21:03:16 #startmeeting 21:03:17 Meeting started Tue Aug 7 21:03:16 2012 UTC. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:03:18 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:03:26 bcwaldon is on his way 21:03:28 Agenda @ http://wiki.openstack.org/Meetings/ProjectMeeting 21:03:31 o/ 21:03:39 #info One week left to propose, review and merge the last Folsom features! 21:03:58 do we have markmc for a 2012.1.2 update ? 21:04:15 otherwise we'll push that at the end 21:04:19 he's not online at the moment... 21:04:28 #topic Keystone status 21:04:35 heckj: then you're in 21:04:47 * heckj waves 21:04:49 markmc: keystone in, you're next 21:04:51 #link https://launchpad.net/keystone/+milestone/folsom-3 21:05:03 How is https://blueprints.launchpad.net/keystone/+spec/pki going ? 21:05:13 quite well 21:05:13 ttx, sorry, had network issues 21:05:30 good feedback from the community on revocation needs around the initial implement, getting nailed down in blueprint right now 21:05:47 ok good 21:05:47 we're not holding PKI back for it, but accelerating getting revocation ready to roll 21:05:53 About document-deployment-suggestions-policy, it's doc-only, right ? So we can postpone it to RC1 if not completed in the week ? 21:06:00 uncertain if we'll make the feature freeze deadline, but giving it a shot 21:06:09 ttx: it's doc only 21:06:18 Quick look at untargeted blueprints on the series plan at: 21:06:23 #link https://blueprints.launchpad.net/keystone/folsom 21:06:35 heckj: agree that prompt-for-password is actually a python-keystoneclient blueprint ? 21:06:43 ttx: yes 21:06:47 wil redirect 21:06:59 heckj: about https://blueprints.launchpad.net/keystone/+spec/iana-register-port ... I suspect it's still blocked at IANA ? 21:07:24 yes, they haven't responded. Not likely to happen at this point, so we'll be staying where we are :-/ 21:07:32 Should we defer to grizzly ? target to F3 and pray ? 21:07:49 grizzly is what I'm thinking 21:08:06 Final question: what's the status of APIv3 one week away from featurefreeze ? 21:08:25 some initial stabs at implementation up for review (thank you dolph!), not as much as we'd hoped. 21:08:51 heckj: is it worth pushing anything in, then ? 21:09:22 heckj: the grizzly branch should be open in a few weeks, once we nail RC1... 21:09:24 I don't believe so - everyone's strapped for time, but not a lot of progress on much other than PKI right now 21:09:54 heckj: agreed, looks like a Grizzly thing now 21:10:05 heckj: anything else ? 21:10:13 not from me 21:10:19 Questions about Keystone ? 21:10:44 #topic 2012.1.2 stable release 21:10:53 markmc: the plan is still to have Keystone/Nova 2012.1.2 on Thursday ? 21:10:58 so, the nova security fix we were waiting for has landed 21:11:10 and we've had ~10 more nova fixes and a couple of keystone fixes 21:11:22 so we have candidates up for testing, or are we waiting for anything more ? 21:11:25 there's one nova fix that I'm worried might be a regression, looking at that 21:11:41 the big change is that we've had ~10 fixes backported by bcwaldon for glance 21:11:47 aw yeah 21:11:49 most haven't landed yet 21:11:54 aw 21:12:03 it'd be nice to do 2012.1.2 with at least some of them 21:12:19 but thinking we should get the important ones landed today 21:12:21 markmc: we can make those land this afternoon 21:12:30 and send out a call for testing for nova, glance and keystone 21:12:34 and then release friday 21:12:44 giving another day for testing glance 21:12:57 pushing out another week would conflict with folsom-3 21:12:58 markmc: ok, so we should see clearer tomorrow, and call for testing candidates then 21:13:17 ttx, yep, hoping to send out the call for testing tonight 21:13:29 what level of testing ? is it unit tests by develoepr? 21:13:30 #info 2012.1.2 may be pushed back to Friday to include Glance 21:13:57 Ravikumar_hp: it's mostly a question of looking at proposed deliveries for things that would not get caught by our rigorous testing 21:14:06 like fucked up tarballs. 21:14:27 markmc: anything else ? 21:14:32 ttx: I promise, no fucked up tarballs this time 21:14:41 yayaya ... 21:15:01 ayoung: i'm honestly not sure how the client should call :5000/v2.0/tenants without initializing it as Client(token='asdf', endpoint='http://keystone:5000/v2.0') 21:15:40 ttx, nope 21:15:42 #topic Swift status 21:15:46 notmyname: hey 21:15:47 hi 21:15:53 You just released 1.6.0, so I don't have so many questions for you :) 21:15:58 :-) 21:16:00 notmyname: should I create a tentative 1.6.1 milestone to target new things against ? 21:16:05 (like for 1.5.1 we can always rename it if you change your mind wrt numbering) 21:16:17 yes. 1.6.1 for now 21:16:28 #action ttx to create 1.6.1 milestone for Swift if nobody beats him to it 21:16:41 notmyname: anything you wanted to mention ? 21:16:48 ya, I have one question for you 21:17:10 I think we have time for about one more release for folsom 21:17:21 when is that scheduled? 21:17:40 notmyname: definitely. Final is scheduled for Sep 22 21:17:55 I expect to have RC1 for the other projects around Sep 6 21:18:06 ok, thanks. that's what I was wonderign 21:18:40 So I'd say you can plan one between Sep 1-13 21:18:56 Err, release is Sep 27 21:19:03 ok 21:19:11 Sep 22 was Diablo. Good ol days 21:19:31 Questions on Swift ? 21:20:28 #topic Glance status 21:20:32 bcwaldon: o/ 21:20:36 #link https://launchpad.net/glance/+milestone/folsom-3 21:20:41 ttx: hey hey 21:20:50 Looking good, let's see in more details: 21:20:57 * https://blueprints.launchpad.net/glance/+spec/api-v2-store-access 21:21:07 What's the status here ? ETA for code being proposed ? 21:21:19 ttx: today/tomorrow 21:21:27 Ok... Is this the last step before we can mark the api-2 super-blueprint completed ? 21:21:43 yes 21:21:46 * https://blueprints.launchpad.net/glance/+spec/glance-deprecate-client 21:21:58 Is this one just a question of finding the 5min to do it, or is it blocked on something else ? (like glance-client-parity ? 21:22:00 that will be done today/tomorrow by me if nobody jumps at it 21:22:00 ) 21:22:04 its a 5-min thing 21:22:17 ok, then I think we look good here 21:22:26 On python-glanceclient side, is there anything blocking the core stuff ? 21:22:31 negative 21:22:34 its in great shape 21:22:40 it does have some outstanding ssl-related reviews 21:22:44 that could use some love 21:23:22 yes 21:23:25 bcwaldon: Anything else ? 21:23:44 yes 21:23:58 I wanted to mention that I'm thinking through images api versioning 21:24:01 (surprise surprise) 21:24:11 and I will be sending out an email with my thoughts 21:24:22 suspense is now killing me 21:24:27 ttx: mind if i jump ahead in the queue, may be losing internet soon. 21:24:35 #bcwaldonisdone 21:24:37 vishy: ok, you're next 21:24:39 Questions on Glance ? 21:25:07 #topic Nova status 21:25:18 #link https://launchpad.net/nova/+milestone/folsom-3 21:25:26 vishy: Let's see progress on essential stuff: 21:25:29 ohai! 21:25:31 * https://blueprints.launchpad.net/nova/+spec/general-host-aggregates (jog0) 21:25:39 This one looks in danger because of lack of review activity ? 21:26:01 ttx: yes, although I think it is ready, just needs a +2 21:26:06 ttx: I got a few reviews today, but still need more reviews 21:26:14 the last patch in the series might not make it 21:26:27 #help nova-core needed to review general-host-aggregates proposed changes 21:26:35 jog0, ttx: but I think we can mark it complete without it 21:27:09 vishy: I agree 21:27:12 vishy, jog0: sure, maybe create a separate blueprint or bug to track the remaining work 21:27:20 ttx: cool 21:27:23 * https://blueprints.launchpad.net/nova/+spec/integrate-python-glanceclient 21:27:27 Is this one complete ? Or more code is needed ? 21:27:31 ttx: +1 21:27:40 I'm blocked on testing that guy 21:27:42 so frustrating 21:28:05 blocked on testing ? How can we unblock you ? 21:28:18 not use glance client three different ways in nova 21:28:28 its all on me 21:28:43 * vishy slaps bcwaldon 21:28:50 :( 21:28:58 bcwaldon: should still be in in time for F3 ? 21:29:16 yes 21:29:27 I dont want openstack folsom to use legacy glance client 21:29:35 wouldnt be the end of the world, but t'would be better 21:29:58 hmm right. We would probably have to grant an exception for that 21:30:07 but would definitely prefer if it got in in time 21:30:15 Others may just already be complete: 21:30:21 * https://blueprints.launchpad.net/nova/+spec/extract-nova-volumes (remaining work is covered in Cinder blueprint ?) 21:30:31 jgriffith: progress on migration? 21:30:40 (or is there anything left to do on Nova's code side ?) 21:30:44 vishy: ttx: pretty much there 21:30:57 db and persistent targets added to cinder-manage 21:30:58 jgriffith: do you have changes to nova-manage? or is it separate? 21:31:09 vishy: I made them all in cinder-manage 21:31:21 great we can mark that implemented then 21:31:25 jgriffith: how's progress on docs? 21:31:29 Writing a script now to do the service switch over etc and doing some tests 21:31:31 vishy: will do 21:31:36 done 21:31:38 :) 21:31:39 annegentle: believe it or not, I am working on it :) 21:31:40 * https://blueprints.launchpad.net/nova/+spec/xenstore-metadata looks complete too 21:31:47 jgriffith: I do :) 21:32:09 as do https://blueprints.launchpad.net/nova/+spec/per-user-quotas and https://blueprints.launchpad.net/nova/+spec/update-flavor-key-value 21:32:21 yes and the plugin one 21:32:28 All 4 ? 21:33:06 novaplugins ? 21:33:59 vishy: still with us ? 21:34:04 yes 21:34:08 just marking them 21:34:10 heh, just checking. 21:34:16 andrewbogott_afk: ping? 21:34:26 looks like he is afk, I'm not sure if he had more to do on that one 21:34:30 vishy: you can update them off-meeting 21:34:33 marked the other three implemented 21:34:34 :) 21:34:35 A couple more questions... 21:34:40 * https://blueprints.launchpad.net/nova/+spec/disable-server-extensions : priority ? 21:35:03 i consider it essential for release 21:35:16 but the later ones can be done as bugs 21:35:41 vishy: mark it High and we'll do bugs 21:35:48 * https://blueprints.launchpad.net/nova/+spec/volume-usage-metering : was that started ? ISTR seeing it discussed somewhere 21:35:52 ttx: marked it, any that miss we will do as bugs 21:36:20 ttx: yes it was discussed, but i think the plan was to do it in cinder instead 21:36:26 haven't seen any props 21:36:27 Yes, that's correct 21:36:37 in Cinder in Folsom ? 21:36:42 or some Grizzly thing ? 21:36:45 ttx: yes Folsom 21:36:53 ok will redirect 21:36:54 ttx: But I haven't heard anyhting about it for a while now 21:37:01 #info In general, if anyone already knows a given feature won't make it, please mark its blueprint "Deferred" 21:37:14 there are more than one in that case 21:37:19 It will help others focus on stuff that is more likely to land. 21:37:25 vishy: Anything else ? 21:37:53 nope, we will have another meeting on thursday 21:37:57 mostly just need help with reviews 21:38:00 actually yes 21:38:27 need lots of help with reviews: 21:38:28 https://review.openstack.org/10726 21:38:45 is a big one that would be nice to get in 21:38:49 + the cells code 21:38:53 #help need help with reviews, including https://review.openstack.org/10726 21:38:53 yikes 21:39:03 yeah, was going to mention cells 21:39:17 vishy: you really want Cells in in Folsom rather than early Grizzly ? 21:39:21 would be good to have a bp to track it every week 21:39:25 #help need documentation on cells 21:39:57 #action comstud to create a Cells blueprint so that we can track it 21:40:20 ttx: early grizzly is probably fine for cells, although experimental support in would be nice for the third parties that are waiting on it 21:40:24 like mercado-libre 21:40:32 vishy: it's sufficiently separate ? 21:40:37 from the core code ? 21:40:42 they may be able to work off of trunk though so isn't super crazy 21:40:45 vishy, if only we had a way of marking things as experimental :) 21:40:53 ttx: yes, both cells and bare metal don't really touch core 21:41:00 and even if it's marked experiemental, people are going to want backports 21:41:06 markmc: yes we do! release notes! 21:41:13 heh, who reads docs 21:41:19 i do. 21:41:20 * markmc hides from annegentle :) 21:41:23 so not sure what experimental does other than set expectations ... doesn't reduce maintenance load 21:41:27 * ttx lies well 21:41:27 * annegentle finds markmc 21:41:31 * jgriffith ducks so annegentle can reach markmc 21:41:32 russellb, markmc: point taken, do you feel the same about the bare metal stuff? 21:42:01 i don't think something should go in as experiemental, if it's not ready, it's just not ready 21:42:01 vishy, haven't looked yet, but it does seem late 21:42:20 vishy, new drivers are slightly different, though - e.g. doesn't add API 21:42:22 russellb: +1 -- bracnhes can be used to track experimental stuff 21:42:27 baremetal stuff does have some docs on the wiki, so that was cool, i don't know about the code state yet 21:42:38 markmc: we have the unfortunate result of having a tremendous amount of work being done by multiple companies 21:42:50 markmc: who are trying to do their best to get into folsom 21:43:01 lets discuss it at the nova meeting/mailing list 21:43:04 Looks like a good topic for the Nova meeting 21:43:06 sure, sounds good 21:43:18 As long as there is no blueprint for it, it doesn't exist from my perspective 21:43:18 i don't think we should decide based on that it seems late, judge it by the quality of the code 21:43:26 and late blueprint I do despise. 21:43:34 but i don't think we should cut slack if there isn't enough time for legitimate feedback to be addressed 21:43:50 ttx: ok I'm done, you can move on! 21:43:52 Other questions on Nova ? 21:44:10 #topic Quantum status 21:44:16 danwent: yo 21:44:18 hi 21:44:20 #link https://launchpad.net/quantum/+milestone/folsom-3 21:44:29 btw, refresh if you haven't in the past half hour 21:44:29 Progress looks generally good, but I think it's time to land and complete stuff 21:44:48 yeah, today was our deadline for all 'features' being pushed to review.openstack.org 21:44:57 some of the stuff is not truly complete though, they are WIP branches 21:44:58 Ah, that explains it 21:45:02 OK, essential stuff review... 21:45:07 * https://blueprints.launchpad.net/quantum/+spec/provider-networks 21:45:13 Proposal is in, missing review: https://review.openstack.org/#/c/10938/ 21:45:17 It's the last part, right ? 21:45:37 yes 21:45:44 * https://blueprints.launchpad.net/quantum/+spec/quantum-v2-public-networks 21:45:49 Still blocked in review at https://review.openstack.org/#/c/9845/ 21:45:53 This is also the last part ? 21:46:12 checking the reviews, it looks like we're getting some input from people who weren't around when the original discussion happened on the ML :( 21:46:30 so I think we're going to simplify the code a bit to avoid the objection, and then merge it. 21:46:31 yay, ignore them :) 21:46:39 ok... 21:46:43 * https://blueprints.launchpad.net/quantum/+spec/quantum-l3-fwd-nat 21:46:53 "Needs code review"... Where is the review ? 21:47:02 I just pushed that today. definitely still needs polish, testing, and style cleanup 21:47:08 Is it the only part needed ? 21:47:15 I think jenkins missed it 21:47:22 yes, all code is there, just not in great shape 21:47:28 Link it here and I'll add it 21:47:36 https://review.openstack.org/#/c/10993/ 21:47:49 OK. In other news was wondering if https://blueprints.launchpad.net/quantum/+spec/quantum-notifications was actually completed ? 21:47:51 i'll add it, as I want to add the cli link too 21:48:13 yes, it was merged a few days ago, but not updated on LP 21:48:24 ok, will do 21:48:26 There are a number of folsom-3 targets that are not in the plan: 21:48:30 cisco-nxos-enables-multiple-ports, metadata-overlapping-networks, metaplugin 21:48:33 Should I just add them with Low prio ? 21:48:59 i actually updated their priority abou 20 mins ago 21:49:08 all med + low 21:49:21 danwent: did you set their series goal as well ? 21:49:27 =Folsom ? 21:49:28 yes 21:49:33 awesome 21:49:42 Finally, looking at untargeted blueprints in https://blueprints.launchpad.net/quantum/folsom : 21:49:43 (at least on two of them… i'll re-run ttx.py later to be sure) 21:49:53 https://blueprints.launchpad.net/quantum/+spec/add-basic-quantum-system-tests looks like it should be a Tempest blueprint rather than a Quantum one ? 21:50:28 yes, probably, though I think that BP is basically dead (no one has worked in it in months) 21:50:35 will redirect 21:50:37 What about https://blueprints.launchpad.net/quantum/+spec/quantum-security-groups and https://blueprints.launchpad.net/quantum/+spec/quantum-client-security-groups ? 21:51:04 can I remove them from the Folsom series goal ? 21:51:19 We're going to work on them very soon after F-3, but we just can't get them in for Folsom. We have a lot of testing + bug fixing to do as it is. 21:51:23 yeah 21:51:31 danwent: ok... Anything else ? 21:51:46 not that I can think of... 21:51:58 just going to need to organize the core devs to focus on the top priority reviews 21:52:08 yes, lots of review backlog now 21:52:08 and then test test test 21:52:12 Questions on Quantum ? 21:52:17 o/ 21:52:28 * ttx pauses 21:52:40 I just wanted to check on the doc plan of record - you're updating wiki pages and Diane's bringing it into "official" docs? 21:52:57 annegentle: that's what we're doing for the API spec. salv-orlando is leading that. 21:53:08 for admin docs, I was planning on doing doc-book directly 21:53:16 ok, and what's the plan for admin? 21:53:19 while pulling in chunks of text from others 21:53:20 heh I can't type fast enough! 21:53:23 :) 21:53:29 Other questions ? 21:53:34 i'll be the point-of-contact for admin docs 21:53:37 ok sounds good. Any WADL for the API so we could put it on api.openstack.org? 21:53:54 I can contact salv-orlando offline with the WADL Q. 21:53:58 i don't think we have it at this point. 21:54:00 yeah, that would be great. 21:54:00 I don't believe we'll have a wadl for next week 21:54:01 sorry 21:54:04 ok then 21:54:08 #topic Cinder status 21:54:12 jgriffith: howdy! 21:54:15 #link https://launchpad.net/cinder/+milestone/folsom-3 21:54:17 Howdy 21:54:26 We already touched it during the Nova part 21:54:30 * https://blueprints.launchpad.net/cinder/+spec/migrate-nova-volumes-to-cinder 21:54:42 How is this doing ? You told me yesterday it was almost done ? 21:54:49 I think we're ready to close that 21:55:03 salv-orlando: I just got a WADL for object storage last week so it's not time-sensitive other than missing out on being on api.openstack.org :) 21:55:19 jgriffith: as in.. propose code ? 21:55:26 or have it merged ? 21:55:33 Yeah, I should push the review today 21:55:49 ok, set to "Needs code review" status when done 21:55:50 The patch will only be db and targets 21:56:01 Everything else is so dependent I'm doing an example script 21:56:07 jgriffith: Anything else on your mind ? 21:56:08 ie switching the service 21:56:16 Lots of things up in the air 21:56:41 multi back-ends and boot from volumes 21:56:43 That's how the last week before FeatureFreeze should look 21:56:49 :) 21:56:53 Questions on Cinder ? 21:57:16 #topic Horizon status 21:57:23 gabrielhurley: o/ 21:57:23 ttx: hello sir 21:57:26 #link https://launchpad.net/horizon/+milestone/folsom-3 21:57:36 Still feeling confident about those 4 incomplete "High" blueprints ? What's the ETA for code proposal on them ? 21:57:55 as soon as I get through this quantum stuff they're next on deck 21:58:15 well, two of them *are* the quantum stuff 21:58:23 right 21:58:37 so code should be pushed for review in the next two days for those ? 21:58:49 I'd say end of the week is reasonable 21:58:56 ok 21:59:01 Looking at https://blueprints.launchpad.net/horizon/folsom ... 21:59:06 Can we assume that anything not targeted to folsom-3 (i.e. identity-backend-type, syspanel-tests and tenant-deletion) should be deferred to Grizzly ? 21:59:14 definitely 21:59:23 ok, will clean up 21:59:29 thanks 21:59:31 gabrielhurley: Anything else you wanted to mention ? 21:59:48 more eyes on the quantum code review would be awesome 21:59:54 #action ttx to create grizzly series for Horizon and defer non-targeted Folsom bps 21:59:58 also, help getting the devstack/quantum docs up-to-date 22:00:11 it's making reviewing/testing that code difficult 22:00:35 that's about it 22:00:41 Questions for Horizon ? 22:01:05 #topic Other Team reports 22:01:11 Any other team lead with a status report ? 22:01:21 o/ 22:01:24 I'll be quick 22:01:26 promise 22:01:30 annegentle: go for it 22:01:54 Reminder to please use DocImpact in your commit messages. And don't be surprised if I just amend your message with DocImpact added. 22:02:01 that's all 22:02:10 quick indeed 22:02:22 #info remember to please use DocImpact in your commit messages. And don't be surprised if I just amend your message with DocImpact added. 22:02:27 #topic Open discussion 22:02:31 Anything else, anyone ? 22:03:01 well the, 22:03:02 n 22:03:03 #endmeeting