21:02:08 #startmeeting 21:02:09 Meeting started Tue Aug 14 21:02:08 2012 UTC. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:02:10 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:02:15 Agenda @ http://wiki.openstack.org/Meetings/ProjectMeeting 21:02:28 #info Folsom-3 should be cut at the end of the day, so we'll look into remaining targets and defer/consider exceptions where appropriate 21:02:42 #topic Actions from previous meeting 21:02:47 * comstud to create a Cells blueprint so that we can track it 21:02:53 vishy: Can't find one yet. I guess this is Grizzly material now ? 21:04:10 ttx: yes appears so 21:04:19 #topic Keystone status 21:04:22 heckj: o/ 21:04:25 #link https://launchpad.net/keystone/+milestone/folsom-3 21:04:39 Not much apparent progress since last week... 21:04:50 document-deployment-suggestions-policy being doc-only can be postponed to RC1... 21:05:02 heckj: https://blueprints.launchpad.net/keystone/+spec/pki : Defer/Will be merged today/Exception requested ? 21:05:12 exception requested 21:05:25 heckj: ETA for that ? 21:05:32 based on feedback from the community, we really want to add in revocation support for the tokens - aiming to have it in within 1 week 21:05:53 heckj: how self-contained is this feature ? 21:05:59 very 21:06:10 so it doesn't impact "regular use" of keystone ? 21:06:37 nope - not even defaults. It's all for new use of PKI - default is the older token mechanisms 21:06:46 heckj: i'm fine with it, especially if it's the only one in Keystone 21:06:56 * https://blueprints.launchpad.net/keystone/+spec/ocf-support : Couldn't find code proposed... Deferred ? 21:06:57 yay 21:07:22 ttx: change was abandoned - so deferring it 21:07:28 heckj: Is that all the features that got in F3 ? Or is there anything we should retroactively create ? 21:07:57 #info FFe for pki blueprint, if it gets merged in one week max 21:08:00 That's really it 21:08:08 From the Folsom blueprints, I'll also defer iana-register-port to Grizzly. 21:08:21 heckj: Looking at F3-targeted bugs now, I see two bugs targeted... Are those really F3 publication blockers, or should they be removed from list ? 21:08:29 ttx: they denied us - so will be relooking at that blueprint entirely 21:09:02 (or pushed back to RC1 milestone) 21:09:05 ttx: not blockers for the release 21:09:24 can be easily kicked back to RC1 milestone - it's also out of the path for normal operation 21:09:24 ok, great 21:09:30 heckj: anything else ? 21:10:12 I need to follow up with mtaylor and jeblair, but we might experiment with a feature branch to continue work on the V3 API while we continue to stabilize keystone for grizzly release 21:10:30 you mean Folsom :) 21:10:36 sounds like a cool experiment 21:10:40 Questions about Keystone ? 21:10:41 yeah - sorry, Folsom 21:10:42 heckj: good idea 21:11:17 #topic Swift status 21:11:21 notmyname: hey 21:11:24 howdy 21:11:25 Great write-up to the list with recent features, thanks. 21:11:36 I created 1.6.1 milestone for you, no date yet. Let me know when you have an ETA. 21:11:45 thanks. it's a great group of contributors 21:12:01 ok. I have an FYI though 21:12:27 I can use that 21:12:33 based on a patch merged yesterday, and depending on what happens before the next release, there is a very good chance we may call the next one 2.0 21:12:51 oooh. 21:13:34 the patch updated the on-disk format of the ring (a great fix from swifterdarrell) but it may require the version updated to communicate that 21:13:42 notmyname: curious what the reason is 21:13:55 I'll probably send an email to the openstack-dev list about it 21:14:03 notmyname: sounds like a good plan 21:14:04 ha, looked up and hit enter, ill shh now 21:14:11 notmyname: anything else ? 21:14:30 nope. swift meetup on Aug 30 if you are in the bay area 21:14:38 #action notmyname to send an email to openstack-dev with 2.0 plans 21:14:42 Questions on Swift ? 21:14:51 s/plans/thoughts 21:15:24 #topic Glance status 21:15:28 bcwaldon: o/ 21:15:31 #link https://launchpad.net/glance/+milestone/folsom-3 21:15:41 Only one left: 21:15:49 * https://blueprints.launchpad.net/glance/+spec/glance-deprecate-client : Looks like we should wait for this one, should be merged today ? 21:16:06 or do you need slightly more time ? 21:16:09 if anyone can offer their opinion on https://review.openstack.org/#/c/11359/, we can get that bp in today 21:16:23 ttx: the code is there, just need to get it reviewed/merged 21:16:48 bcwaldon: ok, lets wait 21:17:01 #help Priority glance review at https://review.openstack.org/#/c/11359/ 21:17:08 hopefully somebody watching this meeting has a few minutes 21:17:13 Is that all the features that got in F3 ? Or is there anything we should retroactively create ? 21:17:25 that would be it, afaik 21:17:32 Looking at F3-targeted bugs now, I see two bugs targeted... 21:17:42 Are they F3 blockers, or more like RC1 targets now ? 21:17:47 rc1 21:17:56 ack 21:17:59 bcwaldon: Anything else ? 21:18:11 I wanted to mention something about the v2.0 Images API spec 21:18:30 go ahead 21:18:32 As Folsom-3 is our feature freeze, this is the point where we say "What's in the code is v2.0" 21:18:58 and I'm working on getting a nice markdown-formatted api spec for people to refer to 21:19:07 kewl 21:19:12 bcwaldon: nice! 21:19:29 Questions on Glance ? 21:20:05 #topic Quantum status 21:20:13 danwent: around ? 21:20:17 yup 21:20:23 #link https://launchpad.net/quantum/+milestone/folsom-3 21:20:28 Pretty long list of stuff under review here :) 21:20:31 still clearing out some bugs 21:20:41 so the way to think of it as there are 3 high priority things we're tracking 21:20:51 and anything medium or below "gets in if it gets in" 21:21:17 the provider net stuff is 90% in, we just keep lobbing off sub-commits 21:21:22 OVS support is all that is left 21:21:44 All the "needs review" here are expected to hit today ? Or do you need slightly more time ? 21:22:09 I wouldn't mind more time for our 3 branches that are high priority and above 21:22:38 but any of the lower stuff should either be merged tonight, or move on, in my opinion 21:22:45 to help people focus on key items and testing. 21:23:01 hmm, we'll see with Nova if it makes sense to delay F3 one more day to include a bit more stuff 21:23:18 I won't ask if there was anything implemented that isn't already part of this list :) 21:23:19 ttx: otherwise, the three big high priority branches can get merged 21:23:30 we'd just lop off smaller bugs to fix any remaining issues. 21:23:50 Looking at https://launchpad.net/quantum/+milestone/folsom-rc1 you targeted quite a few blueprints there 21:24:02 taht's in addition to the 3 you already mentioned, right 21:24:24 ack. two of those are testing infrastructure 21:24:33 and two are in review already 21:24:44 (but unlikely to hit today or tomorrow ?) 21:25:06 if we have an extra day, the ones in review may hit. I just haven't been super involved in the reviews, so I can't say for sure. 21:25:16 which ones are testing infra ? 21:25:26 "agent for test" 21:25:36 "v2 quantum versions of devstack exercise scripts" 21:25:51 basically, those are one item, but they are going into separate repos, so two BPs. 21:26:02 I'm fine with those hitting early in RC1 21:26:18 the L3 CLI stuff is in review, but needs unit tests written 21:26:23 especially the devstack one which probably doesn't land in Quantum anyway 21:26:30 multi-host and rootwrap are the more concerning ones. 21:26:36 yes 21:26:49 What about https://blueprints.launchpad.net/quantum/+spec/ovs-security-filtering ? 21:26:59 if it's really not started, sounds like grizzly to me 21:27:03 this would likely be a very small vif driver 21:27:13 if its about 200 lines of code, it won't go in :) 21:27:40 what about https://blueprints.launchpad.net/quantum/+spec/metadata-overlapping-networks ? 21:27:52 yeah, carlp mentioned that last week. 21:28:11 this is one of those "oh shit.." issues, as nova's metadata API assumes clients have unique ips. 21:28:35 hmm, could almost be considered a gap bug rather than a blueprint 21:28:51 Those all kinda make sense, but added all together it makes me fear for Quantum stability 21:28:53 yeah. we tend to create BPs for everything. 21:29:02 in times when we need some stasis 21:29:21 * ttx should reassign the last one since he is no longer working on it 21:29:34 agreed. as I mentioned, these aren't all for sure things we're looking for, more just things we're considering 21:29:48 i'll do a write-up on this and get your thoughts. 21:30:00 as I mentioned, rootwrap and multi-host seem the most possible to disrupt 21:30:05 danwent: sure, I'll look into them into more detail and talk to you 21:30:19 k, thanks 21:30:23 #action danwent/ttx to discuss how to reduce the nmber of FFe in Quantum 21:30:35 From the untargeted Folsom blueprints, should we just defer: 21:30:40 official-v2-api-spec 21:30:44 nova-quantum-interface-creation 21:30:49 isc-dhcp 21:30:50 ? 21:31:03 v2 spec is going to be done, but its not code, just using an item to track it. 21:31:09 all others are going to be deferred. 21:31:13 ok 21:31:30 On the bugs list, you should only keep F3 blockers 21:31:43 yup, i'm half way to cleaning it out 21:31:43 (those we'll need to fix and backport to milestone-proposed before Thursday) 21:31:56 Ideally all of those should have someone assigned to fix 21:32:03 agreed 21:32:06 Otherwise it doesn't look like it's going to happen 21:32:11 You can push back the others to Folsom-RC1. 21:32:17 danwent: Anything else ? 21:32:26 nope, back to hacking and reviewing :) 21:32:39 danwent: stay around, we migth discuss a one-day delay once we get to Nova 21:32:46 Questions on Quantum ? 21:32:53 ok. do me a favor and ping my irc handle? 21:32:57 sure 21:33:16 i'll be rebasing :)( 21:33:21 #topic Cinder status 21:33:28 jgriffith: around ? 21:33:34 Yep 21:33:36 #link https://launchpad.net/cinder/+milestone/folsom-3 21:33:51 This list grew quite a lot since last week :) 21:33:58 Yes, that it did 21:34:15 Are all those reviews going to land today ? 21:34:23 *should* 21:34:50 Most are close, minor fixes here and there 21:34:52 in other words, can we defer anythign that doesn't make it by the end of the day ? Or would you appreciate just a few more hours ? 21:35:07 Depends on when the end of your today is :) 21:35:14 A few more hours would be great 21:35:24 jgriffith: it's actually the beginning of my tomorrow. 21:35:42 also known as midnight Hawai time 21:35:43 I think everything that's in review now should be closed out by COB mountain time today 21:35:56 ok 21:36:00 https://blueprints.launchpad.net/cinder/+spec/cinder-notifications -> deferred ? 21:36:14 LOL 21:36:23 yes, that's the *one* that won't go probably 21:36:31 yes, deferred 21:36:36 netapp-volume-driver-cmode and create-volume-from-image got reviews merged, are they completed ? 21:37:04 yes 21:37:12 ok, willfix 21:37:13 In the untargeted folsom list, I see: 21:37:19 * https://blueprints.launchpad.net/cinder/+spec/implement-availability-zones 21:37:29 should I mark this one deferred too ? 21:37:41 That made it I believe... 21:37:54 * jgriffith checking... 21:38:16 Yes, that made it 21:38:29 linked stuff says "Partially implement blueprint implement-availability-zones" 21:38:30 https://review.openstack.org/#/c/11035/ 21:38:40 so I was kinda wondering 21:39:02 We can clarify that off-meeting 21:39:14 Only one bug targeted to F3, that's bug 1023311 21:39:14 k.. the second part was cinderclient which is going through now 21:39:15 Launchpad bug 1023311 in cinder "Quotas management is broken" [High,Triaged] https://launchpad.net/bugs/1023311 21:39:27 Is it really a F3 blocker, which will get fixed in master and backported to milestone-proposed before Thursday ? 21:39:37 Or should we just retarget it to RC1 to make sure it's fixed before release time ? 21:39:43 RC1 21:39:56 I've about got it, but honestly won't get back to it in time 21:40:07 Will add implement-availability-zones to F3 targets 21:40:13 jgriffith: Anything else ? 21:40:32 Nope, just trying to push the pypi of cinderclient 21:40:37 that's it 21:40:54 Questions on Cinder ? 21:41:18 jgriffith: moving to docs after f3? 21:41:24 ttx, jgriffith, mtaylor: can we get cinderclient pushed to PyPI? The entry for the package exists but no distributions were ever uploaded. It's non-existence is blocking the last Horizon blueprint (code's done, it just depends on a package that doesn't exist...). 21:41:26 annegentle: Yes 21:41:48 jgriffith: great, thanks. I think the original "doc plan" we outlined on IRC still works. 21:41:50 gabrielhurley: Working on it now, you should see it soon 21:41:54 jgriffith: thanks 21:42:03 #topic Nova status 21:42:06 vishy: hey 21:42:22 just in time 21:44:07 hmm, maybe we'll do horizon first 21:44:12 lol 21:44:22 sorry 21:44:22 #topic Horizon status 21:44:23 :) 21:44:23 heh 21:44:25 arrrh 21:44:27 ahaha 21:44:30 #topic Nova status 21:44:36 #link https://launchpad.net/nova/+milestone/folsom-3 21:44:36 tease! 21:44:53 vishy: this list is a mess :) 21:45:11 yes it is 21:45:12 ok, lets iterate through it... 21:45:17 * https://blueprints.launchpad.net/nova/+spec/general-host-aggregates : Implemented, right ? 21:45:21 yes 21:45:32 * https://blueprints.launchpad.net/nova/+spec/task-management 21:45:33 remaining work of converting availability zones will have to be in grizzly 21:45:53 maoy: comments on that ^^ are we complete for now? remaining stuff done in bugs? 21:46:20 * https://blueprints.launchpad.net/nova/+spec/config-drive-v2 21:46:31 vishy: it's a bit complicated. :) 21:46:46 implemented 21:46:50 marked 21:47:02 maoy: so do we defer it to grizzly then? 21:47:35 vishy: I'd revise the bp for folsom only and get a new one for grizzly 21:47:48 vishy: there is a race condition that should be fixed in f-4 21:47:54 maoy: ok can you revise it to what was done and mark it implemented? 21:47:58 maoy: unfortunately tere is no f4 :) 21:48:07 maoy: yes that can be reported as a bug 21:48:16 ttx: oops. my bad 21:48:21 maoy: and fixed for rc1 21:48:23 vishy: yes i'll update it 21:48:28 maoy: thanks 21:48:30 * https://blueprints.launchpad.net/nova/+spec/novaplugins 21:49:02 looks like defer to me 21:49:08 i still haven't heard from andrew whether it is done 21:49:11 he got stuff merged 21:49:15 but i don't know if it is complete 21:49:28 right, but no more code in 21:49:33 in all cases 21:49:48 * https://blueprints.launchpad.net/nova/+spec/deprecate-createserverext 21:50:04 ttx: so i'm not exactly sure how we should deprecate an extension 21:50:17 so i guess that one is kind of blocked? 21:50:28 i mean we can warn when it is loaded i guess 21:50:44 yeah I would add a DeprecationWarning 21:50:46 right now createserverext is exactly the same as /servers/ 21:50:55 so maybe we should just leave it and kill that blueprint 21:51:05 kill kill 21:51:17 * https://blueprints.launchpad.net/nova/+spec/xenapi-live-block-migration 21:51:17 yeah, probably best - not like it's much maintenance overhead 21:51:22 that should make it 21:51:28 vishy: today ? 21:51:34 i think so 21:51:44 I won't go over all the Low stuff... 21:52:02 can we apply the "what gets in is in, the rest is Grizzly" rule ? 21:52:18 ttx: i have a list of potential stuff to keep an eye on for ffe 21:52:24 https://review.openstack.org/#/c/11276 21:52:27 ';) 21:52:30 ttx: http://etherpad.openstack.org/nova-ffe 21:52:54 and I would also like a special ffe to put everything that goes into cinder into nova-volumes 21:53:06 we can drop it as one big chunk 21:53:23 vishy: would one extra day of review significantly help in getting more in F3 and less in FFe ? 21:53:37 what is the ruling there? cinder v nova-volume? 21:53:45 probably, there is a number of close stuff 21:53:48 I've been pondering delaying F3 by one day since there is so much stuff "almost there" 21:54:00 ttx: i buy that 21:54:36 Would be good to apply some priority treatment though... Concentrate on stuff that is more important first 21:54:44 danwent: would you buy that too ? 21:55:05 as long as its a targeted extension, only at high priority stuff, then yes. 21:55:08 my rationale on the dump from cinder -> volume is: there was uproar about removing nova-volume so we need to keep it and put in security fixes 21:55:16 * ttx kinda prefers to wait one day and get less FFes, so that stuff gets tested in F3 rather than randomly after 21:55:31 a) cinder is the one being gated on so we should keep the code the same 21:55:33 though i'm not sure how to enforce that, other than to actually branch, and just say that its OK to cherry-pick a select set of features over as well as bugs. 21:55:47 b) security backports will be way easier if the code is in sync at folsom release 21:56:12 ttx: do you buy that? 21:56:44 vishy: yes, from security pov 21:56:54 russellb: opinion ? 21:57:10 * russellb reads scrollback.. 21:57:36 #info F3 cut for Quantum/Nova delayed by one day to let a few extra prioritized reviews in, in an effort to limit the number of FFe 21:57:48 #info other projects may opt in 21:57:58 yes, does seem like a good idea to keep the code in sync, backports would def be easier 21:58:10 need to look closely at any compatibility/upgrade issues from the big dump though .... 21:58:13 ttx: is it just the branch cut that is delayed, or also the final release? 21:58:19 (of folsom-3) 21:58:24 also F3 milestone publication, yes 21:58:32 ok 21:58:32 would be done on Friday (god, I hate that) 21:58:53 vishy: Can I mark Deferred everything from https://blueprints.launchpad.net/nova/folsom that is not F3-targeted ? 21:59:01 ttx: yes 21:59:16 On the bugs side, there are 6 bugs targeted to F3... 21:59:24 Could we refine that list to include only real F3 blockers ? 21:59:42 Like "ZOMG I can't start an instance anymore" 22:00:14 pretty sure that qpid one is fixed, *changes it* 22:00:42 vishy: Anything else ? 22:01:08 ttx: may speak for a sec 22:01:09 i would suggest people doing reviews prioritize these: http://etherpad.openstack.org/nova-ffe 22:01:30 #help Nova Review prio to http://etherpad.openstack.org/nova-ffe 22:01:32 i just approved xenapi live migration 22:01:33 primeministerp: sure 22:01:44 would like to open up https://review.openstack.org/#/c/11276 22:01:52 for reprioritization 22:02:04 for review if possible 22:02:21 to a higher priority 22:02:38 3rd on vishy's list 22:02:42 primeministerp: it's on vishy's prio list I just linked 22:02:46 o 22:02:53 * primeministerp needs to refresh 22:03:00 ttx: thx 22:03:00 (row 12) 22:03:04 it's a good candidate for an exception if we don't get it in by the freeze 22:03:07 but let's be honest... there is a risk that it won't get in... it was proposed very late 22:03:17 but yes, that too 22:03:23 it's a lot of code to review.. 22:03:25 ttx: I understand 22:03:34 Frankly I was amazed by the number of last-minute reviews 22:03:41 not just yours 22:03:51 ttx: it was kind of amazing actually :( 22:03:57 It's so much easier to get your code in at any other moment of the 6-month-long cycle 22:03:58 yeah we kinda got slammed in the last week or two 22:04:13 Questions on Nova ? 22:04:17 however quite honestly it would be a major milestone and help our community from a single source of code perspective 22:04:26 * ttx tries not to overrun too much 22:04:31 give us the ability to file bugs properly 22:04:40 etc 22:04:54 ack 22:04:57 'tis on the exception candidate list 22:04:58 #topic Horizon status 22:05:10 gabrielhurley: thanks for your patience 22:05:12 o hai 22:05:14 ;-) 22:05:16 #link https://launchpad.net/horizon/+milestone/folsom-3 22:05:29 I'm so much happier with this milestone than I was last week 22:05:37 switch-to-cinder-client got done ? 22:05:46 it's merging right this second 22:06:02 * https://blueprints.launchpad.net/horizon/+spec/translation-documentation -- sounds like this is doc and can be pushed back to RC1 in case of need 22:06:08 yep 22:06:12 You have an enormous list of F3-targeted bugs, I suspect most of them are not F3 publication blockers... 22:06:20 all the wishlist stuff can be flat out deferred 22:06:27 Should I push back to RC1 anything that is not FixCommitted when I cut the branch ? 22:06:32 yep 22:06:36 awesome 22:06:46 of the remaining 8 there are no blockers, and most of them will be fixed today 22:06:56 it might be end of day tomorrow, so you have a bit more time 22:07:00 even better 22:07:11 gabrielhurley: Anything else you wanted to mention ? 22:07:14 woo hoo extra day! 22:07:25 nope. just a thanks to everyone for helping tear through the end of the F3 milestone :-) 22:08:12 Questions for Horizon ? 22:08:21 ohnoimdead: yay, thank me for another sleepless night 22:08:49 #topic Other Team reports 22:08:56 Any other team lead with a (quick) status report ? 22:09:51 \o 22:10:04 go for it 22:10:06 actually no, looking at the clock :) 22:10:10 I'll put it in an email 22:10:12 heh 22:10:18 #topic Open discussion 22:10:21 Anything else, anyone ? 22:11:25 #endmeeting