19:01:39 #startmeeting tripleo 19:01:40 Meeting started Tue Dec 16 19:01:39 2014 UTC and is due to finish in 60 minutes. The chair is SpamapS. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:01:41 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:01:42 * tchaypo is busy fooding 19:01:45 The meeting name has been set to 'tripleo' 19:01:52 o/ 19:02:00 hey 19:02:19 #link https://wiki.openstack.org/wiki/Meetings/TripleO 19:02:39 * lsmola is busy with kid running around 19:02:44 #topic agenda 19:02:44 * bugs 19:02:44 * reviews 19:02:44 * Projects needing releases 19:02:44 * CD Cloud status 19:02:46 * CI 19:02:49 * Tuskar 19:02:51 * Specs 19:02:54 * open discussion 19:02:56 Remember that anyone can use the link and info commands, not just the moderator - if you have something worth noting in the meeting minutes feel free to tag it 19:02:59 #topic bugs 19:03:16 #link https://bugs.launchpad.net/tripleo/ 19:03:16 #link https://bugs.launchpad.net/diskimage-builder/ 19:03:16 #link https://bugs.launchpad.net/os-refresh-config 19:03:16 #link https://bugs.launchpad.net/os-apply-config 19:03:16 #link https://bugs.launchpad.net/os-collect-config 19:03:19 #link https://bugs.launchpad.net/os-cloud-config 19:03:21 #link https://bugs.launchpad.net/os-net-config 19:03:24 #link https://bugs.launchpad.net/tuskar 19:03:26 #link https://bugs.launchpad.net/python-tuskarclient 19:03:44 Howdy folks! Let's jump right in. Please highlight any non-critical bugs you want to bring to the group's attention. Now, to the criticals. 19:04:38 we can't use no-mergepy yet because of features in heat that are being tweaked 19:05:04 tchaypo: sounds right 19:05:14 anybody what to elaborate on that? 19:05:20 there's a bug that I'd look up if I wasn't on iPad 19:05:30 tchaypo: do you have a list of what is still broken? 19:05:58 We should be marking those as affecting us, perhaps even link to the blueprint. 19:06:01 https://bugs.launchpad.net/tripleo/+bug/1374626 19:06:06 Launchpad bug 1374626 in tripleo "UIDs of data-owning users might change between deployed images" [Critical,Triaged] 19:06:32 That one is assigned to me. It is still 3rd in line behind setting up some CI for tripleo-ansible (almost done..) and working on L3 segmentation for ironic/nova. 19:06:38 tchaypo: also sending it to mailing list would be nice 19:07:23 lsmola: I believe there was a thread where we discussed the main blocker, which is that we have no migration path from mergepy templates to not-mergepy templates that doesn't involve having spare machines. 19:07:48 The external_reference feature is needed to do that properly. 19:07:53 https://launchpad.net/bugs/1401929 19:07:55 Launchpad bug 1401929 in heat "Overzealous validation of images in empty ResourceGroups" [High,In progress] 19:07:59 yea that is a big one 19:08:07 The image validation is a misfeature IMO. 19:08:22 But nobody seems to agree with me in Heat land. ;) 19:08:24 https://bugs.launchpad.net/tripleo/+bug/1385346 19:08:25 Launchpad bug 1385346 in tripleo "upstart service unreliable after introducion of pipe to logger" [Critical,Fix committed] 19:08:47 lsmola: kk, I'll email later 19:09:04 great 19:09:17 greghaynes: ^^ is this solved? I thought we still had issues because of disappearing servers 19:09:49 Not solved, there is a patch https://review.openstack.org/#/c/138844/ 19:09:53 greghaynes: it's marked Fix Committed but I don't really believe it. 19:10:03 I should update that commit msg 19:10:17 greghaynes: ok, so should you attach that patch to the bug? 19:10:23 Yep 19:10:25 greghaynes: ok, marking In Progress and assigning to you. 19:11:17 greghaynes: also did we ever start the discussion on openstack-dev about making syslog default? I recall you falling down an eventlet rabbit hole. 19:12:02 I only got +1 responses (maybe only one?). The eventlet issue is fixed by just writing to /dev/log instead of syslog() 19:12:10 which is what that patch does :) 19:12:39 greghaynes: ok, cool. It would be good to Related-Bug the bug about syslog being default in your patch. 19:12:48 * bnemec needs to catch up on that 19:12:58 righto 19:13:04 https://bugs.launchpad.net/tripleo/+bug/1401300 19:13:06 Launchpad bug 1401300 in tripleo "Keystone Private Key not securely sent to host" [Critical,Confirmed] 19:13:13 Anybody want to bite on this one? 19:13:20 It's really a spec I think. 19:13:55 SpamapS: yeah, that is a new mechanism 19:14:08 I think were going to end up wanting a short term dirty fix (like scp) and a spec for something more sane 19:14:08 We could just accept the risk. 19:14:31 Also did we ever actually turn on https for heat/cfn API access? 19:14:48 Probably not. 19:14:51 Something a lot nicer should come as a result of the barbican work im going but thats about 3 medium sized features out... 19:15:16 i thomk there's a pending change that does that, but im not dure if its doing it intentionally or if it a mistake 19:16:18 Ok, so we need somebody to look at this and _at the very least_ assess the risk and see if we really do need to fix it "zomg right now" or if it is just something we need to document as a shortcoming and drive the long term fix as a "High" bug. 19:16:51 I'm super overcommitted or I'd take it on. Anybody? 19:17:17 ok, I can take on assessing it since im pretty familiar with whats going on 19:17:23 greghaynes: \o/ 19:17:55 everybody buy greghaynes beer in Seattle. Or anything else he wants to consume recreationally within the limits of the law and professionalism ;-) 19:18:22 haha, gluten free tasty beverages are welcome :) 19:18:30 seattle cider co is pretty good 19:19:01 https://bugs.launchpad.net/tripleo/+bug/1401617 19:19:05 Launchpad bug 1401617 in os-cloud-config "Flavor based ramdisk/kernel ID is deprecated in Ironic since Juno" [Critical,In progress] 19:19:22 I'm kind of trying to drive this one on the side. It isn't easy though, as we have to change os-cloud-config and tripleo-incubator in weird ways. 19:19:38 Should get done before they remove the flavor->ramdisk+kernel mapping though. 19:19:44 In fact they may keep that until K+1 19:19:53 https://bugs.launchpad.net/tripleo/+bug/1401957 19:19:55 Launchpad bug 1401957 in tripleo "pip install MySQL-python failing on fedora" [Critical,Triaged] 19:20:27 This one is "solved" but awaiting final publish to mirrors, and only affects Fedora users, so I move to drop it to High. 19:20:32 anyone object? 19:21:17 I take 60s of silence as no. Dropped to High. 19:21:44 we have a workaround in Ci already afaik 19:21:51 high sounds good to me 19:21:59 SpamapS: high is good 19:23:36 Ok all the rest are fix committed I think. 19:23:37 getting back to https urls 19:23:40 https://review.openstack.org/#/c/131795/ 19:24:07 https://bugs.launchpad.net/diskimage-builder/+bug/1402085 19:24:08 Launchpad bug 1402085 in diskimage-builder "pkg-map incompatible with new install-packages element" [Critical,In progress] 19:24:15 that may need some review love 19:25:04 tchaypo: that looks like it needs a bug 19:26:16 yeah, i don't understand what its doing. it seems to have had a couple of +2s and im not sure why 19:26:20 just so we can make sure to drive it to completion 19:26:43 Ok, looks worth a closer.. look. 19:26:59 anyway, long time on bugs. Let's move on. 19:27:14 #topic reviews 19:27:25 oh, one more bug thing 19:27:30 Bug squash day tomorrow 19:27:45 Just a reminder :) 19:28:19 #info There's a dashboard linked from https://wiki.openstack.org/wiki/TripleO#Review_team - look for "TripleO Inbox Dashboard" 19:28:21 #link http://russellbryant.net/openstack-stats/tripleo-openreviews.html 19:28:24 #link http://russellbryant.net/openstack-stats/tripleo-reviewers-30.txt 19:28:27 #link http://russellbryant.net/openstack-stats/tripleo-reviewers-90.txt 19:28:31 greghaynes: RIGHT! should have said that in bugs section. ;) 19:28:45 thanks greghaynes 19:28:53 can you email to remind people about that? 19:29:01 tchaypo: yep, good call 19:29:11 for stevenk it's bug squash day as soon as he wakes up 19:29:34 We're still _very_ slow on reviews since the US Thanksgiving break. 19:30:03 I suspect that wont get better over upcoming holidays 19:30:25 One stat I would like to see is the throughput. 19:30:38 It's great to see our problems, but I'd like to have visibility into our capabilities too. 19:31:00 queue growth is up to 2.2/day on the 30 day report 19:31:09 from memory it was 1.9 last week 19:31:16 whoa, 190 days https://review.openstack.org/#/c/98817/ 19:31:35 #link http://stackalytics.com/report/reviews/tripleo-group/open 19:31:40 different view of things there 19:31:52 That was wip for a lot of that time though. 19:31:53 workflow -1 in jun 19:32:31 tchaypo: are you core? or are the stats just not updated? 19:32:39 im not core 19:32:50 #link http://stackalytics.com/?module=tripleo-group&metric=marks 19:33:22 my review numbers are the highest they've ever been right now 19:33:30 This actually does show throughput 19:33:54 and we're definitely lower than we were in the past 19:34:04 if I can get them higher and keep them there for a few months I'll start asking about being core 19:34:50 tchaypo: in the grand scheme of things, you're beating half the core team. 19:35:02 do we need to drop some inactive members? 19:35:28 I feel like right now would be the wrong time to do that. 19:35:32 but maybe we'll revisit in January. 19:35:34 Yeah 19:35:39 +1 to January 19:35:50 i'm beating numbers, I don't know about quality 19:35:51 things will look worse in january wouldn't they? 19:35:53 Anyway, does anybody have any specific things to discuss about reviews? I feel like we could probably do better. 19:36:07 ya, i'm taking two weeks off for xmas holiday, that won't help my numbers in january :) 19:36:18 but spam apps any feedback would be welcome and helpful :) 19:36:20 slagle_: yes they will, but I'm suggesting we tell those with the worst numbers to step-up or step-out. :) 19:36:57 hehe 19:38:35 the other thing is - perhaps our metrics are measuring the wrong thing. perhaps we should be asking people for input on redefining the core commitments for next year? 19:38:55 it has been about a year since we moved to the 3/day thing, might be worth a discussion 19:39:03 does 3/workday still make sense? 19:39:16 let's table that for the ML 19:39:23 agree 19:39:39 #topic CI 19:39:49 #link http://goodsquishy.com/downloads/s_tripleo-jobs.html 19:40:29 Hmm, <50% pass rate. 19:40:33 http://goodsquishy.com/downloads/tripleo-jobs.html theres a lot of red 19:40:34 68% for nonha .. given the fact that fedora was totally borked for a week.. this seems o-k 19:40:38 bnemec: thats w/ HA 19:40:41 f20 was broken for a bit 19:40:44 also everything was borked for the weekend 19:40:56 Ah 19:41:13 Derekh or I will send the weekly email tomorrow or the day after 19:41:13 Right setuptools fun 19:41:35 greghaynes: I wasn't aware of that, do you have a bug? 19:41:44 tchaypo: anything to report on HP regions? 19:42:22 derekh was adding 15 nodes to hp1 19:42:54 I'm stuck on hp2 right now, I rebuilt the seed and now it's not responding to arp on eth1 and it's broken the limits of my knowledge 19:43:02 tchaypo: https://review.openstack.org/#/c/141659/ 19:43:17 thanks greghaynes 19:43:53 tchaypo: perhaps it's time to shift roles a bit.. seems stalled. We can discuss later. 19:44:08 kk 19:45:06 anything else about CI? 19:46:02 see email soon, nothing I'm aware of 19:46:10 #topic Tuskar 19:46:16 jdob: got anything for us? 19:46:31 SpamapS: nope, I've been diverted all month; we can probably drop this topic from the agenda entirely 19:46:40 no real reason to call out one of the 10 or so projects within tripleo 19:46:49 jdob: ok, can you edit it out of https://wiki.openstack.org/wiki/Meetings/TripleO please? 19:46:51 i'll jump up and down on the ML or reviews section if I need eyes :) 19:46:52 will do 19:47:09 jdob: But it's the most visually attractive project.... ;) 19:47:17 #topic Specs 19:47:33 Haven't seen much in the way of new specs. Anybody have a spec they want to hilight? 19:47:38 * jdob resists urge to comment on himself also being visually attractive 19:47:58 Is the cinder/ha one merged? 19:48:19 jdob: professionalism.. ;) 19:48:28 ya, i'm dripping with it 19:48:52 https://review.openstack.org/#/c/101237/ is cinder ha and still not merged 19:48:59 and ive heard lots of review begging 19:49:19 Yes, review that spec. 19:49:45 and all specs 19:50:03 * SpamapS hums "Let's talk about specs, baby..." 19:50:13 #topic open discussion 19:50:18 professionalism.... 19:50:25 tchaypo: ah-thankyou 19:50:33 we've had a few ML conversations about meeting time for the other meeting 19:50:52 I'd ask if anyone can't make this meeting but clearly they aren't here... 19:50:57 :-/ 19:51:16 also about the purpose of the meetings and if there's anything we can change 19:51:18 tchaypo: did you try kicking out my votes and any for others in the American timezones? 19:51:55 I haven't yet, I've just been manually mentally adjusting votes 19:52:10 tchaypo: I think the meeting is intended to gather realtime participation around common topics like driving bugs and reviews. I think they get tedious sometimes but still are valuable. 19:52:50 Just gives us a time where we can block out and get maximum rapid feedback on the immediately-important topics for the project. 19:53:12 yeah. I think things like the Ci/cd reports don't contribute much value though 19:54:08 tchaypo: Indeed, those seem to be socialized well enough as passive reports and email summaries. 19:54:42 tchaypo: the CI regions _do_ need more love though. So keeping people abreast of the needs there seems like an immediate feedback situation. 19:55:50 Also I expect jerryz to want to discuss his issues as he brings up his region, so alt meeting time might be better for him to discuss that with those who are not in NorAm. 19:56:02 I. 19:56:07 tchaypo: anyway, I echo your concern that we're not using the time super wisely. 19:56:20 well the current favoured times cut him out completely 19:56:31 tchaypo: yeah so I think we can't let that happen. 19:56:37 There are a _FEW_ OpenStack users in China. :) 19:56:37 from memory 19:56:42 and devs. :) 19:57:11 yes :) 19:57:27 tchaypo: keep up the good fight. I suggest resurrecting the discussion in January. 19:57:31 if this meeting was about 2 hours later it would be better for Sydney 19:57:34 Everybody is checking out for holidays. 19:57:35 OH 19:57:40 the meetup! 19:57:42 and bej, I think 19:57:42 it's on 19:58:08 hotel group booking? 19:58:16 none 19:58:23 The hotels downtown are reasonable and many. 19:58:44 I suggest the Westin or Sheraton if you have that choice. 19:59:04 The Hilton (downtown, not airport) and Crowne Plaza are also close and nice enough. :) 19:59:14 I already booked in the Westin. 19:59:27 SpamapS: Did we get the possible scheduling conflict resolved? 19:59:30 Which is HP approved for my HP'ers. :) 19:59:37 This might be useful information to put in the etherpad. 19:59:38 greghaynes: entirely. I sent an email. Did it not make it to the list? 19:59:47 I think I just failed at reading email 19:59:54 got archive happy probably 19:59:57 why only 3 days this time? 20:00:01 bnemec: "Details to follow about access to facility. There are many hotels short walking distance from the office, including a Westin, a Sheraton, and a Hilton." 20:00:13 tchaypo: being mindful of peoples' time. 20:00:57 do we plan to use Monday and Tuesday for informal catch Ups? 20:01:13 or should we just turn up on Tuesday afternoon for wednesday mo 20:01:14 tchaypo: you can of course come early to LA and we can co-work in my basement if you want. ;) 20:01:19 morning? 20:01:30 Or come down to pdx :) 20:01:34 its only a 3hr train ride 20:01:36 i noticed that too, was going to confirm 18 wasn't a typo 20:01:37 ANyway we're over time 20:01:43 PLEASE confirm ASAP. 20:02:03 What's the mid-cycle RSVP level - still 16 out of 20? 20:02:05 We got pushback on a full week. 20:02:26 Geoff___: RSVP right now is 4 right now. 20:02:28 -> #tripleo 20:02:30 #endmeeting