21:02:36 #startmeeting project 21:02:37 Meeting started Tue Jul 30 21:02:36 2013 UTC and is due to finish in 60 minutes. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:02:38 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:02:40 The meeting name has been set to 'project' 21:02:43 #link http://wiki.openstack.org/Meetings/ProjectMeeting 21:02:55 Back in business after last week pause, we'll have a look at havana-3 plans and how reasonable they are (or not) 21:03:04 #info feature freeze is on *September 4* 21:03:15 Note that some projects implement a feature proposal freeze a few weeks before 21:03:28 #topic General stuff 21:03:41 I don't have anything specific 21:03:54 Oh. Yes I have 21:04:21 ttx, I'd like to repeat that we have planned stable/grizzly freeze on Aug 1st 21:04:21 If you have generic contraints you'd like applied to desig nsummit topics, like "no QA session at the same time as Infra sessions" please send them my way ASAp 21:04:33 and release 2013.1.3 on Aug 8th 21:04:47 apevec: heat has several outstanding reviews FYI 21:04:57 apevec: how is it doing so far ? 21:05:09 going* 21:05:32 adam_g has sent status yesterday, there were two exceptions discussed on the stable-maint 21:05:55 shardy, so you want to have heat 2013.1.3 ? 21:06:14 sdague, annegentle, mordred: news from QA/Docs/Infra programs ? 21:06:15 apevec: yes please, but we can discuss that offline :) 21:06:27 so the neutron job is in really bad shape 21:06:37 ttx: we're having fun issues with requirements - but we're working on making them more solid 21:06:38 sdague: we'll talk about Neutron testing in the Neutron section 21:06:47 or after the mtg rather 21:06:53 ttx: this is going to lead to more aggressive alignment on the requirements repo 21:06:56 We're holding Docs Boot Camp in Mountain View Sept 9 and 10, thanks to Mirantis for playing host. 21:06:59 #link wiki.openstack.org/wiki/Docs_Bootcamp_2013 21:07:04 ok, requirements chasing is getting interesting... not there yet 21:07:07 ttx: because doing otherwise is killing us 21:07:19 mordred, any more details? the issues, the changes? 21:07:19 +1 21:07:25 mordred: aggressivity++ 21:07:41 markmc: the issues have to do with the fact that pip installs wind up with state based on the arbitrary sequence 21:07:46 rather than the graphe of needs 21:07:50 ok 21:07:54 so the first thing that sdague and I are working on 21:08:02 yeah, think I saw something go by 21:08:02 fun 21:08:03 anneisgentle: any way to participate remotely? 21:08:07 is getting devstack to run the update.py script in setup_develop 21:08:28 so that we have the same current behavior as devstack, but aligned to what's in requirements 21:08:40 along with that we'll start gating changes to requirements on devstack 21:08:45 shardy: sorry, not at this point 21:08:57 then, we'll start auto-creating and proposing changes to projects on changes to requirements repo 21:08:58 anneisgentle: OK, thanks 21:09:02 shardy: we're always here for you :) Monday office hours. 21:09:08 because, well, we will have already tested that they work 21:09:21 ok, time to get into the specifics 21:09:37 #topic Oslo status 21:09:42 markmc: hi! 21:09:46 #link https://launchpad.net/oslo/+milestone/havana-3 21:09:46 howdy 21:10:01 it's all looking pretty fine 21:10:03 You have 8 tracked blueprints targeted to h3, which is pretty reasonable 21:10:04 apart from messaging 21:10:10 oh, 5 now 21:10:24 basically, oslo.messaging is in gerrit etc. 21:10:37 and now has working kombu and qpid drivers 21:10:46 not thoroughly tested, but the basics work 21:11:03 the big question is whether we can complete their testing and get e.g. nova switched over to using it 21:11:03 Most of your stuff is already under review, which is good news 21:11:20 we've set ourselves a deadline of August 16th for a go/no-go 21:11:26 sounds a bit risky to do it, and a bit risky not to do it 21:11:34 i.e. unless nova is using oslo.messaging at that point, we punt to icehouse 21:11:41 markmc: +1 21:11:44 not doing it shouldn't have an impact on havana 21:11:55 I think it's good to wrap up oslo changes a few weeks before H3 anyway, so that we have time to push oslo-incubator syncs at the end of the milestone 21:11:57 any messaging features we want for havana will go in through oslo-incubator anyway 21:12:15 the other big messaging one is the message security patches 21:12:19 which look in great shape 21:12:29 and the keystone KDS patches look pretty well advance too 21:12:37 so I'm fairly confident that will make it for havana 21:12:53 a fair bit of work to make nova etc. use it, though 21:13:13 ok 21:13:24 markmc: anything else you wanted to raise ? 21:13:32 it begs the question whether we should have e.g. nova blueprints for "use oslo.messaging" and "use oslo message security" 21:13:43 not just nova, all rpc-using projects 21:13:53 markmc: does that sound fair to say that oslo should be finished with its features a few weeks before feature freeze ? 21:14:02 ttx, yes 21:14:24 markmc: sounds like you want blueprints that affect multiple projects! 21:14:39 a blueprint with tasks for multiple projects 21:14:50 i.e. like launchpad bugs 21:14:53 markmc: your wishes may soon be granted. 21:14:57 heh 21:15:01 in the meantime? 21:15:14 in the meantime, tracking as a separate blueprint makes sense to me 21:15:14 would filing blueprints help track stuff? 21:15:16 ok 21:15:22 Questions about Oslo ? 21:15:37 thanks ttx 21:15:43 #topic Keystone status 21:15:49 henrynash: around ? 21:16:43 looks like henrynash is not around, jd__ you up ? 21:16:57 hum? 21:16:59 yep! 21:17:02 #undo 21:17:03 Removing item from minutes: 21:17:04 #topic Ceilometer status 21:17:07 jd__: hey 21:17:09 #link https://launchpad.net/ceilometer/+milestone/havana-3 21:17:40 You have 13 tracked blueprints, which is slightly optimistic compared to your usual velocity 21:17:44 It's also highly dependent on Eoghan and yourself, representing 9 of those 13 21:17:54 How are you both going to make it ? 21:18:16 we believe! 21:18:26 though one of my blueprint "Alarming" isn't a real one 21:18:36 it's just one final to track a bunch of others 21:18:54 I don't know if we can change it status so it's clearer for you to track 21:18:55 ok, we'll track that closely 21:19:22 use-new-rpc-messsage is marked Blocked. Could you explain what it is blocking on ? 21:19:43 ttx: yes, check its dependency graph actually 21:19:54 we're waiting for https://blueprints.launchpad.net/oslo/+spec/trusted-messaging 21:20:07 ok 21:20:10 stacktach-integration depends on logging-uses-notifications and remove-instance-fault-table, but those have been targeted to "next" 21:20:27 these are Nova bp I think 21:20:56 jd__: still sounds difficult to complete your ceilometer bp in h3 if nova's deferred on their side :) 21:21:01 all I can do is remove stacktach-integration, not sure sandywalsh will be happy 21:21:05 yep 21:21:08 russellb: ^could you look into that with sandywalsh and jd__ ? 21:21:15 maybe these aren't reall dep, I'll check with sandywalsh 21:21:18 -l 21:21:25 jd__: anything you wanted to mention ? 21:21:31 all good 21:21:47 Questions on Ceilometer ? 21:21:58 (We'll skip Swift since notmyname is not around) 21:22:13 #topic Glance status 21:22:15 not fair 21:22:21 markwash: busted 21:22:23 was almost done :-) 21:22:35 7 tracked blueprints so far, that sounds pretty reasonable 21:22:49 glance-tests-code-duplication still needs triaging :P 21:22:54 indeed it does 21:22:59 and there is one more that needs to be added to the list 21:22:59 About api-v2-property-protection, we've been carrying it over for the last 12 months now... 21:23:09 Is there any reason to think that it will hit in the next weeks ? 21:23:12 ttx but now people are working on it! :-) 21:23:25 there has been some activity on the list as well 21:23:26 markwash: I heard that one before I think :) 21:23:34 haha 21:23:42 markwash: so this time it's for real ? 21:23:51 this time, its personal 21:24:07 ok, I'll track that one closely :) 21:24:10 I'm not sure if it will make it, but there is active development and folks I trust 21:24:15 markwash: we pushed an advisory (OSSA-2013-018) for Glance python client today 21:24:20 markwash: how about we tag a new client release to get that out to the crowds ? 21:24:34 sounds good, will do 21:24:38 awesome 21:24:48 markwash: anything you wanted to mention ? 21:24:59 nothing from me, one question though 21:25:03 what is "Beta Available" ? 21:25:23 https://wiki.openstack.org/wiki/Blueprints#Implementation 21:25:32 "Implementation is almost complete, code is available in a branch or a draft review now" 21:25:39 okay 21:25:43 ty 21:25:46 Questions on Glance ? 21:26:05 Do we have markmcclain in the house ? 21:26:22 hi 21:26:25 #topic Neutron status 21:26:30 markmcclain: hi! 21:26:32 #link https://launchpad.net/neutron/+milestone/havana-3 21:26:45 * ttx refreshes ttx.py output, takes time 21:27:00 42 targeted blueprints... about twice as many as the total implemented in havana-2 21:27:08 This sounds a very unlikely goal, and could lead to issues prioritizing reviews and get the important stuff in 21:27:16 That said, 22 of those 42 are implemented or under code review already... 21:27:29 So I guess if you pulled an enormous code review effort early on, and pushed back a few of those 43 to "Low" priority... the picture could look a bit more doable 21:28:06 There are a few without an assignee in there, which in my book could have been plain refused: auto-associate-floating-ip, ml2-multi-segment-api 21:28:15 (you should require an assignee to be set before triaging a blueprint) 21:28:45 markmcclain: openvswitch-kernel-vxlan is marked "Deferred"... should it be set to "next" milestone instead ? 21:29:14 I need to doable check with the subteam working on vxlan 21:29:19 might remove it all together 21:29:28 You also have 12 "undefined" blueprints in there that need a priority. Ideally none of them would get higher than "Low" and add to the improbability above :) 21:29:46 I'll fix up the 2 unassigned that I missed 21:29:56 nachi, markmcclain, sdague: I wanted to talk about the Neutron gating tests 21:30:05 We made the neutron gate non-voting due to bug 1194026 21:30:06 Launchpad bug 1194026 in neutron "check_public_network_connectivity fails with timeout" [Critical,In progress] https://launchpad.net/bugs/1194026 21:30:13 I'd like to have those tests re-enabled ASAP so that they don't bitrot to the point where we can't use them anymore 21:30:18 What's the status of that ? Are we now hitting bug 1206307 instead ? 21:30:19 Launchpad bug 1206307 in neutron "neutron-adv-test.sh:773 [Fail] Couldn't ping server in Gating" [Critical,Confirmed] https://launchpad.net/bugs/1206307 21:31:25 sdague had to drop off -- markmcclain do you have news on that side ? 21:31:43 yeah.. I was chatting with nati_ueno yesterday about it and still have problems reproducing outside of the gate 21:31:55 I know that nati_ueno is actively working on the issue 21:32:01 both bugs are still active ? Or the first one was solved ? 21:32:18 the patch we pushed for the first one didn't fully solve the problem 21:32:25 so the bug was reopened 21:32:31 ok, so both bugs are active at this point 21:32:36 in the investigation he found the 2nd one 21:34:02 nachi: feel free to send me an update on your progress... and if you need any help 21:34:09 markmcclain: anything you wanted to raise ? 21:34:19 no 21:34:25 Questions on Neutron ? 21:34:52 #topic Cinder status 21:34:55 jgriffith: hola! 21:35:12 #link https://launchpad.net/cinder/+milestone/havana-3 21:35:42 jgriffith: yt ? 21:36:33 russellb: you there ? 21:36:35 hes riding a horse 21:36:35 yep 21:36:42 #undo 21:36:43 Removing item from minutes: 21:36:46 #topic Nova status 21:36:49 russellb: hey 21:36:51 hi 21:36:51 grrrr 21:36:53 #link https://launchpad.net/nova/+milestone/havana-3 21:36:55 #link http://lists.openstack.org/pipermail/openstack-dev/2013-July/012323.html 21:37:07 * jgriffith snoozes he looses his turn 21:37:15 so .... yeah. 21:37:28 i've deferred 10-15 blueprints or so 21:37:32 russellb: so looking at tracked (>Low) blueprints, you are at 51 21:37:41 compared to 25 delivered during h2 21:37:47 I suspect you'll need to push more to "Low" (a.k.a. "may land, who knows") priority 21:37:59 ttx: that's a good idea 21:38:03 easier than deferring 21:38:10 esp. when you don't know what might make it 21:38:14 i'm also expecting a whole bunch to miss the proposal deadline 21:38:21 When is your "proposed feature freeze" ? 21:38:25 aug 21 21:39:02 i will work on dropping priority of some 21:39:06 hadn't thought of that 21:39:18 it's an easy trick 21:39:29 but otherwise i feel like the best i can do is try to encourage review priorities, and wait to see what code actually shows up 21:39:34 basically helps prioritizing reviews on one side... and lower expectations on the other 21:39:45 true 21:39:48 image-multiple-location depends on effective-template-base-image-preparing which was not targeted to any milestone 21:40:15 new-hypervisor-docker is marked Blocked... what is it blocking on ? 21:40:23 oh, sorry for not updating whiteboard 21:40:27 he's adding devstack support 21:40:43 i wanted to have a clear path to being able to test it 21:40:53 ok, that is sort of implied in the current explanation, but wanted to be sure 21:41:04 wouldn't it be nice if we had tasks on blueprints ? 21:41:04 ok, yes, that's all it is 21:41:15 there's Work Items i guess ... 21:41:29 but not for multiple projects. 21:41:34 yeah. 21:41:39 russellb: anything else you wanted to mention ? 21:41:46 Oh, you may have missed the ping earlier 21:42:00 not really, just thinking about what we can do better to avoid this kind of situation in the future 21:42:08 i guess it's being more aggressive with using Low 21:42:09 :-) 21:42:09 russellb: ^could you look into that with sandywalsh and jd__ 21:42:18 i did miss that, my connection dropped for a bit 21:42:26 so i don't know what that's referring to 21:42:27 stacktach-integration depends on logging-uses-notifications and remove-instance-fault-table, but those have been targeted to "next" 21:42:38 stacktach-integration is a ceilo bp 21:42:45 the other two are nova's 21:42:50 oh 21:42:55 yeah, they hadn't started AFAIK 21:42:57 so i pushed them off 21:42:59 which made my inconsistency detector go bleep 21:43:15 the answer might be to defer the ceilo one as well 21:43:24 so either someone needs to yell at me to say "wait, i'm doing it!" or defer the ceilo one, yeah 21:43:29 could you look into it with sandywalsh and jd__ ? 21:43:36 my deferring triggered some yelling :-) 21:43:37 yes. 21:43:41 Any question on Nova ? 21:43:56 #topic Cinder status 21:43:58 :) 21:44:03 jgriffith:hi!! 21:44:03 hosre is back in teh barn 21:44:06 hola 21:44:12 #link https://launchpad.net/cinder/+milestone/havana-3 21:44:19 17 tracked blueprints, which is more than you delivered in h2 21:44:33 You might want to push some of these "Medium" to "High" or "Low" to help prioritizing the review effort, same suggestion than for russellb 21:44:34 Yeah, but this time we're going to hit them 21:44:48 I'll comb through next week 21:44:54 bare-metal-volumes and cinder-nfs-driver-qos need an assignee 21:44:56 most are in progress pushes from H2 21:44:59 (you should require an assignee to be set before giving a blueprint a priority) 21:45:08 those will likely be dropped due to lack of interest 21:45:14 cinder-volume-driver-optional-iscsi-support needs to be triaged too 21:45:23 but I need to give folks a chance to speak up tomorrow 21:45:32 sounds good 21:45:33 finally, db-cleanup (Medium) depends on db-archiving (Low) ? 21:45:47 quite frankly that whole thing is a bit of a mess 21:45:52 maybe adjust prio or remove link if it doesn't make sense 21:46:06 db-cleanup might be an "ongoing" task 21:46:11 I'll adjust, but I'm also not convinced any longer on the whole thing 21:46:15 unless it has clear objectives :) 21:46:23 good point 21:46:37 jgriffith: we have one Nova blueprint blocked on https://review.openstack.org/#/c/30974/ so please give that one some review priority 21:46:57 I have reviewed 21:47:04 jgriffith: anything on your mind ? apart from horses breaking out of the barn ? 21:47:14 haha 21:47:25 I'm concerened about encryption and db stuff 21:47:36 also the quota stuff 21:47:46 that merged into oslo before merging in Cinder or noava 21:47:48 nova 21:47:53 which seems backwards 21:47:56 other than that 21:48:02 nothing really 21:48:29 once the code is separated, it doesn't sound backward to implement improvements in oslo and sync them back 21:48:41 Questions on Cinder ? 21:48:49 sure, if the code works 21:48:53 anyway... 21:48:54 jgriffith: indeed 21:48:57 #topic Heat status 21:49:02 shardy: o/ 21:49:03 o/ 21:49:04 #link https://launchpad.net/heat/+milestone/havana-3 21:49:19 I know I need to start bumping (a lot) of stuff 21:49:22 * ttx refreshes 21:49:33 31 tracked blueprints, while you implemented 8 during h1 and 8 during h2 21:49:38 That no longer qualifies as optimism, it's pure madness :) 21:49:42 Not been ready to burst new-contributors bubbles just yet, but yeah 21:49:44 heh 21:49:57 You definitely need to more liberally use the "Low" priority to designate stuff that "may" make it 21:50:00 to be fair, a lot of stuff has been carried over, specifically the HOT functionality 21:50:11 That will help in communicating which features are the most likely to land to the rest of the world 21:50:13 which is all starting to come together, combined with the watch-ceilometer stuff 21:50:18 Because it's a quite unlikely those 30 will make it, and I prefer to be pleasantly surpised than doing bulk deferrals on feature freeze day 21:50:47 ttx: agree, I'll start getting it into shape after discussions at our meeting tomorrow 21:50:50 You also have a few waiting to have a priority set: 21:50:52 software-configuration-provider, update-stack-new-resource-state, heat-multicloud, oslo-db-support 21:51:17 ttx: that is deliberate - those appeared while I was on holiday and I need to discuss whether they just get bumped immediately 21:51:42 If you push things to "Low" they will disappear from my radar and i'll stop complaining about crazy goals :) 21:51:50 shardy: anything else you want to raise ? 21:51:53 * shardy sets everything to low 21:51:55 ;) 21:52:05 that would be cheating. 21:52:08 no, not atm, will try to get things more realistic for next week :) 21:52:16 great! 21:52:19 Questions about Heat ? 21:52:25 seem to have herded the cats into a corner atm ;) 21:52:39 #topic Horizon status 21:52:44 gabrielhurley: feeling better ? 21:53:05 somewhat 21:53:09 #link https://launchpad.net/horizon/+milestone/havana-3 21:53:13 16 tracked blueprints while you got 7 implemented in h2 (and 8 in h1) 21:53:21 So that's pretty optimistic... though you have about half of them implemented or in review already 21:53:22 yeah. I expect some will slip 21:53:29 there's a lot with priority "low" 21:53:39 So you might make it if you do a big review push early on 21:53:54 certainly possible 21:54:02 but probably optimistic 21:54:06 I will refine as we go 21:54:08 Has work started on https://blueprints.launchpad.net/horizon/+spec/rbac ? 21:54:26 david-lyle has it in his head. he may have started code, but he's generally on it. 21:54:45 ok 21:55:09 gabrielhurley: some will slip, but ideally those who slip should be Low ones :) 21:55:19 that's the goal 21:55:20 :-) 21:55:26 that's what the roadmap is about: setting expectations as good as we can 21:55:34 gabrielhurley: anything you wanted to mention ? 21:55:45 not off the top of my head 21:55:59 Questions on Horizon ? 21:56:21 Looks like we don't have henrynash, who was supposed to replace dolphm this week 21:56:35 anyone else feeling like defending keystone ? 21:57:08 ok, short update for the record 21:57:14 #topic Keystone status 21:57:31 6 tracked blueprints in havana-3, which is reasonable 21:57:41 endpoint-filtering needs an assignee (or be removed from havana-3) 21:58:03 #topic Incubated projects 21:58:10 hub_cap, devananda: news ? 21:58:14 * NobodyCam for devananda "we are still working on it. no real news yet" 21:58:20 heh 21:58:32 heyo 21:58:41 hub_cap: how is it going ? 21:58:53 no real news. ive got the rpm integration almost done as a POC 21:59:10 a guy from mirantis and someone else both offered to finish it up 21:59:27 so we shoudl be able to run on ubuntu or fedora images now 21:59:32 well, once the reviews land ;) 21:59:37 "Impl Cinder Volume extend for resize volume" is marked essential for your havana release, is it really essential ? As in "won't release without it" ? 21:59:38 and heat integration is next 21:59:51 well ttx, our cli tool and api already have it 22:00:01 and we impl'd it internally at rax, FAIL 22:00:11 ok. Assigning someone to that BP is probably a good idea then :) 22:00:20 and haomaiwang said he wants to do it i believe 22:00:22 ok, time is up, anything else you wanted to mention ? 22:00:24 ill make sure its assigned this wk 22:00:33 * hub_cap hugs everyone 22:00:35 that is all 22:00:37 #endmeeting