21:03:01 #startmeeting 21:03:02 Meeting started Tue Mar 13 21:03:01 2012 UTC. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:03:03 Useful Commands: #action #agreed #help #info #idea #link #topic. 21:03:24 Today's agenda: http://wiki.openstack.org/Meetings/ProjectMeeting 21:03:30 General progress towards RC1, for those interested: 21:03:39 #link http://wiki.openstack.org/releasestatus/rc1.html 21:03:45 (everyone likes graphs) 21:03:55 #topic Actions from previous meeting 21:04:05 * jaypipes to find volunteer for 934492: done 21:04:15 * vishy and ttx to discuss how to catch release-critical bugs 21:04:27 So far the RC bugs found us, so we didn't need to go a long way to catch them. 21:04:38 But I guess we still need to have a discussion on that. 21:04:50 * vishy to take leftover RC1 bugs (if any) to the ML for volunteers 21:04:59 All RC bugs are assigned, so we are good here. 21:05:11 * ttx and mtaylor to investigate actions around branches when we "open Folsom" 21:05:22 Investigation was done, I think the process is now mostly defined. 21:05:30 #topic Keystone status 21:05:35 heckj: o/ 21:05:42 #link https://launchpad.net/keystone/+milestone/essex-rc1 21:05:52 I put out a call earlier today to the ML that has a very nice response 21:06:01 14 RC bugs left (was 11 last week) 21:06:09 We are very behind on bugs, including some new ones found just last night, but they're all now assigned and moving 21:06:10 yes, that was cool 21:06:34 heckj: some bugs look a bit stuck 21:06:35 I'm afraid the other projects are quite ahead of us, but we're picking up steam to get an RC1 cut out the door 21:06:43 bug 950364 21:06:43 Launchpad bug 950364 in keystone "the owner field in glance is tenant_name" [Critical,In progress] https://launchpad.net/bugs/950364 21:06:54 * ttx refreshes 21:07:06 bug 924029 21:07:06 Launchpad bug 924029 in openstack-manuals "services.xsd has incorrect service type for Glance" [Medium,Confirmed] https://launchpad.net/bugs/924029 21:07:23 bug 944372 21:07:23 Launchpad bug 944372 in keystone "docs/middleware_architecture out of sync with auth_token" [High,Confirmed] https://launchpad.net/bugs/944372 21:07:39 and bug 942793 and bug 949373 look stuck on the outcome of the cfg discussion 21:07:40 Launchpad bug 942793 in keystone "keystone-all ignores all command line options" [High,Confirmed] https://launchpad.net/bugs/942793 21:07:41 Launchpad bug 949373 in keystone "Default configuration values are ignored" [High,Confirmed] https://launchpad.net/bugs/949373 21:07:56 Can the situation be unblocked for the last two ? 21:08:47 I believe so - markmc has a number of patches pending that should resolve them, and we'll move it forward with those reviews shortly. They might potentially impact the testing of keystone, so we'll walk through them carefully. 21:09:17 It's fundamentally about moving from a global config to a non-global - plenty on the list about that 21:09:29 heckj: based on the current list, what is your rough ETA for the RC ? 21:09:42 You have a good bugfix rate, but the issues keep on coming :) 21:10:33 My main concern is that in an ideal world, with Keystone being used by all the other pieces, we need it in RC state first 21:10:33 The flurry of patches from markmc is the biggest unknown. I'll be focusing on those this week, along with a few items assigned for me to fix as bugs themselves. I'd like to aim for friday for them all to be resolved, but I don't know what I don't know. 21:10:51 heckj: sounds good. 21:11:01 Does anyone have a bug that they think should be added to the RC list ? 21:11:07 Most of the bugs around keystone are *not* preventing it from being used by other projects, and those get absolutely critical immediate attention. 21:11:28 I've heard good anecdotal evidence that it's working well in some limited essex-4/rc testing as well 21:11:33 heckj: remember you can use the "essex-rc-potential" tag to mark non-release-blocking targets of opportunity, if you have any 21:11:46 ttx: thanks, will do 21:12:07 heckj: anything else ? 21:12:17 ttx: should doc bugs be consider blockers 21:12:25 anotherjesse: ttx good question 21:12:27 not from me. Questions? 21:12:36 anotherjesse: ++ 21:12:41 anotherjesse: yeah that's a good question. Depends on where the doc lands, I guess 21:12:47 * annegentle says if it's not documented it doesn't exist :) 21:12:51 I'll happily pull that off the blocker list if it isn't - have been assuming it was 21:12:55 annegentle: ++ 21:13:13 ttx: the keystone.openstack.org / nova.openstack.org docs are in the projects themselves 21:13:16 If the doc is included in the release deliverable, I'd say a basic level of doc is a blocker yes 21:13:36 but we could push changes to those after an RC is cut and still update the website ... 21:13:57 ttx: anotherjesse yep, I'd rather not have doc inaccuracies ship with code, but yes the web lets us fix faster 21:14:20 anotherjesse: so it boils down to what "basic level of doc" is :) 21:14:55 ttx: I'm okay with basic level being install and configure, anotherjesse what do you think? 21:14:56 anotherjesse: we can certainly complement the doc after the RC in various ways, but the doc shipped with the project should be accurate and kinda useful 21:15:19 keystone has good command-line help 21:15:42 annegentle: +1, anything required to run the code is a release blocker in my book. the rest is nice-to-have. 21:15:42 so at one point doc bugs should move to "essex-rc-potential" tag 21:15:51 we can probably talk about docs in general at the end 21:15:57 ok 21:16:02 Questions about Keystone ? 21:16:02 not really keystone specific 21:16:23 agreed 21:16:37 #topic Swift status 21:16:42 notmyname: o/ 21:16:45 good afternoon 21:16:50 notmyname: how is 1.4.8 going ? 21:17:06 good. here's the plan: 21:17:20 review and merge everything possible by the end of this week 21:17:33 make packages for milestone-proposed next monday 21:17:35 do QA 21:17:45 and release 1.4.8 on thurs mar 22nd 21:17:53 1.4.8 will be our essex release 21:17:53 notmyname: do you have new features planned, or only bugfixes ? 21:18:04 notmyname: (asking because some distros have asked me) 21:18:16 mostly bugfixes, but I hope to have object versioning in this week 21:18:59 notmyname: ok 21:19:06 plan looks alright to me 21:19:10 notmyname: please ensure docs are updated based on features/bug fixes 21:19:19 annegentle: indeed 21:19:33 Daviey: sounds like an FFe for object versioning if you want to ship 1.4.8 in Precise 21:19:38 * annegentle high fives notmyname 21:19:43 notmyname: Anything else ? 21:19:54 no, I don't think so 21:20:01 Questions on Swift ? 21:20:06 yes 21:20:14 bcwaldon: go ahead 21:20:27 Looks like there's a review in gerrit that needs notmyname's approval 21:20:28 https://review.openstack.org/#change,4893 21:20:40 chmouel noted it was a blocker for a bug in glance 21:21:05 it already has 5 +1's, so it shouldn't be difficult to review :) 21:21:24 bcwaldon: it's one of the several reviews that is partially reviewed (has at least +1 or +2) and needs to get merged this week 21:21:48 notmyname: yep, as long as this week is the goal 21:22:00 moving to Glance, unless there is more to discuss on that 21:22:05 not from me 21:22:06 I'd expect most of the outstanding reviews to land this week 21:22:13 ok, thanks! 21:22:17 #topic Glance status 21:22:24 jaypipes: yo 21:22:31 #link https://launchpad.net/glance/+milestone/essex-rc1 21:22:39 ttx: oy 21:22:45 5 RC bugs left (was 4 last week) 21:23:01 jaypipes: based on the current list, what is your ETA for the RC ? 21:23:17 (My main concern about Glance is the low bugfix rate, but the number of bugs mitigates that) 21:23:32 ttx: Sorry, was AFK.. Swift: 1.4.8 is really concerning me. I'm thinking 1.4.7 might be safer for our LTS. 21:23:39 ttx: I think no more than end of this week.. 21:23:42 ttx: just approved a fix for one of them and assigned the last unclaimed bug to myself 21:23:47 ttx: 3/5 bugs are keystone related 21:23:57 yeah me! 21:23:57 jaypipes: ok 21:24:07 ttx: and one is waiting for a second +2 (nullable name) 21:24:12 heckj: get used to it, we blame keystone for everything 21:24:18 jaypipes: you got it 5 minutes ago :) 21:24:27 jaypipes: bcwaldon +2'd it, but it failed on python 2.7 unit tests. 21:24:46 Daviey: your call, obviously 21:24:58 Anyone have a bug that should be added to the Glance RC list ? 21:25:19 Vek: should be a random cache thing... should re-run that 21:25:28 I just hit "retrigger" 21:25:49 coolio. 21:26:07 heckj: just about to push tests and a fix for https://bugs.launchpad.net/keystone/+bug/954089 21:26:08 Launchpad bug 954089 in keystone "service-list returns empty set for TemplatedCatalog backend" [Critical,In progress] 21:26:18 sweet! Thank you! 21:26:22 jaypipes: Anything else ? 21:26:22 welcome 21:26:38 ttx: nope. the dev/cdrom bug is really low priority. that's the only one I'd kick out of RC. 21:27:00 ttx: but... the good news is that Tempest against Essex is now passing all 128 integration tests... 21:27:06 w00t. 21:27:07 you can tag it with the magic tag if you want it to be in scope but optional 21:27:19 128? cool. 21:27:24 ttx: yeah, was going to do that.. also mark it low hanging.. 21:27:31 Questions on Glance ? 21:28:43 #topic Nova status 21:28:46 vishy: hey 21:28:50 #link https://launchpad.net/nova/+milestone/essex-rc1 21:28:51 hi 21:29:17 wow, 3 RC bugs left (down from 17 last week) 21:29:20 so we are very close 21:29:25 in fact I have one more 21:29:30 all of those were reported today 21:29:35 we were done for a second :) 21:29:43 yes, I saw that 21:29:58 I almost released RC1, but... no. 21:30:11 just added this one: https://bugs.launchpad.net/nova/+bug/953831 21:30:12 Launchpad bug 953831 in nova "when creating instance root image file,nova does not check backing file is ready.file is " [High,In progress] 21:30:20 just was verifying that there is an actual race condition there 21:30:27 Bug 954086 needs an assignee 21:30:28 Launchpad bug 954086 in nova "nova fails when requesting volumes using '--insecure' flag" [High,Triaged] https://launchpad.net/bugs/954086 21:30:29 should have a fix proposed soon 21:30:41 Who works on that one ? 21:30:50 ttx: yes, that one seems important but I haven't found someone to take it yet 21:31:05 bcwaldon: ^^ can you look at that? 21:31:31 vishy: ten-four 21:31:50 All remaining bugs look in good progress. 21:31:55 there are also three remaining rc potential bugs https://bugs.launchpad.net/nova/+bugs?field.tag=essex-rc-potential 21:32:22 So... my main concern about Nova is potential unknown showstoppers 21:32:31 vishy: At some point I think we'll need to go through the ~75 Nova bugs in "New/undecided" state 21:32:39 and prioritize them, to make sure we didn't miss an Essex blocker 21:32:41 ttx: If it works for an hr in devstack, we are solid, no? :) 21:32:54 Daviey: that's my metric 21:33:31 Daviey: jaypipes has tempest running against devstack and non-devstack deploys as well ;) 21:33:45 anotherjesse: That is great! 21:33:45 yup. 21:34:05 Does anyone have a bug that should be added to the Nova RC list ? 21:34:12 Test, test, test :) 21:34:18 ++ 21:34:24 ttx: jog0 will have one momentarily :( 21:34:26 ttx: I've been grabbing any bugs that come through on the ml and I haven't seen any 21:34:47 ttx: although it has been going super fast lately, I'm barely able to keep up 21:34:51 the bugmail has been rather daunting recently -- great work everyone 21:34:52 I also have a question: nova VSA support is broken what should be done about that for Essex? 21:35:11 jog0: probably just going to have to be broken 21:35:13 three options: ship broken code, fix it, delete it 21:35:19 vishy: ttx: do we consider the 8 doc bugs against openstack-manuals actually nova doc bugs? I don't think so but they are config doc items 21:35:19 what use is broken code? 21:35:50 bcwaldon: I gues my concern is ripping it out without accidentally touching something else 21:35:51 annegentle: if they need to be fixed in nova itself, yes 21:36:06 bcwaldon: clearly we have to leave the db tables in for example 21:36:18 I'm for ripping out known-broken..; though I don't want to make Twitter explode again 21:36:27 vishy: why should we leave db tables? 21:36:45 bcwaldon: because i think db migrations at this point is dangerous 21:36:50 vishy: +1 21:36:54 ttx: they are all configuration docs that don't exist anywhere that I know of. These docs could exist in nova but don't "have" to other than by the definition of "install and config are basic level docs" 21:36:58 even to delete a few self-contained tables? 21:37:23 bcwaldon: what is the value of pulling it out to that extent 21:37:26 annegentle: I think they should be non-blocking nova bugs then (tagged "essex-rc-potential") 21:37:29 ? 21:37:32 ttx: okay 21:37:43 vishy: I dont want to ship it and have users think it works for the next 6 months 21:38:12 vishy: but I agree we could leave underlying code and still get that 21:38:22 vishy: by removing apis 21:38:30 bcwaldon: we should discuss that on the bug itself. But I agree it needs to appear at the very least in our rc-potential list 21:38:41 need to move on 21:38:44 ok, we can discuss this later 21:38:46 bcwaldon, ttx: ok lets discuss offline 21:38:49 vishy: Anything else ? 21:38:55 jog0: you have another bug to target? 21:39:07 or is that what bcwaldon was referring to? 21:39:11 yes, but its small 21:39:18 nova/openstack/common/cfg.py: 'nova.api.openstack.contrib.standard_extensions' 21:39:36 jog0: file and you should get targeted 21:39:46 Questions on Nova ? 21:39:48 jog0: yes link me the bug and i will target if appropriate 21:39:50 it will be in shortly 21:40:08 #topic Horizon status 21:40:14 ohnoimdead: o/ 21:40:19 #link https://launchpad.net/horizon/+milestone/essex-rc1 21:40:20 \o/ 21:40:33 6 RC bugs left (down from 18 last week), all assigned 21:40:41 that number just fell :) 21:40:55 look, magic! 21:41:06 I wonder where those bugs are gone now. 21:41:09 yeah, we had a few we were hoping to get in but were not blockers (didn't know about essex-rc-potential 21:41:15 cool 21:41:26 I have no particular concern about Horizon, sounds well on track 21:41:32 gabriel is banging on his final few plus doc blueprint and will be done today 21:41:41 huzzah 21:41:44 andy has a review out for his with comments and jake's is nearly done 21:41:54 ohnoimdead: Anything particular in mind ? Any bug in the other projects blocking you ? 21:42:23 the big concern right now that has the most impact on horizon is the mysql connection bug which joe is well aware of 21:42:33 ohnoimdead: I have a TryStack member writing a Dashboard how-to to back port to Diablo docs, also hoping to get an Essex tutorial on Dashboard as well. 21:43:05 annegentle: I'm gonna have *lots* of docs forthcoming, but not for the RC1 21:43:12 annegentle: is someone on the hook for essex dash tutorial? 21:43:19 ohnoimdead: does that mysql thing have a bug number ? 21:43:31 gabrielhurley: ok, sounds good, let me know what types so I don't find other volunteers :) 21:43:40 annegentle: will do. 21:43:48 https://bugs.launchpad.net/keystone/+bug/943031 21:43:50 Launchpad bug 943031 in quantum "After MySQL connection failure (or timeout), first request reports MySQL Server has gone away, following requests work" [High,In progress] 21:44:05 ohnoimdead: yes, though I had to send him screenshots so he doesn't have a working Essex afaik 21:44:20 annegentle: um... :p 21:44:34 OK. Questions for Horizon ? 21:44:35 ohnoimdead: so yeah the diablo back port is more likely :) 21:44:56 annegentle: gabriel and i can probably put something together after rc1 21:45:06 it's a nice example of how TryStack helps with "makin' it real" 21:45:37 Moving on 21:45:40 #topic Incubated projects and other Team reports 21:45:45 danwent, troytoman: yo 21:45:54 hi 21:45:56 https://launchpad.net/quantum/+milestone/essex-rc1 21:46:10 still have a fair number of things targeted, but they are all in review except for a docs change 21:46:24 I think we're in good shape for our target to have an RC1 by tomorrow. 21:46:38 that's about it. 21:46:47 danwent: oh, ok. Ping me when you want it :) 21:46:52 great, thanks! 21:47:11 annegentle: More discussion on docs ? 21:47:35 ttx: two people have signed up for the remaining nova configuration doc tasks, six more needed 21:47:58 annegentle: what are the remaining tasks ? 21:48:03 I've pulled in all 485 nova flags to http://wiki.openstack.org/NovaConfigOptions 21:48:26 vishy: would be good to avoid adding new flags, now :) 21:49:41 I think I have prioritized the absolutely essential but with 485 flags I know there are still gaps 21:49:45 ttx: agreed 21:49:56 annegentle: could you link to the remaining nova configuration doc tasks, so that I can add it to the meeting minutes ? 21:49:59 ttx: which is a bummer because i thought we might need another couple 21:50:04 also the removal of -- in front of all the "things formerly known as flags" is going to be tough to merge 21:50:12 ttx: yeah was searching 21:50:26 vishy: you can still negociate with anne, but docteam needs to know when/if we add one 21:51:05 remaining doc priorities: 21:51:09 bug 953129 21:51:09 Launchpad bug 953129 in openstack-manuals "Docs need Essex overview of RBAC model and policy.json configuration" [High,Confirmed] https://launchpad.net/bugs/953129 21:51:22 bug 953137 21:51:22 Launchpad bug 953137 in openstack-manuals "Need docs for Configuring the ec2 compatibility api for Compute in Essex" [High,Confirmed] https://launchpad.net/bugs/953137 21:51:27 bug 953134 21:51:27 Launchpad bug 953134 in openstack-manuals "Docs need Essex info for configuring nova-api" [High,Confirmed] https://launchpad.net/bugs/953134 21:51:38 bug 953138 21:51:40 Launchpad bug 953138 in openstack-manuals "Need doc update for Essex for configuring Compute's storage system" [High,Confirmed] https://launchpad.net/bugs/953138 21:51:44 bug 953148 21:51:44 Launchpad bug 953148 in openstack-manuals "Metadata configuration needs to be documented in Essex" [High,Confirmed] https://launchpad.net/bugs/953148 21:51:49 bug 953151 21:51:50 Launchpad bug 953151 in openstack-manuals "Network configuration - Linux bridging, OVS, Multi-nic not well documented for Essex" [High,Confirmed] https://launchpad.net/bugs/953151 21:51:59 annegentle: those are openstack-manuals, or in-nova ? 21:52:13 ttx: those are against openstack-manuals but could be also copied to nova 21:52:35 annegentle: ok, coordinate with vishy if you need to add them to RC1 scope 21:52:42 ttx: ok will do 21:52:43 annegentle: anything else ? 21:52:52 also looks like anotherjesse had something? 21:53:12 annegentle: are you referring to the conversation in #dev 21:53:31 anotherjesse: no your hint at the beginning of the meeting 21:53:35 but we could push changes to those after an RC is cut and still update the website ... 21:53:36 o/ (would like to know the status of tempest) 21:54:05 annegentle: my assumption is we keep working on docs improvements even after the RC 21:54:08 I think once the very basics are covered, doc bugs should not block release 21:54:11 but some of those docs live in code 21:54:16 anotherjesse: absolutely 21:54:19 and get pushed to the websites if needed 21:54:28 (openstack-manuals, wiki, etc.) 21:54:35 so would we push a new RC if there were doc updates? 21:54:37 really I'm trying to target bare-minimum as highest priority I promise 21:54:55 anotherjesse: probably just a best effort to keep in mind 21:55:02 anotherjesse: that wouldn't warrant a reroll in my book 21:55:13 anotherjesse: but we would include them in if we rerolled 21:55:17 docs will have errors, code will have errors 21:55:21 i.e. "essex-rc-potential" 21:55:30 annegentle: !!! what !!! my code is perfect 21:55:35 except https://bugs.launchpad.net/nova/+bug/954377 21:55:36 mine is too. 21:55:37 Launchpad bug 954377 in nova "ec2 metadata service regressed for block-device-mapping and empty kernel-id, ramdisk-id" [High,In progress] 21:55:47 maybe a comment in the code-shipped doc segments that "the most up-to-date documentation can be found at " 21:55:51 Any other team lead with a status report ? 21:56:10 Vek: that might be good - add to the sphinx docs template 21:56:34 jaypipes: o/ (would like to know the status of tempest) 21:56:39 Vek: anotherjesse yeah that's in the Sphinx template for nova/swift/keystone at least 21:57:16 * ttx sneakily switches to open discussion while everyone is busy talking 21:57:20 #topic Open discussion 21:57:28 jaypipes: specifically, how is it coming towards being stable enough to run as a gate... 21:57:35 #info The sessions proposal website for the design summit is now open ! 21:57:40 See http://wiki.openstack.org/Summit for details 21:57:45 #link http://summit.openstack.org 21:59:14 heckj: looks like you'll have to ask jaypipes on another channel. 21:59:39 Anything else, anyone ? 21:59:43 yeah, I'll catch him later - not in the critical path for anything 22:00:25 ok then... have a good week, everyone 22:00:28 #endmeeting