21:02:22 #startmeeting project 21:02:23 Meeting started Tue Sep 4 21:02:22 2012 UTC. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:02:24 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:02:25 The meeting name has been set to 'project' 21:02:34 Agenda @ http://wiki.openstack.org/Meetings/ProjectMeeting 21:02:46 #info In particular we'll look into progress towards folsom-rc1 and Swift 1.7.0 21:03:04 #info Remember RC1 is a release candidate, so it will be your Folsom release unless we decide new bugs were found that warrant a respin 21:03:11 #link http://wiki.openstack.org/ReleaseCycle 21:03:24 #info I'd like us to produce our RC1s ASAP next week, and in all cases before September 18 21:03:36 so that they can see solid regression testing before final release 21:03:45 and we can all focus on documentation and migration doc 21:03:58 #topic Keystone status 21:04:03 heckj: hello! 21:04:08 #link https://launchpad.net/keystone/+milestone/folsom-rc1 21:04:18 All feature blueprints completed, a few bugs left targeted 21:04:33 we're looking pretty good 21:04:33 heckj: The two v3api bugs... are they still necessary ? aren't they Grizzly/Apiv3 ? 21:04:52 nope - need to move them out from that targeting 21:05:05 I can do that for you. 21:05:15 Anything else (not on the list) that should definitely be fixed before we can build a release candidate ? 21:05:49 I'd suggest bug 1022614 21:05:50 Launchpad bug 1022614 in keystone "Memcache token backend does not expire tokens" [High,In progress] https://launchpad.net/bugs/1022614 21:06:16 heckj: what do you think ? 21:06:20 * heckj looks 21:06:52 but the review was abandoned... 21:08:34 documentation related bug - issue was related to timezone offsets 21:08:50 (or at least that's the easiest fix) 21:09:02 woul dbe a good RC1 bug 21:09:19 ok, so maybe its status should be updated (and importance reset) 21:09:25 heckj: Looking at that list you seem reasonably close to publishing a RC1 ? 21:09:38 i.e. early next week should be ok ? 21:09:47 yeah, I think we're pretty close. We had a new bug come in today, but so far it's been nicely stable 21:09:55 heckj: anything else ? 21:09:57 right now, next week is looking good 21:09:59 nothing els 21:10:05 Questions about Keystone ? 21:11:43 notmyname: around ? 21:12:25 let's skip and go back 21:12:28 #topic Glance status 21:12:32 bcwaldon: o/ 21:12:37 ttx: sorry here 21:12:51 notmyname: argh 21:12:55 ttx: I'll wait 21:12:58 bcwaldon: ok :) 21:13:03 #topic Swift status 21:13:09 #link https://launchpad.net/swift/+milestone/1.7.0 21:13:19 #info So the milestone-proposed branch was cut earlier today... please test 21:13:21 1.7.0 is ready for QA testing 21:13:25 #link https://github.com/openstack/swift/commits/milestone-proposed 21:13:28 #link http://tarballs.openstack.org/swift/swift-milestone-proposed.tar.gz 21:13:38 notmyname: Any vague ETA on the QA sign-off on this ? 21:14:08 the goal is still sept 13, as it was. we simply cut it early so we don't run into QA scheduling difficulty like we did last time 21:14:17 OK. Do you have any blueprint implemented in that version that should be linked on this page ? 21:14:20 so that's the end date, but if it's done earlier, great! 21:14:25 Maybe https://blueprints.launchpad.net/swift/+spec/zone-concept-change ? 21:14:57 ya, I need to go through them and bring it up to date 21:15:17 #action notmyname to go through 1.7.0 features and target blueprints accordingly 21:15:25 notmyname: you confirm that this should be the final Folsom release for Swift, unless shit happens ? 21:15:34 (shit being a critical bug or regression found between now and final release that would warrant a 1.7.1) 21:16:13 correct. 1.7.0 should be the final release for folsom. bugs will be backported to 1.7.0 until QA has certified it. anything after that will only happen in folsom if it's a major security issue 21:16:20 or other critical issue 21:16:23 OK, that makes it like the equivalent of a folsom-rc1 for you. "Good for release until proven otherwise" 21:16:32 notmyname: anything else ? 21:16:41 correct (but I think we'll have at least one patch) 21:16:59 nothing else from me 21:16:59 Questions on Swift ? 21:17:44 #topic Glance status 21:17:48 bcwaldon: we are back online 21:17:52 ttx: hey there 21:17:52 #link https://launchpad.net/glance/+milestone/folsom-rc1 21:18:03 All feature blueprints completed, 12 bugs still targeted, almost all under review 21:18:14 Confident you can nail all those down this week ? 21:18:26 ttx: yep 21:18:36 ttx: just need help to get the reviews done 21:18:38 This one looked abandoned last time I looked: bug 1039815 21:18:39 Launchpad bug 1039815 in glance "v2 API should provide official openstack image property schemas" [High,In progress] https://launchpad.net/bugs/1039815 21:18:51 ttx: ah, it is abandoned 21:19:02 should move off-radar ? 21:19:04 that needs to be targeted at grizzler 21:19:17 ok, untargeting then 21:19:35 This one is unassigned: bug 1042823... you're on it ? 21:19:36 Launchpad bug 1042823 in glance "glance-control reload doesn't work with multiprocessing" [Critical,Triaged] https://launchpad.net/bugs/1042823 21:19:41 ttx: yes 21:19:56 bcwaldon: Looks all good. Anything else ? 21:20:09 ttx: nope, going well 21:20:17 Questions on Glance ? 21:20:55 #topic Quantum status 21:20:58 hello 21:20:59 danwent: hey 21:21:04 #link https://launchpad.net/quantum/+milestone/folsom-rc1 21:21:12 Looks like you finally got all the features in ? 21:21:18 yes. 21:21:24 we have two non-trivial reviews still in progress 21:21:30 one is the nova security groups stuff. 21:21:36 Cool, more focus on bugfixing, doc and extinguishing fires 21:21:40 vish +2'd, but we still need a second nova core reviewer 21:21:51 #help nova review: https://review.openstack.org/#/c/12173/ 21:22:06 the other is an L3 that we have core devs on, and am not worried about. 21:22:24 my goal was to be completely in bugfix/doc stage by this meeting. we didn't quite make it, but we're pretty close 21:22:35 Is the RC1 bug list current with known RC1 blockers ? 21:23:01 what I've been doing is that anything High or above is a blocker for RC1 21:23:15 danwent: Confident you can fix them all this week ? 21:23:18 (we don't allow any commits that don't have an RC1 targeted bug) 21:23:29 Given your team bugfix rate it definitely sounds doable. 21:23:43 i'm very confident we can fix all of the known issues now. Question mark is just around what else will pop up as we do more testing and documenting 21:23:55 indeed 21:24:03 danwent: Would you like to talk about bug 1023169 ? 21:24:04 Launchpad bug 1023169 in quantum "update nova to report quantum floating IPs" [High,In progress] https://launchpad.net/bugs/1023169 21:24:11 I also started email threads with you about floating-ip/nova stuff, and devstack stuff. we can handle that offline if you like. 21:24:18 this is related to the email I recently sent you 21:24:35 right. Maybe we can handle this offline and at the nova meeting 21:24:46 will likely drop this, unless you/vish feel we shoudl do it. Not clear it makes sense to do a half-way nova/quantum proxy for floating IPs. 21:24:50 that's the only orange flag I've on quantum 21:24:53 ttx: agreed 21:25:00 devstack gating is the other one 21:25:12 but we have an existing thread about that too. 21:25:26 #info orange flag with https://launchpad.net/bugs/1023169 21:25:28 Launchpad bug 1023169 in quantum "update nova to report quantum floating IPs" [High,In progress] 21:25:40 well, devstack gating is not really a release issue, more like a general QA issue 21:25:47 agreed. 21:26:03 Looking at bug triaging, you have 8 untriaged, but from known quantum dudes, so it should be a ballpark walk. 21:26:16 danwent: Anything else ? 21:26:26 ok, will take a look. was trying to get people to bug triage during review day… didn't work well :P 21:26:28 nope 21:26:33 Questions on Quantum ? 21:27:09 #topic Cinder status 21:27:13 jgriffith: o/ 21:27:17 #link https://launchpad.net/cinder/+milestone/folsom-rc1 21:27:17 howdy 21:27:25 All feature blueprints completed, 1 bug left targeted 21:27:32 Is this one the only known blocker for RC1 ? 21:27:41 Pretty much 21:27:46 The pylint bugs we'll see 21:27:54 so you should be ontrack for RC1 early next week 21:28:00 Yup 21:28:08 You should keep on triaging incoming bugs, make sure there isn't anything reported that should definitely be fixed before release 21:28:10 With the exception of the giant sync up 21:28:18 giant sync up ? 21:28:24 nova-volume/cinder 21:28:43 that's cinder -> nova-volume ? 21:28:49 Yes 21:28:55 jgriffith: https://bugs.launchpad.net/nova/+bug/1028092 21:28:57 Launchpad bug 1028092 in nova "Resizing a Xen instance with attached volumes fails" [Undecided,In progress] 21:29:01 ok, so mostly affecting nova 21:29:04 I would really like for that to get in 21:29:06 if possible 21:29:11 review for it is up 21:29:20 creiht: That's listed under NOva 21:29:23 :) 21:29:33 jgriffith: that can easily be fixed :) 21:29:37 * creiht waves a bunch of hands 21:29:45 jgriffith: want me to move it under cinder ? 21:29:45 Yes, but it fails Jenkins and has a -1 21:29:47 Needs fixed 21:29:54 ttx: I can do it no problem 21:29:56 BUT 21:29:59 jgriffith: yes renuka is working on it 21:30:16 well I want it to be in nova-volume, whatever that means for cinder is up to you :) 21:30:17 Renuka and I discussed whether it's feasible to get xen SM in cinder at this point or not 21:30:26 It's changed dramatically over the last week or so 21:30:40 creiht: Yeah, I know your position 21:30:48 :) 21:31:04 Of course, once we start moving to cinder, then I will worry about it ;) 21:31:07 Talking to Renuka it likely will be Grizzly for Cinder 21:31:42 jgriffith: if that is the case, then will cinder not be a bit for bit copy of nova-volume? 21:31:46 jgriffith: i'll move it to cinder and put the RC1 flag on it, just so that it's on release radar 21:32:03 then you can wontfix or defer it 21:32:14 creiht: can't put one by you 21:32:18 :) 21:32:25 heh 21:32:46 ttx: I'll talk to Renuka again 21:32:50 jgriffith: Anything else ? 21:32:54 ttx: If it's doable we'll update 21:33:03 not trying to stir things up... just that anyone running Xen SM will be stuck with nova-volume until it is available in cinder 21:33:07 ttx: The only other concern I have is snapshot deletes and migration 21:33:20 ttx: But not for this meeting 21:33:26 Questions on Cinder ? 21:33:57 creiht: This is way maintaining both SUCKS 21:34:03 s/way/why/ 21:34:13 #topic Nova status 21:34:18 vishy: hi! 21:34:26 ttx is in a hurry today 21:34:26 #link https://launchpad.net/nova/+milestone/folsom-rc1 21:34:33 but I step off my soapbox 21:35:01 creiht: I always fear that Nova might take a bit long :) 21:36:08 hi 21:36:19 vishy: So we are all set on the feature front ? 21:36:27 yes 21:36:28 There was a request posted last week by Nachi about metabridge driver... 21:36:29 features done 21:36:33 http://lists.openstack.org/pipermail/openstack-dev/2012-August/000844.html 21:36:38 i think that one is too late 21:36:47 danwent, opinion on this one ? 21:37:04 vishy: that's what I think too... unless it's actually an integration bug and not really a feature 21:37:06 i'm ok passing on it 21:37:20 vif-drivers can be installed separately if needed anyway 21:37:31 its a completely standalone module 21:37:35 ok. i'll admit not having understood what it was about, so I happily defer to your judgement 21:37:53 (note: i haven't looked at the patch, but if this is a standard vif driver, that's the case) 21:37:53 #info metabridge driver FFE denied, no more FFE anyway at this point 21:38:17 vishy: Looking at RC bugs, you have 19 targeted 21:38:26 Is that a full view of the RC1 blockers, or do you still have to go through the list ? 21:39:05 i haven't gone through it yet, but i would assume they are all blockers for nwo 21:39:08 (there are actually 21 now, I just added bug 1046020 and bug 1046040) 21:39:09 Launchpad bug 1046020 in nova "attach_volume fails for LXC due to wrong root_device_name and path" [High,Confirmed] https://launchpad.net/bugs/1046020 21:39:11 i will pare it down 21:39:11 Launchpad bug 1046040 in nova "Project-specific flavors broke Horizon flavor admin" [High,In progress] https://launchpad.net/bugs/1046040 21:39:20 which sounded like regressions to me 21:39:25 they are 21:39:46 vishy: There are a number of unassigned ones... I fear those might not get completed this week 21:40:28 vishy: +1 to paring down as we progress in the week. We'll review again at the Nova meeting Thursday 21:41:05 sounds good 21:41:10 Looking at triaging now... we are still at ~49, so not much progress in the last week 21:41:28 I'll try to go over some of them (again), but this is not good news if we couldn't summon enough goodwill in Nova to get that done in 3 weeks :( 21:42:16 #info orange flag: bug triaging still lagging behind, some issues might fly below radar 21:42:42 vishy: Anything else that keeps you awake those nights ? 21:42:51 nope just bugs 21:42:57 xml progress is going well 21:43:06 we should at least have base api and important extensions covered 21:43:32 I doubt we will have all extensions but we will at least know where we stand for the important stuff 21:43:48 sounds good. It will always be incrementally better than essex 21:44:12 wonder if we should have yet another "XML?" session at the design summit 21:44:45 or maybe an XML-themed party where "enterprise users" can try to convince us with cocktails 21:44:59 Questions on Nova ? 21:45:08 haha 21:45:17 i'm there 21:45:35 "XML night: enterprise users pay drinks" 21:45:39 +1 21:45:58 #topic Horizon status 21:46:04 'ello 21:46:06 gabrielhurley: hey 21:46:09 #link https://launchpad.net/horizon/+milestone/folsom-rc1 21:46:13 So quantum-public-network was merged ? 21:46:28 it was. turned out to be a pretty minimal change to Horizon and totally backwards-compatible 21:46:35 Awesome. So you're done with features now, looking pretty good 21:46:39 once it made it through nova and quantum it was easy 21:46:40 yep 21:46:45 found an important nova breakage today 21:46:50 vish is working on that 21:46:54 otherwise bugs are doing fine 21:46:55 yes, I targeted it earlier 21:47:00 I saw, thanks 21:47:02 7 Horizon targeted bugs left, all relatively low-prio 21:47:12 yeah, the rest are all nice-to-haves 21:47:28 gabrielhurley: You still look on top of triaging and targeting... Anything else you wanted to mention ? 21:47:47 testing of the quantum integration is appreciated 21:47:49 that's about it 21:48:16 seems like the following issues if floating a bit between nova + horizon: https://bugs.launchpad.net/horizon/+bug/1040956 21:48:17 Launchpad bug 1040956 in nova ""Unable to get quota information" in horizon when using quantum" [Critical,Confirmed] 21:48:34 not sure if there is agreement on the right place to fix 21:48:50 ^^that 21:48:50 gabrielhurley: Error: "^that" is not a valid command. 21:49:02 danwent: nice find 21:49:22 * ttx reads -- vishy: comment ? 21:49:25 my last understanding was that it was more of a nova-side fix 21:50:00 i think the "right fix" is either to actually proxy floating ips in nova to quantum, or to fix it in horizon by having horizon recognize that quantum does not support nova floating IP commands. 21:50:06 I just triaged it into RC1 to follow up on it 21:50:10 there's a third one-line trivial fix in nova, but its really just a band-aid 21:50:15 thanks 21:50:41 if nova decides not to fix it we can band-aid it in horizon 21:50:46 danwent: so it's linked to taht other floating IP/Quantum support issue, right ? 21:50:59 ttx: yes 21:51:30 ok, that's an identified orange flag, no need to raise another one 21:51:42 Other questions for Horizon ? 21:51:43 ttx: ah, its a slightly different bug 21:51:49 but close enough. 21:52:06 danwent: right, not same bug, but same release decision to make, I guess 21:52:12 agreed 21:52:45 #topic Other Team reports 21:52:51 annegentle: around ? 21:53:03 yep 21:53:14 annegentle: how is doc shaping up ? 21:53:34 My report hasn't changed much from what I sent to the mailing list last week. Lots of Qs over "our" three day weekend but not a lot of patches 21:53:42 #help More docs please. 21:53:52 I hope that people will gradually shift focus toward docs as RC1s go out 21:53:54 I'll send another weekly update tomorrow. 21:54:04 annegentle: quantum team is planning on drafting admin docs in google docs first, is there a major concern about that? 21:54:11 annegentle: one concern that was raised last week is about migration doc 21:54:25 danwent: I don't have much visibility to it if it's not in openstack-manuals, but I can take a look 21:54:27 idea is that its pretty easy for people to collaborate and merge their content 21:54:46 danwent: ditto for openstack-manuals :) I don't mind simple first though. 21:54:49 annegentle: how is that handled currently (migration) ? 21:54:52 annegentle: ok, will follow-up later 21:55:29 annegentle: at which doc level should we address that ? 21:55:31 migration sections exist for Keystone, say. Not really an overall "how to migrate" section done yet, though I did open a doc bug. 21:55:57 ttx: if "that" is migration it's probably for ops/admins first 21:56:16 annegentle: some suggested that we should maintain a lower-level MIGRATION text file in source code, not sure how good that would be though 21:56:49 ttx: oh I saw that on the -dev list and responded there. Honestly dev-level MIGRATION text isn't really what people are looking for. They want operator level stuff, unless I'm missing some concept. 21:57:04 no one really said an opinion on my response so far 21:57:13 annegentle: no, I'd agree with you. Also per-project doesn't really make sense there 21:57:59 since you might want to expain you should update keystone before glance or the other way around 21:58:10 annegentle: anything else ? 21:58:21 Any other team lead with a status report ? 21:59:40 nope, thanks for asking 21:59:47 #topic Open discussion 21:59:53 #info If you want to be elected PTL for openstack-common, Glance, Nova, Keystone, Swift, Quantum, Cinder or Horizon, remember to submit your candidacy before EOD tomorrow ! 22:00:27 In related news, I'll try to finalize the summit session suggestion site tomorrow 22:00:50 so we all start planning the design summit next month 22:00:54 Anything else, anyone ? 22:01:34 ok then... 22:01:37 #endmeeting