20:01:05 #startmeeting TripleO 20:01:07 Meeting started Mon Aug 26 20:01:05 2013 UTC and is due to finish in 60 minutes. The chair is SpamapS. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:01:08 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 20:01:10 The meeting name has been set to 'tripleo' 20:01:35 #topic Roll Call! 20:01:40 o/ 20:01:43 yo 20:02:09 cue crickets 20:02:12 * SpamapS wonders just how much of TripleO is off to burn the man :) 20:02:28 lol, I alos think its an English holiday 20:02:54 I do believe if we have but 2 active participants, I may declare this meeting lacking in quorum. 20:03:10 I seconded to motion 20:03:22 I third that 20:03:24 s/to/the/ 20:03:37 no say it ain 20:03:47 so say it ain't so 20:03:53 oh crap 20:04:08 ok lets meet :) 20:04:25 o/ 20:04:38 #topic bugs 20:04:52 https://bugs.launchpad.net/tripleo/ 20:05:10 https://bugs.launchpad.net/diskimage-builder/ 20:05:16 https://bugs.launchpad.net/os-refresh-config/ 20:05:22 https://bugs.launchpad.net/os-apply-config/ 20:05:31 https://bugs.launchpad.net/os-collect-config/ 20:06:15 https://bugs.launchpad.net/tripleo/+bug/1213967 seemed rather nasty 20:06:17 Launchpad bug 1213967 in nova "Seed VM filtering out compute resources" [Critical,In progress] 20:07:03 SpamapS: that was because of me 20:07:15 and the lack of tests in the baremetal driver 20:07:39 are we just waiting on lifeless to respond to reviews? 20:08:14 for that bug I don't think so the nova patch was reverted 20:08:52 so do we just need to mark the bug as fix committed manually in nova, and fix released in tripleo? 20:10:00 jog0: ^^ ? 20:10:19 SpamapS: yes 20:10:27 https://review.openstack.org/#/c/43235/ 20:10:44 although a related bug should be opened in nova for lack of tests in barematal driver 20:10:46 Cool. The two criticals assigned to me should be addressed/closed/split/something by end of this week. I've been distracted by Heat issues the last week or so. 20:11:09 jog0: sounds like an action for you. :) 20:11:26 (and agreed) 20:11:33 ci currently has a baremetal test bug open 20:11:40 probably low priority though, as it is an ironic release away from being deprecated :) 20:11:58 #link https://bugs.launchpad.net/openstack-ci/+bug/1082795 20:11:59 Launchpad bug 1082795 in openstack-ci "Add baremetal testing" [High,Triaged] 20:11:59 any other bugs people want to bring up? 20:12:37 the other ones lead to the Grizzly POC we did.. so.. 20:12:42 #topic Grizzly test rack status 20:13:03 My plan has been to go after the POC: bugs after I tackle the other two criticals. Still the plan. 20:13:22 Have not heard anything specific about the test rack lately though. 20:13:35 so if nobody else has anything 20:13:39 * SpamapS pauses 30s for lag 20:14:28 #topic CI virtualized testing progress 20:14:31 SpamapS: https://bugs.launchpad.net/nova/+bug/1217082 20:14:32 Launchpad bug 1217082 in nova "baremetal driver needs better tests" [Undecided,New] 20:14:36 jog0: cripes 20:14:44 jog0: oh ok that was just "here's the bug" ;) 20:14:48 * SpamapS un-cripes jog0 20:14:52 no updates since last week (I got busy with other infra things) 20:14:53 SpamapS: haha yeah 20:14:56 pleia2: ^^ you're up. 20:15:33 jog0: perhaps you just want to instead add nova to https://bugs.launchpad.net/openstack-ci/+bug/1082795 ? 20:15:34 Launchpad bug 1082795 in openstack-ci "Add baremetal testing" [High,Triaged] 20:15:50 pleia2: perhaps you and I should buy a round of beers for the team for every week we get distracted from our TripleO tasks. ;) 20:15:53 pleia2: that too but I want unit tests 20:16:01 * jog0 updates bug 20:16:04 jog0: oh ok :) 20:16:07 SpamapS: haha 20:16:23 #topic open discussion 20:16:43 I have one actaully 20:16:48 upgrade ordering 20:16:51 few more reviews on https://review.openstack.org/#/c/30441/22/neutron/db/extradhcpopt_db.py after we added the new features 20:17:14 so I recently got stuck by the fact that nova has no rpcapi compat testing 20:17:37 so a a first pass I am just trying everything trunk and grizzly compute 20:17:57 does it seem reasonable to assume nova-compute nodes will be the last step in an upgrade? 20:18:46 jog0: yes. I've been thinking actually about the fact that we really need something that systematically controls orchestration changes in Heat.. 20:19:21 SpamapS: yeah, well for now i am ignoring the actual upgrade and testing the mid-upgrade code path 20:19:29 but to do so I need to define what that is 20:19:57 jog0: I think all of the components should be able to express, at least in documentation, what they need done and when, for an upgrade. 20:20:29 jog0: so if nova compute needs all other RPC partners to be speaking the "new" RPC, then it should say that in the manual. :) 20:20:30 SpamapS: ++, having this docuemnted in triplo will be really helpful 20:21:03 SpamapS: well with upgrade_levels the rpc upgrade stuff is a little easier 20:21:06 in theory 20:21:15 Yeah I'd hope all new things can talk to all old things. 20:21:23 SpamapS: exactly 20:21:51 But I would accept that there may be a few special pieces that need ordering. 20:22:05 SpamapS: yeah, for example nova-compute should be last 20:22:12 I think 20:22:37 nova-compute may also be one where the nodes are just decomissioned and allowed to die-off through attrition.. so they should be able to stay online for a long, long time. 20:23:11 so early testing looks pretty decent for running grizzly nova-compute with trunk 20:23:26 after I reverted a patch, tempest only had 10 failures 20:23:33 only 10! 20:23:36 :) 20:23:55 so just 1% failure rate 20:24:45 ok, good stuff to keep in mind. 20:24:46 anything else? 20:25:09 nothing over here 20:25:20 thanks SpamapS 20:25:31 By the power vested in me by the distinguished technologist Robert Collins and the great foundation of OpenStack, I release you all back to productivity. May the code be with you. 20:25:38 #endmeeting