15:00:19 #startmeeting ceilometer 15:00:20 Meeting started Thu Sep 19 15:00:19 2013 UTC and is due to finish in 60 minutes. The chair is jd__. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:22 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:24 The meeting name has been set to 'ceilometer' 15:00:34 o/ 15:00:37 hello everybody 15:00:44 o/ 15:00:53 o/ 15:01:04 o/ 15:01:05 o/ 15:01:25 hello 15:02:08 o/ 15:02:16 #topic Review Havana RC1 milestone 15:02:24 #link https://launchpad.net/ceilometer/+milestone/havana-rc1 15:02:59 so we still got our blueprints unmerged, that's bad 15:03:08 o/ 15:03:16 though I think we're coming to an end fortuantely 15:03:16 * eglynn busily working thru' review comments on https://review.openstack.org/44751 15:03:31 ... will have another iteration up shortly 15:03:31 the release manager virtual deadline is tomorrow 15:03:53 so nose on the grindstone like eglynn would say 15:04:14 ... grinding like crazy :) 15:04:25 same for me :) 15:04:51 I also went through our bug list to mark the ones important for RC1 15:05:02 so the list on the page should be up to date 15:05:17 don't forget to add this as a milestone if you want this to be fixed for rc1 when reporting new bugs 15:05:30 and also feel free to review the list yourself if you think I missed something important :) 15:05:48 there's also a couple of bugs that you can see at https://launchpad.net/ceilometer/+milestone/havana-rc1 that don't have assignees 15:05:48 I need to chase dprince on https://bugs.launchpad.net/ceilometer/+bug/1224666 15:05:49 Launchpad bug 1224666 in ceilometer "alarm_history_project_fkey Constraint fails on el6 MySQL" [High,Triaged] 15:06:07 so volunteers welcome 15:06:08 (seems to me that issue should have gone away since https://review.openstack.org/45306 landed) 15:06:26 eglynn: ack, I guess you can test it? 15:06:36 jd__: yep, I've done so 15:06:37 end of my speech for RC1, questions? 15:06:48 (will confirm with dprince in any case ...) 15:08:28 #topic Release python-ceilometerclient? 15:08:37 I don't think we need that yet 15:08:49 we may want to release once sileht changes on alarming got merged 15:08:55 s/got/get/ 15:09:04 yep wait on https://review.openstack.org/#/c/46707/ 15:09:14 exactly 15:09:16 ya 15:10:05 #topic Open discussion 15:10:23 I'm out of topic :) 15:10:33 wsme support return code \o/ 15:10:42 lol 15:10:47 heh 15:10:53 204s, 201s, the joy of it! :) 15:11:27 Looks like the tests got resolved. Thanks for that. =] 15:11:39 guys, do you have any docs related to Autoscaling integration with heat? 15:11:59 nadya, would like to read that also 15:12:09 not that I'm aware of, asalkeld may have something on the wiki 15:12:21 I would hope it doesn't need any documentation ;) 15:12:23 (IIRc he's back from vacation on Monday) 15:12:48 all the changes will be in their part? 15:13:05 nadya, a example for heat template: https://github.com/openstack/heat-templates/blob/master/cfn/F17/AutoScalingCeilometer.yaml 15:13:25 sileht, thanks! 15:13:36 nadya, all you needs have been merged 15:13:37 just in case you missed the email, we're pushing hard for more notification support in openstack projects. 15:13:45 #link http://www.sandywalsh.com/2013/09/notification-usage-in-openstack-report.html 15:14:01 beat the drum "more events!" :) 15:14:19 btw, PTL elections will start tomorrow https://wiki.openstack.org/wiki/PTL_Elections_Fall_2013 and I think I heard ttx needed officials to help, if some of you are interested 15:14:21 sandywalsh, cool reports 15:14:42 hi now i do IPMI implementation for monitoring physical devices. What data you prefer to collect there? 15:15:29 lexx: everything? 15:15:48 lexx: do you mean monitoring over IPMI itself? 15:15:55 jd__: agreed. 15:16:21 lexx: (... does that expose everything we currently gather from the hypervisor in the libvirt case?) 15:16:48 eglynn: I think it's remotely read, no? lexx? 15:17:01 yes 15:17:11 lexx: are you building off this bp: https://blueprints.launchpad.net/ceilometer/+spec/ipmi-inspector-for-monitoring-physical-devices ? 15:17:24 yes 15:17:33 lexx: cool cool 15:17:38 * eglynn reads ... 15:17:49 we already have snmp implementation 15:18:34 but i should know which data should collected by IPMI implementation 15:19:37 whatever you can get? 15:19:54 yeah that was kinda my original question 15:19:55 lexx, what are the advantages over snmp Hardware agent? 15:20:08 i.e. what is available for collection over IPMI? 15:20:37 speed Fan, voltage, temperature 15:20:38 https://pypi.python.org/pypi/pyipmi/0.8.0 15:20:52 power on/off would be great (especially for heat) 15:21:12 lexx, oh cool 15:21:19 bios change for security 15:21:52 os failure 15:22:12 sounds like a combination of events and meters 15:22:31 one more question from me. I sae several hbase-related bps on launchpad. they are approved but not started since May. Is there any activity about that anythere :)? 15:22:51 sandywalsh, that is great, that is exactly what we need in Tuskar ans TripleO 15:24:01 sandywalsh, also health monitor of host in Horizon will be great 15:24:12 nadya: are you volunteering? :) 15:24:31 gordc, thinking about this :) 15:25:00 nadya, I just implemented duplicating resource metadata in the meter collection from the HBase driver. That's the only activity on it that I'm aware of. 15:25:07 nadya: i'd say hbase is probably the least actively supported of the backends we have... i don't think anyone will fight over it with you. 15:25:18 sandywalsh, I didn't find any documentation for pypmi. I prefer to use a ipmitool in subprocess 15:25:44 nadya, But, I'm not an HBase maintainer, just wanted to address the bug I was working on for every driver. 15:26:07 lsmola, yeah, I can see ironic using that too 15:26:13 gordc, ok, I see. Thank you for answer 15:26:44 lexx, there seems to be several implementations, which is good 15:26:54 thomasm, thanks for info! 15:26:59 nadya, sure thing 15:27:41 nadya: definitely cool if you can take over HBase maintenance :) 15:27:42 sandywalsh, cool, I think tripleo is still using Nova-baremetal, but we should switch to to Ironic soon 15:28:20 lexx, putting you as a depency for Tuskar integration 15:28:59 lexx, will put it also in Horizon, once i figure out where :-) though it will be something with statistics of Hosts 15:29:50 jd__, I love HBase :) But I'm new in ceilometer so I need to dive into it first 15:30:00 nadya: sure :) 15:30:58 lsmola, no in not. But statistics of hosts already collect by SNMP 15:32:13 lexx, if you mean the Hardware agent, it is collecting only few of them 15:32:51 yes 15:32:58 lexx, the one named by you and sandywalsh are not there as far a I know 15:33:13 lexx, would be cool to add it 15:33:40 ok 15:33:41 lexx, for tuskar and a tripleo, we are monitoring only the hardware, so this will be very welcome 15:34:51 lexx, is the some way to monitor e.g. disk health? 15:35:14 lexx, in general, we are finding a way how to measure the health of the Hardware 15:35:44 lexx, so any failures will be usable i guess 15:36:42 okey 15:37:04 lexx, cool 15:37:22 lexx, thank you very much 15:37:32 no problem) 15:38:15 #endmeeting