20:00:37 #startmeeting tripleo 20:00:38 Minutes (text): http://eavesdrop.openstack.org/meetings/ironic/2013/ironic.2013-08-05-19.00.txt 20:00:39 g'morning, lifeless :) 20:00:39 thanks everyone :) 20:00:39 Log: http://eavesdrop.openstack.org/meetings/ironic/2013/ironic.2013-08-05-19.00.log.html 20:00:40 Meeting started Mon Aug 5 20:00:37 2013 UTC and is due to finish in 60 minutes. The chair is lifeless. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:00:41 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 20:00:44 The meeting name has been set to 'tripleo' 20:01:10 #topic agenda 20:01:22 bugs 20:01:23 Grizzly test rack status 20:01:23 CI virtualized testing progress 20:01:23 open discussion 20:01:26 #topic bugs 20:01:38 https://bugs.launchpad.net/tripleo/ 20:01:39 https://bugs.launchpad.net/diskimage-builder/ 20:01:39 https://bugs.launchpad.net/os-refresh-config 20:01:39 https://bugs.launchpad.net/os-apply-config 20:01:39 https://bugs.launchpad.net/os-collect-config 20:01:47 devananda: morning devananda :) 20:04:29 I have a minor update on https://bugs.launchpad.net/tripleo/+bug/1184484 20:04:31 Launchpad bug 1184484 in tripleo "Quantum default settings will cause deadlocks due to overflow of sqlalchemy_pool" [Critical,Triaged] 20:04:32 ugh, sec, destructo cat 20:04:39 which is I was unable to reproduce 20:05:17 jog0: interesting 20:05:30 jog0: we can still make it happen in the POC, which you have access to now 20:06:42 lifeless: in that case lets sync up after the meeting and work on it 20:06:46 ok 20:07:00 bug 1182249 20:07:01 Launchpad bug 1182249 in tripleo "quantum configuration is overly hardcoded" [Critical,In progress] https://launchpad.net/bugs/1182249 20:07:07 and bug 1183223 20:07:08 Launchpad bug 1183223 in tripleo "nova-compute.yaml missing parameters" [Critical,In progress] https://launchpad.net/bugs/1183223 20:07:09 both spzala 20:07:10 bah 20:07:14 both SpamapS 20:08:26 the three key POC bugs have noone on them yet. 20:08:34 We probably need a volunteer? 20:10:11 SpamapS: ping? 20:10:43 lifeless: pong sorry 20:10:54 * SpamapS got distracted and lost track of time 20:12:01 lifeless: re the POC bugs.. I've been working on 1202322 so long, have not even really had time to think about them. 20:12:12 bug 1202322 20:12:13 Launchpad bug 1202322 in tripleo "cfn-hup restarts everything every 5 minutes because of heat-localip" [Critical,In progress] https://launchpad.net/bugs/1202322 20:12:33 lifeless: I think it may be solved now that we have os-collect-config in t-i-e .. have not had a chance to check yet though. 20:12:49 I will follow up with markmcclain about https://bugs.launchpad.net/tripleo/+bug/1189385 20:12:52 Launchpad bug 1189385 in tripleo "quantum-server hung up it's listening port" [Critical,Triaged] 20:13:03 lifeless: then bug 1182249 and bug 1183223 will be mostly about reviewing them to see if they still apply and addressing the issues left over. 20:13:06 Launchpad bug 1182249 in tripleo "quantum configuration is overly hardcoded" [Critical,In progress] https://launchpad.net/bugs/1182249 20:13:07 Launchpad bug 1183223 in tripleo "nova-compute.yaml missing parameters" [Critical,In progress] https://launchpad.net/bugs/1183223 20:13:29 ok 20:13:40 is there anything the rest of the team can do to help you ? 20:14:11 test .. first bit landed today 20:14:22 and hopefully we can roll the rest out today 20:14:44 ok 20:14:48 any other bugs for discussion ? 20:15:30 * SpamapS may have tunnel vision, so no 20:16:00 ok 20:16:01 so 20:16:02 Grizzly test rack status 20:16:09 #agenda Grizzly test rack status 20:16:17 This is unchanged as far as I know 20:16:25 I got jog0 access to it on friday 20:17:04 remember we do have spare machines in the rack, someone just needs to climb through the records to resurrect them; contact me for the spreadsheet if you wish to do that 20:17:35 we still basically only have one shared vlan tho right? 20:17:51 the space machines should be in separated vlans 20:18:27 there is some confusion around what did and did not get done to the network 20:18:35 that is one of the things to dig into 20:18:51 I've been told with great authority conflicting things :) 20:19:43 Depending on which things are false there are 1/2/3 broadcast domains in the rack's native vlans, ignoring the IPMI connections 20:20:36 ok, - hopefully someone will find the time soon. 20:20:38 moving on 20:20:44 #topic CI virtualized testing progress 20:20:46 o/ 20:21:13 the mic is open 20:21:13 so last week I narrowed down the instructions I should be useing to get basic seed going 20:21:19 still iscsi errors 20:21:43 ok 20:21:55 can we do a focused buddy-debug session today ? 20:22:00 I did some tcpdumps to try to pinpoint what exactly the problem is (and what it's doing) but so far haven't been able to narrow down the dumps to be specific enough to be useful 20:22:15 so I think a better understanding of what it's trying to do would help 20:22:18 lifeless: that'd be great 20:22:37 once I understand what it's doing with iscsi I can do more direct, exact testing 20:23:19 ok 20:23:28 oh, devananda also let me know that we'll need a new power manager driver when we go to lxc 20:23:46 for the record, what it does is repartition the disk and then copy the raw image onto it 20:23:46 since the current one is specifically for kvm, said to follow up with NobodyCam when I'm at that point 20:24:09 hm 20:24:11 we don't need a new power manager, as we're not planning on using lxc to emulate physical nodes; just the bare metal compute controller. 20:24:15 AIUI. 20:24:15 unless we use libvirt_lxc ;) 20:24:18 pleia2: you'll need a few new bits inside of virtual_power_driver_settings, but that should be it -- not a whole new power driver 20:24:19 ah, ok 20:24:28 devananda: ^ 20:24:32 or, not :) 20:24:33 ah, and what he said 20:24:39 even better then, thanks 20:24:51 lxc vm's can't boot arbitrary kernels, so make poor test beds for emulated hardware. 20:24:56 AIUI, IMBW. 20:25:29 anyhow, after-meeting. 20:25:29 lifeless: i believe you're right on that. containers in general dont, AFAIK, diverge from the host's kernel very well 20:25:50 that's it from me 20:26:11 #optic open discussion 20:26:12 bah 20:26:16 #topic open discussion 20:26:33 Anyone that hasn't booked travel yet, you really should :) 20:26:43 is there an agenda available for the Seattle meeting? 20:26:53 in case people cant make it to the whole week 20:27:12 devananda: heh, LXC absolutely can't diverge from host kernel.. since LXC is really just namespaced syscalls. 20:27:50 ccrouch: good point; lets start an etherpad and discuss some broad buckets of timing on the list 20:27:50 ccrouch: good point. We should probably write one up now that we have a rough idea of the attendance. 20:28:17 ccrouch: it is a working sprint rather than a series-of-meetings however, so there is only so much contact-with-the-enemy we can predict 20:28:45 ok gotya, understood 20:29:04 SpamapS: i was being cautious in my statement, just in case there is some crazy way to do it with openvz or something else that i'm unaware of 20:30:02 ccrouch: https://etherpad.openstack.org/tripleo-havana-sprint 20:32:18 lifeless: thanks, to a certain extent the agenda may impact which days to attend, and who should attend, so I appreicate you putting that up 20:32:34 ccrouch: suggestions welcome! 20:36:45 ccrouch: of course, like any good agenda, it will be exploded and unrecognizable by the end of day 2. 20:37:13 :-) 20:37:13 i will pass the URL around 20:37:13 BTW how much interest is there in topics which are targetted at the newer/occasional patch contributor? versus the jedi-level of yourself and SpamapS 20:37:45 the broad theme is to get tripleo in shape to handle H well 20:38:12 a second theme is to work with interested folk in a high bandwidth environment to do what they are interested in doing. 20:38:36 I'll be extremely happy if anyone that pops in works on what they want and will actively help out with those things. 20:38:46 scaling up the team is more important than any single work item IMO 20:39:27 agreed 20:39:46 I see value in gathering entry level work around a couple of days 20:39:52 seems like the first 2 days would be the days to do that. 20:40:02 jog0: heh, itworld didn't contact me for comment :) 20:40:09 SpamapS: agreed 20:40:12 and then Wednesday.. the jedi trials. ;) 20:40:35 we can open thursday with some yoga - sith salutation 20:40:44 lol 20:41:07 buddah on lotus opens the shakras for infinite recursion 20:41:42 Ok I've dropped my suggestions tin. 20:41:44 in 20:42:14 lifeless: want to send that to the ML and we can sort of ratify it next Monday? 20:42:26 sure 20:43:01 then anybody who hasn't booked yet can choose the days they want if they can't handle drink^H^H^H^Heveloping with all of us all week. 20:43:51 dreveloping ? 20:44:13 oh no, there are 4 ^H's, interpreter error. 20:44:16 haha 20:44:18 ok, I think we're done here. 20:44:22 last call for stuff 20:44:23 maybe you've been dreveloping? 20:44:37 no, but I do have some pink champagne from duty free. 20:44:44 Perhaps I should try it :) 20:45:04 * SpamapS seconds the motion that this should end.. soon 20:45:32 #endmeeting