19:00:25 #startmeeting tripleo 19:00:26 Meeting started Tue Oct 1 19:00:25 2013 UTC and is due to finish in 60 minutes. The chair is lifeless. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:00:27 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:00:29 The meeting name has been set to 'tripleo' 19:00:32 #topic agenda 19:00:34 o/ 19:00:42 bugs 19:00:42 Projects needing releases 19:00:42 Grizzly test rack status 19:00:42 CI virtualized testing progress 19:00:42 Insert one-off agenda items here 19:00:45 Tuskar UI design update discussion 19:00:47 CD proposal 19:00:50 open discussion 19:00:52 #topic bugs 19:01:06 #link https://bugs.launchpad.net/tripleo/ 19:01:06 #link https://bugs.launchpad.net/diskimage-builder/ 19:01:06 #link https://bugs.launchpad.net/os-refresh-config 19:01:06 #link https://bugs.launchpad.net/os-apply-config 19:01:08 #link https://bugs.launchpad.net/os-collect-config 19:01:11 #link https://bugs.launchpad.net/tuskar 19:01:13 #link https://bugs.launchpad.net/tuskar-ui 19:01:16 #link https://bugs.launchpad.net/python-tuskarclient 19:01:34 Hey everyone, welcome to the first unified meeting 19:02:52 so, in dib we have one critical 19:03:32 which should be fix-released 19:04:22 https://bugs.launchpad.net/os-collect-config/+bug/1233656 looks like it will be a fun bug 19:04:44 I'm going to bet it's global requirements time... 19:05:03 I see https://bugs.launchpad.net/tuskar has lots of undecided bugs 19:05:12 shadower: hey 19:05:32 shadower: we haven't updated the tuskar bug teams I think; I can't triage on tuskar 19:05:54 shadower: should we do the same thing we did for review teams? just merge, and update the projects to use the combined team? 19:05:55 lifeless, I think he is not here 19:06:11 lifeless, most of the brno team is on cloud party 19:06:18 DOH 19:06:22 And they didn't invite us? 19:06:33 well it's in Brno :-) 19:07:08 Speaking as a not-in-charge Tuskar person, though, merging teams in LP makes sense to me 19:07:32 ok 19:07:59 make sense for me too 19:08:08 +1 19:08:09 #action shadower to reconfigure tuskar,tuskar-ui,python-tuskarclient in LP to use the tripleo team for driver/bug triage etc. 19:08:11 +1 19:08:21 #action lifeless to pull into the tripleo team the existing tuskar team members 19:08:35 those two things combined should DTRT 19:08:53 so there's a bunch of triage needed in tuskar 19:09:05 and tuskar-ui 19:09:15 python-tuskarclient is in better shape 19:09:32 Reminder: everyone should triage; it doesn't need to be closely held :) 19:09:48 Once the teams are sorted, I'm sure that will be easier 19:11:27 we have bug 1233596 19:11:36 hmm, no urlifier, pok 19:11:50 https://bugs.launchpad.net/tripleo/+bug/1233596 19:11:54 lets ensure thats fixed asap 19:12:18 should be an easy pie, AFAIK 19:12:19 the existing POC bugs can be closed I think 19:13:29 I think we're going to have to close https://bugs.launchpad.net/tripleo/+bug/1189385 as cannot-reproduce 19:13:32 as it hasn't happened again 19:13:40 any thoughts 19:13:41 ? 19:13:43 lifeless: seems like tuskar saw that "six < 1.4" issue too (https://bugs.launchpad.net/os-collect-config/+bug/1233656) @ https://review.openstack.org/#/c/44689/ 19:14:39 lifeless: agreed on POC bugs, they are too far away from any future-focused work to be useful and they don't have data points. 19:16:37 lifeless: re neutron/quantum wonkiness.. I think it is a data point and perhaps just deserves demotion to Medium until we see it again. 19:17:04 anyone but me and derekh seeing this https://bugs.launchpad.net/tripleo/+bug/1231930 ? 19:18:04 rpodolyaka1: I haven't heard of it no 19:18:10 rpodolyaka1: you see derekh's theory on gre? 19:18:24 marios: interesting 19:18:55 lifeless: I applied the fix, he proposed in the bug description and it works. not sure, how correct this is 19:19:06 SpamapS: where does the <1.4.0 come from for six? 19:19:18 lifeless: global reqs 19:19:23 SpamapS: not anymore 19:19:26 I believe there is a patch in queue to address 19:19:39 https://review.openstack.org/#/c/49194/1/requirements.txt 19:19:45 SpamapS: six 19:19:46 lifeless: though derekh fixed this in Fedora by adding of an iptables rule (this doesn't fix the problem for Ubuntu 13.04/OVS 1.9) 19:19:51 SpamapS: is whats in global-requirements. 19:19:56 rpodolyaka1: oh, you're seeing it in Ubuntu ? 19:20:04 lifeless: yep 19:20:15 lifeless: approve that and I'll tag a release 19:20:41 looks like an OVS bug to me, though I can be wrong 19:20:57 we'd better ping Neutron guys to take a look at this one 19:21:05 SpamapS: done 19:21:12 rpodolyaka1: if the iptable rule doesn't help for your situation then look like a different problem 19:21:15 rpodolyaka1: so, clearly something wrong 19:21:34 rpodolyaka1: I don't think the hard timeot is appropriate; the timeout is to prevent cruft building up over time 19:21:50 rpodolyaka1: I definatly had a VM I couldn't contact, and then could after allowing gre 19:22:00 rpodolyaka1: tested several times 19:22:39 derekh: Fedora has kernel 3.11 and ovs 1.11 19:22:39 lifeless: sweet ty 19:23:20 * rpodolyaka1 is tired of fighting with ovs/neutron and blames versions of software... 19:23:20 rpodolyaka1: yup 19:23:34 is there anyone setup with an overcloud that *doesn't* show this problem ? 19:23:36 lifeless: sorry, yeah let's merge the LP teams 19:24:11 shadower: if you change the owner of the tuskar teams, I can do the merging for you 19:24:15 shadower: (to me that is)( 19:26:07 lifeless: done for tuskar and tuskar-ui. I can't change the pythonclient, will ask the cli guys to do that tomorrow 19:26:22 rpodolyaka1: before ovs was updated on fedora, (and we had ovs 1.10, this command was returing a -1 19:26:24 ovs-vsctl get Interface gre-1 ofport 19:26:28 shadower: ack, thanks 19:27:29 derekh: I checked neutron logs on compute/notcompute - all OVS calls exit codes equal 0 19:28:08 rpodolyaka1: derekh: lets take this offline from the meeting. 19:28:16 sure 19:28:23 rpodolyaka1: derekh: I suggest expanding the eyeballs that look at it e.g. via the -dev list. 19:28:33 ther's nothing obviously tripleo specific about the fail 19:28:55 okie dokie 19:29:12 #topic Projects needing releases 19:29:30 I've added this section because I want to ensure we don't leave useful things locked up in git HEAT 19:29:33 bah 19:29:35 HEAD 19:30:40 etoomany names 19:31:12 so 19:31:15 lifeless: is there a quick git command to say "show me everything since the last tag" ? 19:31:25 git log 0.0.4..HEAD 19:31:30 oh, last tag 19:31:31 possibly 19:31:40 well getting the last tag isn't _that_ difficult 19:31:42 so dib needs a release. 19:32:02 occ needs a release 19:33:09 oac needs one 19:33:49 orc needs one 19:34:09 orc? 19:34:15 Update stackforge references to openstack 19:34:18 affects metadata 19:34:28 Yeah surprised that we have any changes. ;) 19:34:42 #action SpamapS release all the things 19:35:05 shadower: / anyone with tuskar history :) - has python-tuskarclient done a release? 19:35:06 lets try and land anything in the patch queue first 19:35:14 likewise the tuskar api and ui ? 19:35:30 lifeless: don't believe so 19:35:33 (no) 19:35:40 so, unreleased code is unusable code :> 19:35:45 for any of them 19:36:05 yea 19:36:15 I think we should put this out to the list for discussion, but I'd like to get the tuskar bits on the release train sooner rather than later 19:36:24 python-tuskarclient isn't contentious 19:36:39 however ttx may have opinions about incremental delivery of web services. 19:36:47 yea, it'd be great to point people to something we know works 19:36:59 first step: get it work :) 19:37:00 if there are no objections, I'll raise the subject on the dev list 19:37:05 +1 19:37:12 lifeless: speaking of releases 19:37:28 what about another t-i-e release? 19:37:35 oh, I did not check it 19:37:44 +1 19:37:49 and tht too probably. 19:37:53 SpamapS: ^ ALL the things 19:38:02 tht needs one too 19:38:04 :) 19:38:06 speaking of which, would be awesome if we could land https://review.openstack.org/#/c/47589/ in there too 19:38:09 more things.... 19:38:10 SpamapS: btw I think 'land everything first' is an antipattern. 19:38:11 (tie) 19:38:16 t-h-t 19:38:19 lifeless: fair enough 19:38:21 SpamapS: we have inventory, ship it. 19:38:44 https://git.openstack.org/cgit/openstack/tripleo-heat-templates/refs/ no releases so far 19:38:45 SpamapS: review shouldn't be rushed; and putting release after review pressures review. 19:38:45 lifeless: but its like, an entire command, _each_ repo.. to do a release. 19:38:49 _a whole command_ 19:38:54 marios: seems to be working now... going to try once again tomorrow and remove WIP status 19:38:55 SpamapS: I know, right. 19:39:14 rpodolyaka1: awesome, was going to ping you about that tomorrow 19:39:48 marios: depends on https://review.openstack.org/#/c/47826/ be merged into tuskar-api though 19:40:17 ok, anything else on releases ? 19:40:20 SpamapS: thoughts on a t-h-t release ? 19:40:40 rpodolyaka1: noted, lets talk tomorrow 19:41:35 ccrouch: its been a grab bag of fail for a while now... but I've been pondering adding tox support so it runs its tests in the gate... releases naturally would follow that. 19:42:47 * ccrouch grabs his bag of fail and runs 19:42:48 I agree; though I don't think we /have/ to have CI to do 0.y.z releases 19:42:51 per semver. 19:43:05 #topic Grizzly test rack status 19:43:08 * ccrouch would be pretty stoked with just a tag :-) 19:44:18 ccrouch: I'd love to have the first release have a merge.py that isn't god awful. 19:44:29 but that wouldn't be a first release would it? 19:44:31 SpamapS: release early, release often :> 19:44:41 "if your first release isn't embarassing..." 19:44:48 then it's not an MVP 19:44:54 so the grizzly test rack is going to die 19:45:03 I hope we'll agree to replace it with a trunk POC rack 19:45:15 Please a) see my email to the list 19:45:31 and b) give feedback! So far it's been an overwhelming silence... 19:45:44 with a few direct IRC comments saying 'hell yeah'[6~ 19:46:15 * shadower should be able to find some code review & email time tomorrow 19:47:02 #topic CI virtualized testing progress 19:47:18 pleia2: , clarkb:, mordred: and I spoke in person in Seattle about this 19:47:24 there is a new plan 19:47:37 you have to say new plan in a loud deep voice 19:47:53 New Plan 19:48:00 there you go 19:48:04 hehe 19:48:08 https://etherpad.openstack.org/tripleo-test-cluster 19:48:13 #link https://etherpad.openstack.org/tripleo-test-cluster 19:48:40 I propose that the tooling for this can be brought up in parallel with the cloud we'll need to run it on 19:48:58 we may have one machine already that it can run on 19:49:09 and we have a (non-state-preserving) overcloud we can run it on already 19:49:17 kindof 19:49:56 pleia2: does that make sense to you? Are you able to push the infra glue forward if the tripleo folk push the cloud it'll run in forward? 19:50:17 I'll make sure that we have a set of action items in the spec, and probably fairly specific kards for it in kanban 19:50:21 lifeless: yeah, working closely with clarkb on this so between the two of us we can get whatever you need done 19:50:35 pleia2: ok cool; I will drill in in detail with you post meeting then ? 19:50:47 lifeless: sounds good 19:50:47 #action lifeless and pleia2 to talk turkey on ci story 19:51:04 #topic Tuskar UI design update discussion 19:52:16 so there are some tuskar ui updates based on the list discussion 19:52:20 the designers would love feedback 19:52:31 " 19:52:32 Updated processes in Tuskar-UI in order to be more flexible. 19:52:32 Change is reflected in wireframes + there is a link for youtube stream (posted in openstack-dev list, with [Tuskar] prefix in subject). 19:52:35 Change is demonstrated in UI but it affects API a lot. API needs to be as flexible as possible to support multiple architectures and concepts which were presented in the UI. 19:52:39 " 19:52:49 Please have a look at this and discuss on the list 19:53:02 It's a high leverage chance to influence things for quite some time 19:53:17 #topic CD proposal 19:53:34 (Sorry for pushing through, we have a lot of content today) 19:53:43 This is about running a reference TripleO cloud 19:53:58 Which we'd all be collaboratively responsible for 19:54:17 * SpamapS loves the idea 19:54:30 I sent a mail to the list 19:54:43 has anyone had a chance to see it and think through/got questions ? 19:56:02 ok, deferring to the list then 19:56:14 #topic open discussion 19:56:55 * matty_dubs reading that mail now 19:58:54 As far as open discussion goes, and hearing a lot of silence, I should say that I'm excited that tuskar and tripleo are united! 19:59:08 yaaay 19:59:22 :-) 19:59:55 matty_dubs: cool :) 20:00:00 ok, out of time 20:00:02 #endmeeting