joined #openstack-ceilometer | 15:22 | |
| *** rbak has quit IRC | 15:23 | |
| *** rbak has joined #openstack-ceilometer | 15:24 | |
| gordc | arcimboldo: cumulative just denotes that the value will always increase (although it's not exactly true for libvirt cputime) | 15:25 |
| *** Ephur has quit IRC | 15:25 | |
| *** rbak has quit IRC | 15:31 | |
| arcimboldo | gordc, why not? You just get the cputime, which is the amount of cpu time used so far | 15:32 |
| arcimboldo | so it will always increase, and if I want to know how much cpu time it used so far, I have to get the latest value | 15:32 |
| arcimboldo | actually, I should do cpu.max - cpu.min, in case I want to know how much cpu-time has been used *during said period* | 15:32 |
| *** Ruetobas has joined #openstack-ceilometer | 15:33 | |
| gordc | arcimboldo: the cputime libvirt provides will reset if you reboot | 15:33 |
| arcimboldo | gordc, if you reboot the machine? | 15:33 |
| arcimboldo | s/machine/server/ | 15:33 |
| arcimboldo | but then it's totally useless! | 15:33 |
| *** rbak has joined #openstack-ceilometer | 15:34 | |
| arcimboldo | is there a delta instead of a sample? | 15:34 |
| gordc | reboot the instance. let me find the bug | 15:34 |
| *** eglynn has joined #openstack-ceilometer | 15:35 | |
| gordc | arcimboldo: the cpu value i get from ceilometer matches the total cpu_time i get usnig virsh | 15:36 |
| *** Ruetobas has quit IRC | 15:37 | |
| *** bdossant has quit IRC | 15:38 | |
| arcimboldo | gordc, yes, I think my issue was that I was using client.statistics('cpu').sum | 15:38 |
| *** nadya_ has quit IRC | 15:39 | |
| gordc | oh. yeah, that's just add all the cpu_time from all the instances. | 15:39 |
| arcimboldo | well, my query was for just one instance, but still, it's wrong using sum | 15:39 |
| arcimboldo | but again: how can I know the total cpu time used by a VM, even if it has been live-migrated or rebooted? | 15:39 |
| *** ahilan has joined #openstack-ceilometer | 15:40 | |
| arcimboldo | If the value was a delta, I could have just summed all the samples from START_TIME to END_TIME | 15:40 |
| arcimboldo | but if I only have the value reported by virsh, I cannot compute the cpu time... | 15:41 |
| arcimboldo | the best solution, at this point, is to get the wallclock from nova and multiply by the cpu_util.avg | 15:41 |
| *** Ruetobas has joined #openstack-ceilometer | 15:43 | |
| *** Ruetobas has quit IRC | 15:43 | |
| gordc | arcimboldo: if we capture delta i don't think it'll be entirely accurate. we poll for cputime so all the time between poll and reboot would not be captured. | 15:43 |
| *** ankita_wagh has joined #openstack-ceilometer | 15:45 | |
| *** adurbin_ has joined #openstack-ceilometer | 15:45 | |
| gordc | arcimboldo: https://bugs.launchpad.net/ceilometer/+bug/1417949 | 15:46 |
| openstack | Launchpad bug 1417949 in Ceilometer "CPU samples reset to 0 for shutdown instances" [Undecided,Triaged] | 15:46 |
| gordc | i vaguely remember asking nova dev about it and didn't reach a solution | 15:47 |
| arcimboldo | You should add the last value to all following samples... | 15:48 |
| *** rjaiswal has joined #openstack-ceilometer | 15:48 | |
| arcimboldo | but it's not feasable | 15:48 |
| gordc | arcimboldo: right. again, it wouldn't be completely accurate... and it'd be a bit of a performance hit doing maths each time. i don't have a great answer tbh (not a libvirt expert myself) | 15:49 |
| *** rbak has quit IRC | 15:53 | |
| rjaiswal | gordc: put some comments in https://review.openstack.org/#/c/208068/ , wanted to chat about it | 15:53 |
| *** rbak has joined #openstack-ceilometer | 15:56 | |
| gordc | rjaiswal: sure thing. | 15:56 |
| gordc | rjaiswal: regarding the partitioning? | 15:58 |
| rjaiswal | gordc: yes | 15:58 |
| gordc | right, so the issue is the reload_pipeline method is also used for partitioning. if you look at _refresh_agent, that gets called anytime an agent joins/dies. | 16:00 |
| *** _cjones_ has joined #openstack-ceilometer | 16:00 | |
| gordc | it uses reload_pipeline to recalculate all teh queues it needs to write to. | 16:00 |
| rjaiswal | ok | 16:00 |
| rjaiswal | i was thinking that enabling the event pipeline reload would ensure that only the event notifiers would get re-built | 16:01 |
| gordc | if we have partitioning on but the refresh_pipeline_cfg off, it will never recalculate the target queues. | 16:01 |
| gordc | rjaiswal: understood. maybe we can add a param which forces them to be run if it's called by workload partitioning? | 16:03 |
| rjaiswal | gordc: looking at the code now | 16:03 |
| rjaiswal | gordc: ah, got it | 16:05 |
| *** exploreshaifali has joined #openstack-ceilometer | 16:06 | |
| *** eglynn has quit IRC | 16:06 | |
| gordc | rjaiswal: cool cool | 16:06 |
| *** fawadkhaliq has joined #openstack-ceilometer | 16:08 | |
| rjaiswal | thanks gordc for the pointer, i wasnt aware of that change to refresh_agent | 16:10 |
| rjaiswal | shud have looked at the code | 16:11 |
| rjaiswal | gordc: i will add a param and submit a patch | 16:11 |
| gordc | rjaiswal: no worries. i probably should've promoted that change a bit more. | 16:11 |
| *** yprokule has quit IRC | 16:17 | |
| *** arcimboldo has quit IRC | 16:23 | |
| *** ankita_wagh has quit IRC | 16:28 | |
| *** ankita_wagh has joined #openstack-ceilometer | 16:31 | |
| *** thumpba has quit IRC | 16:32 | |
| *** nadya has joined #openstack-ceilometer | 16:39 | |
| *** ankita_wagh has quit IRC | 16:42 | |
| *** nadya has quit IRC | 16:44 | |
| openstackgerrit | Pradeep Kilambi proposed openstack/ceilometer: Add support to compute cpu meters in yaml https://review.openstack.org/210619 | 16:49 |
| openstackgerrit | Rohit Jaiswal proposed openstack/ceilometer: Adds support for dynamic event pipeline https://review.openstack.org/208068 | 16:52 |
| cdent | gordc: we heard anything about summit proposals yet? | 16:57 |
| openstackgerrit | Merged openstack/ceilometer: Use start status of coodinator in tooz https://review.openstack.org/201443 | 16:58 |
| *** ankita_wagh has joined #openstack-ceilometer | 17:01 | |
| *** rbak has quit IRC | 17:07 | |
| *** jmatthew_ has joined #openstack-ceilometer | 17:07 | |
| rjaiswal | gordc: request you to triage:https://bugs.launchpad.net/ceilometer/+bug/1486660 and https://bugs.launchpad.net/ceilometermiddleware/+bug/1485711 | 17:09 |
| openstack | Launchpad bug 1486660 in Ceilometer "Sample not requeued or ack'ed on sample publishing failure " [Undecided,New] - Assigned to Rohit Jaiswal (rohit-jaiswal-3) | 17:09 |
| openstack | Launchpad bug 1485711 in ceilometermiddleware "ceilomiddleware.swift ignore the swift_source in environ" [Undecided,New] - Assigned to Rohit Jaiswal (rohit-jaiswal-3) | 17:09 |
| *** rbak has joined #openstack-ceilometer | 17:10 | |
| *** rbak has quit IRC | 17:10 | |
| *** yassine__ has quit IRC | 17:10 | |
| *** rbak has joined #openstack-ceilometer | 17:19 | |
| openstackgerrit | Pradeep Kilambi proposed openstack/ceilometer: Add support to compute cpu meters in yaml https://review.openstack.org/210619 | 17:20 |
| *** fawadk has joined #openstack-ceilometer | 17:20 | |
| gordc | cdent: i haven't heard anything about speaker sessions. i assume it'll be an email. | 17:21 |
| cdent | ack | 17:21 |
| *** fawadkhaliq has quit IRC | 17:21 | |
| gordc | cdent: just reminded me to ask my manager for travel approval. | 17:23 |
| * cdent is gordc's pa | 17:24 | |
| gordc | i'll get you to sign my permission form. | 17:24 |
| cdent | sorry, I was speaking british: PA, personal assistant, secretary, that sort of thing | 17:26 |
| cdent | but you know, if you need someone to sign your permission slip and put a sandwich in your lunch box, that's cool | 17:26 |
| gordc | sandwich? try harder! i demand caviar. | 17:27 |
| cdent | beluga, of course | 17:27 |
| gordc | the secretary part is a little more useful though... let's go with that service. | 17:28 |
| * gordc will google tell me i'm an idiot if i google "beluga caviar" | 17:28 | |
| gordc | oh... there's a beluga fish... i was really scared what result i'd get with "beluga caviar" | 17:30 |
| cdent | I don't think you've seen enough bond movies | 17:30 |
| gordc | cdent: i think it's more that my diet consists of "anything under 10$" | 17:31 |
| *** rbak has quit IRC | 17:31 | |
| gordc | why? is it referenced a lot in bond films? | 17:31 |
| cdent | I think so, usually when he's trying to impress the ladies: he'll order the most expensive champagne and most expensive caviar and their eyebrows will raise | 17:32 |
| * gordc takes notes | 17:33 | |
| *** rbak has joined #openstack-ceilometer | 17:35 | |
| EmilienM | gordc: don't touch beluga, never | 17:36 |
| gordc | EmilienM: whale or fish? or both? | 17:37 |
| cdent | ethics | 17:37 |
| EmilienM | gordc: whale | 17:37 |
| cdent | gordc: that reminds me of another movie for you | 17:38 |
| cdent | millers crossing | 17:38 |
| *** exploreshaifali has quit IRC | 17:38 | |
| *** rbak has quit IRC | 17:39 | |
| cdent | it's not whale though, it's sturgeon | 17:39 |
| gordc | cdent: big coen bros fan? | 17:43 |
| cdent | they rarely fail to entertain | 17:43 |
| cdent | and there's a lot of talk about efffics | 17:44 |
| gordc | will watch... when i just watched the trailer i thought it was the movie Macaulay Culkin watches from home alone... then i realised it was the same year. | 17:44 |
| gordc | EmilienM: you see beluga in quebec? | 17:45 |
| EmilienM | gordc: in Tadoussac, 2h30 from here | 17:45 |
| EmilienM | the only whales present all year in Saint Laurent | 17:45 |
| gordc | EmilienM: you should catch one and build a really big fish tank. | 17:46 |
| gordc | catch two so they have friends. | 17:46 |
| *** eglynn has joined #openstack-ceilometer | 17:49 | |
| cdent | gordc: you are full of big ideas, have you considered a career in politics | 17:49 |
| EmilienM | gordc: I'll go in jail if I do that | 17:50 |
| EmilienM | they are protected and that's good | 17:50 |
| gordc | cdent: i have. i need to form my own party. something that covers far-right and far-left but not the middle. | 17:51 |
| cdent | I think, based on "les extremes se touchent" you just declared yourself a fascist | 17:52 |
| gordc | cdent: hahahah! done! | 17:53 |
| gordc | we vote in a few months in Canada. i'll keep you posted. | 17:53 |
| gordc | EmilienM: you'll get deported. | 17:53 |
| gordc | i'll take care of your apartment | 17:54 |
| *** afazkas has joined #openstack-ceilometer | 17:54 | |
| cdent | thanks | 17:55 |
| * cdent has dinner | 18:01 | |
| gordc | rjaiswal: i commented on ceilometermiddleware bug... not quite sure about details | 18:03 |
| pradk | gordc, Do you know if anyone covering ceilometer for the ops meetup? was just curious | 18:03 |
| pradk | donno if they even have any topics around it | 18:04 |
| gordc | pradk: i don't believe so. last i checked, there wasn't anything ceilometer related, a lot of it seemed to be about external monitoring tools (sensu, nagios, etc...) | 18:04 |
| pradk | ah ok | 18:05 |
| gordc | pradk: https://etherpad.openstack.org/p/PAO-ops-meetup this is the only etherpad i'm aware of | 18:07 |
| *** nadya has joined #openstack-ceilometer | 18:09 | |
| *** nadya has quit IRC | 18:10 | |
| *** rbak has joined #openstack-ceilometer | 18:20 | |
| *** spzala has joined #openstack-ceilometer | 18:22 | |
| rjaiswal | thanks gordc, will look at the code to confirm that | 18:22 |
| *** rbak has quit IRC | 18:28 | |
| *** rbak has joined #openstack-ceilometer | 18:29 | |
| *** rbak has quit IRC | 18:31 | |
| *** rbak has joined #openstack-ceilometer | 18:31 | |
| *** rbak has quit IRC | 18:33 | |
| *** jmatthew_ has quit IRC | 18:41 | |
| *** exploreshaifali has joined #openstack-ceilometer | 18:42 | |
| *** fawadk has quit IRC | 19:04 | |
| *** nadya has joined #openstack-ceilometer | 19:11 | |
| *** cdent has quit IRC | 19:11 | |
| *** hhuang has quit IRC | 19:13 | |
| *** afazkas has quit IRC | 19:13 | |
| *** afazekas has joined #openstack-ceilometer | 19:14 | |
| *** nadya has quit IRC | 19:16 | |
| pradk | gordc, any idea why the tempest tests could be failing https://review.openstack.org/#/c/210619/ .. we got this when the meters.yaml was missing in devstack | 19:28 |
| *** hhuang has joined #openstack-ceilometer | 19:29 | |
| gordc | pradk: i don't know why but the ceilometer tempest tests are running slow as hell | 19:37 |
| *** shardy_ has joined #openstack-ceilometer | 19:41 | |
| *** shardy has quit IRC | 19:42 | |
| pradk | gordc, hmm ok.. the error looks like its because the sample wasnt added to the db on time | 19:43 |
| pradk | we use to get this error when the yaml was missing | 19:43 |
| pradk | i guess i'll recheck and see | 19:43 |
| *** nadya has joined #openstack-ceilometer | 19:44 | |
| gordc | it's image.size meter. nothing to do you patch | 19:44 |
| gordc | if you look at the tests. they are all running super slow. | 19:44 |
| pradk | yea | 19:44 |
| gordc | even the non-telemetry ones... ours is the longest because we depend on actions of others. | 19:44 |
| *** afazekas has quit IRC | 19:46 | |
| *** nadya has quit IRC | 19:46 | |
| *** shardy_ has quit IRC | 19:46 | |
| *** eglynn has quit IRC | 19:46 | |
| *** shardy has joined #openstack-ceilometer | 19:47 | |
| *** belmoreira has joined #openstack-ceilometer | 19:50 | |
| *** rbak has joined #openstack-ceilometer | 19:51 | |
| gordc | pradk: i think this is the begnning of the end of cycle gate implosion... | 19:52 |
| pradk | yea had to start sometime, thought we still had a week or so..guess not | 19:52 |
| gordc | i honestly don't think any of our tempest tests will pass if it gets worse. | 19:53 |
| *** alejandrito has joined #openstack-ceilometer | 19:59 | |
| *** alejandrito_ has joined #openstack-ceilometer | 20:00 | |
| *** alejandrito has quit IRC | 20:00 | |
| gordc | pradk: i have no idea how a test took 2 seconds on one check[1], and 120s on another check | 20:09 |
| gordc | [1] http://logs.openstack.org/68/208068/8/check/gate-tempest-dsvm-postgres-full/65b1830/console.html#_2015-08-19_18_18_18_490 | 20:09 |
| gordc | [2] http://logs.openstack.org/19/210619/12/check/gate-tempest-dsvm-postgres-full/79c2a07/console.html#_2015-08-19_18_56_37_921 | 20:10 |
| pradk | weird | 20:10 |
| gordc | roughly same time, test has nothing to do with the middleware or storage driver changes... | 20:11 |
| gordc | s/test/patch/ | 20:11 |
| gordc | they both seem to be on the same test node too | 20:12 |
| *** rbak has quit IRC | 20:12 | |
| pradk | yea i'm about to bring up my devstack instance with tempest, will see if its something i can reproduce locally.. i doubt it though | 20:13 |
| pradk | sounds like its perhaps load on the node or some process hanging | 20:13 |
| gordc | possibly? i don't know what to hope for. if it's a gate thing, it's not our fault but we'll have a hell of a time committing... if it is our fault, we have to find a bug and fix it but we'll still have a good gate... | 20:15 |
| gordc | seems like we're screwed either way. | 20:15 |
| *** rbak has joined #openstack-ceilometer | 20:15 | |
| pradk | gordc, could be increase the timeout or whatever and see if it help? has not been added to the database within 196 seconds ..where would that limit be set? | 20:17 |
| gordc | pradk: we could. but if we haven't received a notification in 3mins, something must be up | 20:18 |
| *** alejandrito_ has quit IRC | 20:18 | |
| *** _cjones_ has quit IRC | 20:21 | |
| *** r-daneel has joined #openstack-ceilometer | 20:24 | |
| *** julim has quit IRC | 20:25 | |
| *** spzala has left #openstack-ceilometer | 20:28 | |
| *** rbak has quit IRC | 20:29 | |
| *** rbak has joined #openstack-ceilometer | 20:33 | |
| *** exploreshaifali has quit IRC | 20:41 | |
| *** eglynn has joined #openstack-ceilometer | 20:42 | |
| *** belmoreira has quit IRC | 20:44 | |
| pradk | gordc, you see my response to your why we need multi comment in profiler one? | 20:46 |
| gordc | pradk: nope... looking now | 20:46 |
| pradk | thx | 20:46 |
| *** nadya has joined #openstack-ceilometer | 20:46 | |
| *** nadya has quit IRC | 20:51 | |
| *** Ephur has joined #openstack-ceilometer | 20:51 | |
| gordc | pradk: oh... i wonder if we can name 'multi' something better. seems very hacky | 20:52 |
| gordc | pradk: i still don't understand why we need to have changes in meter.notifications.py | 20:53 |
| pradk | gordc, you mean the volume change? | 20:53 |
| gordc | yeah. normalise_as_list should return exactly that. | 20:54 |
| pradk | gordc, so when i tested with profiler tests, i was getting back 1 instead of [1] reason being the parse_fields method return the value as is if its of type int as we dont lookup any jsonpath's in that case | 20:55 |
| *** Ephur has quit IRC | 20:56 | |
| pradk | gordc, in normalise_as_list we return whatever values = d.parse_fields(value, body, True) gives if it doesnt fit either conditions below | 20:57 |
| pradk | gordc, perhaps we could convert to list in this method instead | 20:57 |
| pradk | this could also handle the case where we have multiple meter with volume =1 | 20:59 |
| pradk | regarding renaming multi, | 20:59 |
| gordc | oh. ok. i see the int clause in parse_fields | 21:00 |
| gordc | yeah, we should put it in noramlise_as_list method | 21:00 |
| pradk | i'm open to it.. multi fits most cases, perhaps not this one | 21:01 |
| pradk | sure i can move it to that method | 21:01 |
| gordc | re: rename multi. if we don't, we should at least explain put 'multi' if meter is not a static text | 21:02 |
| pradk | instead of multi what if we call it query or lookup or search ? search: ['name'] | 21:05 |
| pradk | basically what it would mean is we search for that field in the jsonpath | 21:06 |
| pradk | just thinking out loud | 21:06 |
| *** Ephur has joined #openstack-ceilometer | 21:07 | |
| gordc | hmm... i like the names. | 21:08 |
| gordc | the only thing is i think people will wonder why we don't need for anything other than scenarios where name is a lookup | 21:08 |
| pradk | yea i guess thats the case with any name if we dont document it | 21:11 |
| pradk | so we could leave it multi or change the logic to differentiate between payload.name vs actual static name and do the lookup | 21:12 |
| *** _cjones_ has joined #openstack-ceilometer | 21:12 | |
| pradk | which will be tricky as both are strings | 21:12 |
| gordc | ... lookup/search are fine... i'm leaning more towards lookup | 21:13 |
| pradk | cool i'll go with lookup | 21:21 |
| *** shardy has quit IRC | 21:31 | |
| *** thorst has quit IRC | 21:32 | |
| openstackgerrit | Pradeep Kilambi proposed openstack/ceilometer: Move profiler meters to yaml https://review.openstack.org/213832 | 21:33 |
| *** rjaiswal has quit IRC | 21:40 | |
| *** eglynn has quit IRC | 21:43 | |
| openstackgerrit | gordon chung proposed openstack/ceilometer: add flexible grouping key https://review.openstack.org/204140 | 21:49 |
| openstackgerrit | gordon chung proposed openstack/ceilometer: allow configurable pipeline partitioning https://review.openstack.org/214813 | 21:49 |
| *** gordc has quit IRC | 21:52 | |
| *** rbak has quit IRC | 22:00 | |
| *** rbak has joined #openstack-ceilometer | 22:01 | |
| *** rbak has quit IRC | 22:05 | |
| *** tiny-hands has quit IRC | 22:07 | |
| *** eglynn has joined #openstack-ceilometer | 22:10 | |
| *** ahilan has quit IRC | 22:17 | |
| *** pradk has quit IRC | 22:23 | |
| *** rbak has joined #openstack-ceilometer | 22:25 | |
| *** rbak has quit IRC | 22:30 | |
| *** rbak has joined #openstack-ceilometer | 22:32 | |
| *** rbak has quit IRC | 22:33 | |
| *** rbak has joined #openstack-ceilometer | 22:33 | |
| *** rbak has quit IRC | 22:37 | |
| *** rbak has joined #openstack-ceilometer | 22:38 | |
| *** rbak has quit IRC | 22:48 | |
| *** r-daneel has quit IRC | 22:49 | |
| *** rbak has joined #openstack-ceilometer | 22:51 | |
| openstackgerrit | Merged openstack/ceilometer: Change and move the workers options to corresponding service section https://review.openstack.org/209333 | 22:52 |
| *** rbak has quit IRC | 23:01 | |
| *** rbak has joined #openstack-ceilometer | 23:28 | |
| *** rbak has quit IRC | 23:33 | |
| *** ankita_wagh has quit IRC | 23:46 | |
| *** ankita_wagh has joined #openstack-ceilometer | 23:46 | |
| *** ankita_w_ has joined #openstack-ceilometer | 23:47 | |
| *** ankita_wagh has quit IRC | 23:47 | |
| *** ankita_w_ has quit IRC | 23:51 | |
| *** edmondsw has quit IRC | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!