21:02:28 #startmeeting project 21:02:29 Meeting started Tue Sep 25 21:02:28 2012 UTC. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:02:30 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:02:31 The meeting name has been set to 'project' 21:02:39 Agenda @ http://wiki.openstack.org/Meetings/ProjectMeeting 21:02:49 2 days to final Folsom release! 21:03:06 #info We'll soon be in show-stoppers mode -- new RCs should only be triggered by release deliverables issues or annoying regressions with obviously safe fixes 21:03:38 #topic Keystone status 21:03:43 heckj: o/ 21:03:49 #info Keystone has a long-standing RC1 that might be a winner! 21:03:57 about to break it 21:04:07 bug 1056373 -> for an RC2 21:04:08 Launchpad bug 1056373 in keystone "memcache driver needs protection against non-string keys" [Critical,In progress] https://launchpad.net/bugs/1056373 21:04:09 hah :p 21:04:16 hah! 21:04:35 not even properly tagged so that it flies below my radar, nice 21:04:54 found and linked it up this morning - was a comment on a fixed bug elsewhere 21:05:16 heckj: business summary ? Why couldn't we fix it in stable/folsom ? 21:06:05 utf-8 broken-ness that none of us tested previously - actually intend to backport to essex as well - not an explicit regression so much as a nasty, nasty bug 21:06:40 I don't understand how users can trigger it? 21:07:11 soren: dolphm_ nailed it down this morning, associated review has tests to verify the fix 21:08:10 heckj: technically it could be fixed post-release but you've been pretty calm in the RC front, so... 21:08:19 that's it though - code review in progress, looking pretty good 21:08:24 heckj: how safe is that fix ? 21:08:29 ttx: yeah - would just prefer to get it fixed ASAP. 21:08:33 ttx: very 21:08:58 could easily do backports for resolving it if you'd prefer 21:09:19 heckj: ok, will open a RC2 window, and tag tomorrow morning at the latest 21:09:38 #info Keystone RC2 window with bug 1056373 in 21:09:38 Launchpad bug 1056373 in keystone "memcache driver needs protection against non-string keys" [Critical,In progress] https://launchpad.net/bugs/1056373 21:10:00 heckj: nothing else, right ? 21:10:07 ttx: process question - we need to fix in master and also backport into milestone-proposed, correct? 21:10:21 heckj: yes, push the fix to master then backport to MP 21:10:33 cool, thanks. Nothing else 21:10:36 heckj: I can handle that part if the fix is committed to master 21:10:46 will ping you just after meeting 21:10:51 Looking at release notes status now: 21:10:54 #link http://wiki.openstack.org/ReleaseNotes/Folsom#OpenStack_Identity_.28Keystone.29 21:10:59 heckj: No upgrade notes ? 21:11:53 ttx: i'll put some in there re: PKI tokens 21:12:01 Also, could you add links to bugs in the known issues section ? 21:12:06 yep 21:12:13 ok 21:12:15 How is planning for the Design Summit "Keystone" topic going so far ? 21:12:24 I see 6 sessions proposed over a total of 9 available slots, not bad 21:12:37 pretty good shape - 3 open slots, expecting some to come in last minute based on last summit. 21:12:40 core bits are all there 21:12:54 heckj: anything else ? 21:12:59 nope 21:13:03 Questions about Keystone ? 21:13:35 heckj: milestone opened @ https://launchpad.net/keystone/+milestone/folsom-rc2 with bug targeted 21:13:46 heckj: please land fix in master asap 21:13:58 #topic Swift status 21:14:03 hi 21:14:04 notmyname: hi! 21:14:08 #info Swift has 1.7.2 as Folsom candidate 21:14:13 notmyname: Everything still looking good on that side ? 21:14:24 mostly, but there is one possible thing that has come up 21:14:29 https://bugs.launchpad.net/swift/+bug/1055834 21:14:30 Launchpad bug 1055834 in swift "Memory Leak" [Critical,Fix committed] 21:14:36 Was wondering if that was a regression or something that's been around forever 21:14:43 this was just patched 21:15:02 ya, I think it's been around for a while 21:15:11 ie not a regression from some recent patch 21:15:23 then I'd say let's fix it in a later version 21:15:38 rather than rush a 1.7.3 for folsom inclusion 21:15:48 or 1.7.4 or whatever 21:15:53 notmyname: thoughts ? 21:15:56 of course, I'd prefer to have it the release so that I can remove the "Known Issues" ;-) 21:16:27 if people have been living with it so far, must not be that obvious of a leak 21:16:28 I don't know what's possible from your side, but from our side it's in master and vetted 21:17:21 notmyname: hrrm. 21:17:30 * ttx looks at the fix 21:18:04 I'm going to push for the fix to go in folsom, but we'll have a release soon after. either way, it will get fixed. it's more an issue of how soon people get it (ie if they only use the openstack 6-month releases) 21:18:58 Looks a bit dangerous to me 21:19:18 notmyname: distros can carry the patch if they want 21:20:16 notmyname: and due to swift using full versions we'd have to play tricks with versions again. 21:20:18 I don't really have a response to that :-/ 21:20:33 sigh 21:20:36 lol 21:20:58 I'll comment... Patch not added because it would be inconvenient 21:21:01 >:) 21:21:28 well, usually when we cut a release we have RAX QA go over it for 4 days 21:21:44 now you're telling me a one-hour-old patch is safe to ship... 21:21:58 well, it's actually already in RAX prod 21:22:14 ah. that counts 21:22:27 notmyname: ok, let's do it ... 1.7.3 ? 1.7.4 ? 21:23:09 notmyname: are you up to aligning the versioning on the different branches ? 21:23:19 sorry, lag here 21:23:53 ya, so 1.7.4 for milestone proposed and 1.7.5 for master. I'll include the version bump in the backport 21:24:03 notmyname: go for it 21:24:15 I can cut/release tomorrow morning if the branches are all set 21:24:32 another topic: 21:24:34 How is planning for the "Swift" topic at the design summit going so far ? 21:24:35 they'll be ready this afternoon 21:24:56 good. we have a few proposals so far, and I've got a backlog of stuff to talk about to 21:25:04 we're having a swift meeting next monday about it 21:25:05 http://lists.openstack.org/pipermail/openstack-dev/2012-September/001368.html 21:25:05 notmyname: ok, just send me an email confirming the commit ID for 1.7.4 and I'll make it happen 21:25:07 #link http://lists.openstack.org/pipermail/openstack-dev/2012-September/001368.html 21:25:19 notmyname: was wondering if we couldn't give the two Tuesday morning slots to the "QA" topic 21:25:26 given that "QA" topic is already over capacity (10 proposed for only 5 available slots) 21:25:33 Swift would have 8 session slots (instead of 10) and QA would have 7 (instead of 5) 21:25:43 would your schedule fit into that ? 21:25:55 ah. ya, let me see what's in the queue, but there is a chance of that. who should I coordinate with about that? 21:26:10 notmyname: me a jaypipes 21:26:14 ttx: you'll probably be able to steal two or so from Horizon, too, just FYI 21:26:15 s/a/and/ 21:26:22 ok, will do 21:26:26 gabrielhurley: it's more difficult to steal from horizon 21:26:34 because the topics are not contiguous 21:26:39 ah 21:26:48 notmyname: anything else ? 21:27:05 Questions on Swift ? 21:27:14 SF Bay meetup on Oct 11, if you're in the area 21:27:19 that's all I have 21:27:53 #topic Glance status 21:28:00 bcwaldon: o/ 21:28:06 #info Glance had its RC2 published earlier today 21:28:52 notmyname: created https://launchpad.net/swift/+milestone/1.7.4 21:29:01 looks like we don't have bcwaldon 21:29:15 danwent: around ? 21:29:29 yup 21:29:34 #topic Quantum status 21:29:43 #info Quantum had its RC2 published Friday and we have a RC3 already cooking 21:29:50 which should be published just after the meeting 21:30:01 danwent: Is there anything on the rc-potential list that we should add ? 21:30:06 #link https://bugs.launchpad.net/quantum/+bugs?field.tag=folsom-rc-potential 21:30:18 Don't think any of those is a show-stopper that needs to be fixed /before/ release 21:30:31 i'm not planning on doing anything for the zmq stuff for the folsom release 21:30:44 things still seem up in the air and rabbit + qpid are fully supported 21:31:06 just want people to be aware 21:31:18 two other issues are very minor 21:31:31 danwent: should we remove zmq from quantum completely ? 21:31:51 ttx/danwent: I believe only dhcp_agent won't work. 21:31:51 ISTR someone suggested that, haven't looked up the thread since meetign started 21:32:23 the other agents should work, but you might need to use an external matchmaker to reasonably scale more than a few nodes. 21:32:26 ewindisch: ok, so in theory someone could use the plugin agents with zmq, but not dhcp 21:32:27 (the matchmaker is pluggable) 21:32:42 danwent: yes - in theory, anyway. 21:33:23 I'm certainly okay with ZeroMQ marked as experimental, especially in relation to Quantum. 21:33:32 I would say we document that zmq is experimental for Quantum and note that in particular DHCP won't work. 21:33:38 danwent: sounds good 21:33:39 :) 21:33:52 danwent: so we are good to cut RC3 ? 21:33:56 yes 21:34:06 Looking at release notes status now: 21:34:11 #link http://wiki.openstack.org/ReleaseNotes/Folsom#OpenStack_Network_Service_.28Quantum.29 21:34:19 Same remark: you should add link to bug in the known issues section 21:34:25 And a bit more verbosity couldn't hurt. 21:34:35 ok. 21:34:38 Planning for the "Quantum" topic at the Design Summit... 21:34:44 I see 9 proposals over a total of 22 slots so far 21:34:54 basically, the core team hasn't even started proposing yet 21:35:03 yeah, that's what I suspected 21:35:05 they have (thankfully) been focused on folsom :) 21:35:08 danwent: Anything else ? 21:35:28 nope. after thursday, i'll start encouraging folks to propose sessions 21:35:35 (and people still asking me why the design summit track is not aligned in CFP with the other tracks 21:35:37 ) 21:35:53 Questions on Quantum ? 21:35:59 danwent; also worth noting that it needs the binary from Nova :( 21:36:14 jgriffith: around? 21:36:27 ewindisch: ok, can you send me a note with what you want in the docs about zmq? 21:36:32 ttx: howdy 21:36:34 okay 21:36:35 #topic Cinder status 21:36:42 #info Cinder got its RC2 out earlier today 21:36:48 Looking at the RC-potential list now: 21:36:53 #link https://bugs.launchpad.net/cinder/+bugs?field.tag=folsom-rc-potential 21:37:04 bug 1055401 21:37:05 Launchpad bug 1055401 in nova "[ietd] can not delete volume sucessfully" [High,In progress] https://launchpad.net/bugs/1055401 21:37:10 bug 1056246 21:37:11 Launchpad bug 1056246 in nova "using tgt-admin --conf option creates false sense of security, volume attach still fails after tgtd restart" [High,Fix committed] https://launchpad.net/bugs/1056246 21:37:28 Are those regressions ? Failures in new features ? 21:37:42 Regressions for the most part 21:38:06 Results of new changes that went in that weren't found in Ubuntu, but in Fedora test day 21:38:12 the --conf thing is a bit scary to me.. reverting it restores the bug that it was supposed to fix, no ? 21:38:27 ttx: Not really... that was my first thought 21:38:33 ttx: it didn't really fix it 21:38:43 ttx: It fixes it for one use case 21:38:47 ttx: (didn't survive accross tgtd restarts) 21:39:01 ttx: So that part is still all good 21:39:13 ttx: I took the conf file out a while back as it wasn't *needed* 21:39:16 so we restore the original (critical) bug ? 21:39:24 rather than half-fixing it ? 21:39:32 ttx: No 21:39:57 ttx: It's been a bit of a sorted tail... the original fix for the critical bug I dind't have the conf file option in there anywya 21:40:14 ttx, the issue is that distros/users need to modify the tgtd config file - /etc/tgtd/targets.conf 21:40:16 ttx: it wasn't even really a half-fix, just gave a false sense of initial security 21:40:18 It was pointed out that if we have the persist file we should use it in tgt-adm udpates 21:40:25 ttx, the "fix" we're reverting just masked that requirement 21:40:36 hhmkay. so you'd like to do a RC3 over those two fixes ? 21:40:42 ttx, I added some details to http://etherpad.openstack.org/nova-folsom 21:40:48 rather tha fixing them in stable/folsom ? 21:41:02 ttx: Those two and if we're doing one might as well add: https://review.openstack.org/#/c/13581/ 21:41:50 jgriffith: ok... I'll delay the decision on that one until we do Nova 21:41:58 ttx: fair enough 21:41:59 since that would also trigger a Nova RC3 21:42:12 The first two will as well FWITW 21:42:26 yeah 21:42:34 Cinder release notes at: 21:42:37 #link http://wiki.openstack.org/ReleaseNotes/Folsom#OpenStack_Block_Storage_.28Cinder.29 21:42:44 jgriffith: Not so much in there so far, planning to work on it ? 21:43:04 ttx: Yes I certainly am 21:43:16 Time is running short 21:43:18 Cinder topic at the Design Summit so far: 21:43:23 5 proposals for 7 available slots, looking good 21:43:34 jgriffith: Anything else before we switch to Nova ? 21:43:35 Yep, someobody's going to get cut :) 21:43:38 Nope 21:43:42 Questions on Cinder ? 21:43:46 vishy: around ? 21:43:54 yup 21:43:58 #topic Nova status 21:44:03 #info Nova got its RC2 out earlier today 21:44:23 #link https://bugs.launchpad.net/nova/+bugs?field.tag=folsom-rc-potential 21:44:40 if we trigger a RC3 due to Cinder... anything else you'd add to it ? 21:44:48 yes 21:44:59 in addition to the above-mentioned 3 ? 21:45:20 https://review.openstack.org/#/c/13649/ 21:45:23 that one 21:45:53 ok then, let's do another round 21:46:07 * ttx is in a good mood or what 21:46:10 I haven't found any other's that seem worth the risk yet 21:46:23 there are a few other good fixes but i think they are fine through stable/folsom 21:46:26 vishy: I'd rather have a very short list at this point 21:46:31 I marked them folsom-backport-potential 21:46:49 given that time to go through gate will quickly become a factor in borking the release 21:47:15 so let me open Cinder RC3 and Nova RC3 and target appropriately 21:48:22 jgriffith: https://launchpad.net/cinder/+milestone/folsom-rc3 <- please target your 3 bugs to it 21:48:42 vishy: https://launchpad.net/nova/+milestone/folsom-rc3 <-- please target your 4 bugs to it 21:48:56 or 5 if that last review really fixes two bugs 21:49:04 Let's have a look at Nova release notes: 21:49:09 http://wiki.openstack.org/ReleaseNotes/Folsom#OpenStack_Compute_.28Nova.29 21:49:18 Quite empty so far, more in prep @ http://etherpad.openstack.org/nova-folsom 21:49:27 vishy: Maybe time to move it over ? 21:50:36 ttx: soon. I want to add a bit more to fetures 21:51:02 vishy: oh btw anything we'd do wrt zmq in that RC3 ? 21:51:30 yes ignore it 21:51:33 :) 21:51:51 I'm fine with that if you are. 21:51:54 looks like the change didnt' go into quantum anyway, so we are leaving the old implementation in for folsom 21:52:02 Design Summit planning now... I see 18 proposals for 31 available slots 21:52:15 Would be good to spend some of the next Nova meeting(s) looking into what's missing 21:52:24 vishy: Anything else ? 21:52:28 ttx: agreed. I haven't even looked at it yet 21:52:55 we ahve two weeks between release and summit this time around, so not as much as a rush 21:53:04 Questions on Nova ? 21:53:18 #topic Glance status 21:53:25 bcwaldon: o/ 21:53:25 vishy, I added "zmq is experimental" to the nova release notes 21:53:31 #info Glance had its RC2 published earlier today 21:53:33 ttx: my apologies for lateness 21:53:38 No folsom-rc-potential bugs: 21:53:40 ttx: well 21:53:42 ttx: refresh 21:53:51 markmc: good call :) 21:53:51 * ttx sobs 21:53:56 ttx: eglynn has one 21:54:16 ttx: its not a dealbreaker, but it might help us get to critical mass for an rc3 21:54:21 eglynn: remind me not to buy you beers at the summit. 21:54:28 burn 21:54:40 * eglynn slinks off ... 21:54:50 * russellb buys eglynn two beers for having to take that from ttx 21:54:58 bcwaldon: that would be the only fix ? 21:55:06 ttx: so far, yes 21:55:28 ttx: and it's very minor 21:56:03 the only reason why I'd consider it is that it's less mainful to change config opts before release than after 21:56:10 painful* 21:56:46 it's a new option, right ? 21:56:47 ttx: ok, fair point 21:56:51 ttx: its new for folsom 21:56:53 rabbit_durable_queues 21:56:58 right, ok then 21:57:02 yep, new in glance, existing elsewhere 21:57:13 ttx: ok, maybe that raises priority 21:57:15 eglynn: it's fixed everywhere else ? 21:57:26 ttx only broken in glance 21:57:35 eglynn: ok then 21:57:44 eglynn: it was an attempt to get parity in glance with the rest of the projects, one minor oversight though 21:57:47 will open RC3 and target only that one to it 21:57:51 ttx: ^ 21:58:01 ttx: ok, sorry :( 21:58:06 Release notes status: 21:58:09 ttx: I'm just trying to keep you busy 21:58:11 #link http://wiki.openstack.org/ReleaseNotes/Folsom#OpenStack_Image_Service_.28Glance.29 21:58:13 ttx: thanks! 21:58:28 nothing like 6 respins on D-1 21:58:43 maybe 7, gabrielhurley hasn't talked yet 21:58:46 ttx: way to put things in perspective 21:59:08 Release notes could use a bit more verbosity, otherwise looks good 21:59:14 Also add link to bug in the known issues section 21:59:21 "Glance" topic at the design summit: 2 proposals for 5 available slots so far 21:59:26 bcwaldon: Anything else ? 22:00:09 ttx: no sir, I havent looked at those proposals, either 22:00:20 ttx: I'll shift into planning phase next week 22:00:25 https://launchpad.net/glance/+milestone/folsom-rc3 with one bug on it 22:00:31 Questions on Glance ? 22:00:50 #topic Horizon status 22:00:57 gabrielhurley: around ? 22:00:58 RC2 stands as good, no new blocking bugs; release notes will be pasted into wiki momentarily; summit track needs more work. ;-) 22:01:21 other PTLs: see ? that's how to do it 22:01:24 lol 22:01:33 ttx: don't you start 22:01:34 * heckj trips gabrielhurley 22:01:37 gabrielhurley: show off! 22:01:47 I have to sit through an hour of y'all talking! 22:01:52 blah blah blah 22:02:05 Questions for Horizon ? 22:02:05 <3 22:02:18 gabrielhurley: I love that I was way late but still got stuck back in rotation before you 22:02:18 gabrielhurley: show off :P 22:02:33 more difficult for heckj; he does not know what the exam will be about since he goes first 22:02:44 * gabrielhurley doesn't envy heckj 22:02:48 that's ok - I just make it up as I'm going along anyway 22:02:56 #topic Other Team reports 22:03:01 who needs planning and forethough 22:03:07 annegentle: around ? how is documentation going ? 22:03:18 * markmc has a stable/essex update 22:03:20 around 22:03:29 markmc: priority to the ladies 22:03:35 ttx, yep :) 22:03:42 why thank you 22:03:58 lots of good doc reviews going through. I'm doing my best to keep Quantum's builds going well. 22:04:09 The redesign of the docs landing pages is now done. 22:04:25 annegentle: good work on that redesign 22:04:25 Realized today I hadn't linked in all the docs.openstack.org/developer links 22:04:31 annegentle: so we are in reasonably good shape for release time ? 22:04:36 bcwaldon: thanks 22:04:43 even if it's always work in progress ? 22:04:47 ttx: I'd like some help with a scientific way to call docs "done" 22:04:58 ttx: the bug list is still over 100 22:05:07 ttx: and they're tagged and triaged quite well 22:05:22 so, I'm looking for the best backlog/task/bug management for docs ideas 22:05:32 because, we'll need to freeze at some point for translators 22:05:49 annegentle: yes, we need to discuss that. Doc / release interlock 22:05:53 also, I'm still seeking one more Documentation topic for the Design Summit - one presenter had to relinquish his spot 22:06:09 annegentle: maybe more suggestions will come up 22:06:19 that's it, thanks for listening 22:06:23 not a big deal if you have an empty slot though, gives flexibility in scheduling 22:06:33 markmc: go ahead 22:06:42 so, stable/folsom will open this week 22:06:52 Friday hopefully 22:06:54 and stable/essex will go into mothball mode 22:07:09 I'm hoping to do a 2012.1.3 release 22:07:10 * ttx googles 22:07:16 thinking 2012-10-11 22:07:21 I can't do it next week 22:07:24 latest status is: 22:07:32 nova: 10 fixes merged, 4 pending, 2 other potentials 22:07:32 glance: nothing merged or pending, 1 potential 22:07:32 keystone: 2 security fixes (critical/high), 3 others merged 22:07:32 horizon: 1 security fix (medium), 3 others merged, 2 more potential 22:07:35 .. 22:07:46 so all but glance already have stuff worth doing a release for 22:08:06 basically, I reckon this would be a nice way to close out stable/essex and move on to stable/folsom 22:08:50 markmc: I suspect some will pick up stable/essex branch maint 22:09:01 but yes, stable release wise... 22:09:06 ttx, right, "mothball" is too strong :) 22:09:16 markmc: markwash has a patch in review that is a major candidate for glance stable/essex 22:09:21 markmc: sounds like a plan 22:09:33 markmc: anything else ? 22:09:42 ttx, nope, thanks 22:09:42 Any other team lead with a status report ? 22:09:53 #topic Open discussion 22:10:00 #info TC elections are running until the end of day Thursday! 22:10:04 Anything else, anyone ? 22:10:11 bcwaldon, this one? https://bugs.launchpad.net/glance/essex/+bug/1012820 22:10:12 Launchpad bug 1012820 in glance "Diablo->Essex migration breaks Nova image_ref" [High,In progress] 22:10:19 ttx: backport to milestone/proposed for Keystone RC2 is up and processing: https://review.openstack.org/#/c/13667/ 22:10:20 markmc: yes sir 22:10:27 bcwaldon, saw your comment that you'll backport a fix alright 22:10:44 ok, let's make it happen 22:10:45 #endmeeting