16:00:12 #startmeeting interopwg 16:00:13 Meeting started Wed Apr 19 16:00:12 2017 UTC and is due to finish in 60 minutes. The chair is markvoelker. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:15 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:17 The meeting name has been set to 'interopwg' 16:00:23 #link https://etherpad.openstack.org/p/DefCoreRoble.20 Today's agenda 16:00:45 Hi folks, who's around today? 16:01:10 hello! 16:01:35 * markvoelker notes that eglute is away today and hogepodge is traveling 16:02:14 Hmm...actually we may have rather a lot of folks away today from the looks of it 16:02:57 ahhhh for dockercon, yes? 16:03:12 Yep, and a few folks on unrelated trips too 16:03:19 I think we may actually be short of quorum today 16:03:42 * mguiney nods 16:03:53 mguiney: while I have you here, how's the Glance stuff going? I saw Brian's question on the review and will answer shortly 16:04:02 I'm still picking through the scoring myself, but I'll get that done too. =) 16:04:29 o/ 16:04:40 sorry for missing two previous comments 16:04:45 meetings... 16:05:12 zhipeng_: believe me, I can sympathize 16:05:15 o/ 16:05:17 =) 16:05:25 :) 16:05:27 the Glance stuff is going failry well, i removed the capability that was not delivered in ocata 16:05:28 o/ 16:05:39 o/ 16:05:43 hello all! 16:06:35 o/ 16:06:53 markvoelker do you guys still need the input from Nova ? I know I'm behind schedule on this 16:06:59 markvoelker, did you see the tc logs from yesterday? Glance may be in trouble 16:06:59 I'm at a booth at Docker Con in Austin, so my reactions may be slow if I'm pulled away. 16:07:12 Ok, looks like we have enough folks here to be useful after all. =) 16:07:13 zhipeng_: I sent a patch up for Nova 16:07:20 hogepodge, the booth bunny! 16:07:38 hogepodge will check it out ! :) 16:07:54 So let me provide the feedback anyways 16:08:01 Ok, let's just walk through the agenda then. 16:08:06 #topic 2017.08 16:08:29 We have a number of patches up and a few outstanding. It's that time again where we really need to start finishing these up. 16:08:55 (I for one am behind due to a bout with viruses this past week) 16:09:45 So let's take a quick walk through the projects here and just do a quick check to see if there are specific points folks want to discuss now 16:10:14 first up, Nova 16:10:38 #link https://review.openstack.org/#/c/456753/ Nova scoring patch 16:11:09 For Nova 16:11:12 Hmm, feels like we discussed suspend/resume at some point in the past.... 16:11:16 again sorry for the late feedback 16:11:25 nova still needs a bit of work, but sean has signed off on Cinder so we should be looking at touchup reviews. 16:11:26 yes especially on that 16:11:33 Oh, right...we did last time. =p I see now. 16:11:44 Matt told me that suspend/resume should not be a priority 16:11:50 in his opinion 16:12:05 because not all in-tree drivers support it 16:12:19 he commented last time when shemail added it in 2017.01 16:12:20 yeah, that's right. can make a note of it. 16:12:25 but was ignored somehow 16:12:30 Yeah, that showed up last time around too 16:12:34 In that patch I also restored all of the previous scoring so it could be updated 16:12:37 (see https://review.openstack.org/#/c/385781/) 16:12:44 and flavor listing he does recommend :) 16:13:11 AZs he does not have a preference, and he suggest us to ask feedback from UC on that 16:13:14 if we got time 16:13:25 can we make a note of these in the review again? 16:13:29 he thinks AZs are common practices 16:13:34 sure no problem 16:13:39 I will add in review 16:13:48 zhipeng_: let me know if you want to take over on the patch 16:14:21 I think I could start with reviewing it since I'm still learning the process :) 16:14:27 zhipeng_: I think I agree with listing AZ's. My recollection is there's a fair amount of client/tool support for that too, but I'd have to go look. 16:14:43 markvoelker I agree 16:14:46 yeah, he is annoyed that AZs are in cinder but not in nova 16:15:02 (annoyed is the wrong word... surprised?) 16:15:14 the last suggestion I got from Matt is about the boot from volume 16:15:25 if we have tempest test cases for that 16:15:27 we could add it 16:16:16 zhipeng_: worth a look, but we need to move quickly to investigate. We're low on time. 16:16:45 I will take a quick study on that and provide a feedback in the code review 16:16:46 I'd say if someone can tackle that this week, great, otherwise maybe just tack a note onto the scoring sheet so we can investigate next time 16:17:10 I will try to do it this week 16:17:24 on one more thing, microversions 16:17:34 i remember we discussed it about three weeks ago 16:17:43 but can't remember the conclusions ... 16:18:11 maybe we could investigate microversions next time ? 16:19:31 ++ 16:19:41 zhipeng_: It's a broad topic, you'll have to remind of the context. =) Was this about defining a minimum microversion (which we probably don't want and is sort of implicit by the supported releases anyway)? Or about a feature exposed by a certain microvserion? Or...? 16:20:39 le tme try to find the quote from Matt 16:21:35 Anything new in Nova is on a microversion > 2.1. 16:21:51 so i think it is about a minimum microversion 16:22:28 Ok. So each Guideline we publish contains a list of supported releases, and to be included a capability must be supported by all of them. 16:22:59 That sort of implies that if a feature became available as a microversion in, say, Ocata, it's probably not a good candidate since the guideline covers earlier releases too 16:22:59 I think shade only assumes 2.0 because.... 16:23:16 okey understood :) 16:23:28 well that is all I got from NOva ptl 16:23:35 on defcore 2017.08 16:23:40 Ok, anything else we need to talk about re: nova scoring? 16:23:51 but at this point, supported versions have microversions 16:24:36 but feature reflected in microversion might not be popular yet. 16:24:57 Rockyg: sure, and shade is just one client to consider. 16:25:11 Ok, let's move on to Cinder 16:25:46 #link https://review.openstack.org/#/c/456440/ Cinder scoring 16:25:49 might need to make sure the minimum features of the release are reflected in user survey. so we can ge that info 16:26:33 I think the only real change here was moving upload to required from advisory 16:27:10 And there were also some changes in individual scores due to v2 -> v3 transition 16:27:15 For interop, the important ishould be the core features regardless of microversion support 16:27:36 Seems pretty straightfoward 16:27:49 Anything to discuss here? 16:28:59 Ok, moving on to Glance then 16:29:02 not really, just comments on anything i missed so I can clean it up 16:29:16 #link https://review.openstack.org/#/c/451167/ Glance scoring patch 16:29:36 i actually had one quick question about this 16:30:03 mguiney: sure 16:30:41 ah, it looks like you already got around to answering it in the patch comments 16:31:09 mguiney: =) Yeah, dropped that in while we were waiting for quorum. 16:31:36 ok, cool! well that answers my question, I suppose 16:31:44 * markvoelker figures there's a RAFT joke about acknoledged writes in there somewhere, but avoids it 16:31:55 heh 16:32:01 Ok, anything else on Glance folks? 16:32:18 well if that is something that does not require adjustment, I think the Glance scoring may be A-Ok 16:32:25 cool :) 16:32:32 ++ 16:33:34 Allright then: neutron 16:34:04 #link https://media.giphy.com/media/109LQSvunrcFNe/giphy.gif Neutron scoring 16:34:32 Unfortuantely I caught a nasty bug this week and haven't been able to clear my backlog of half-finished stuff. =( 16:35:05 :-( 16:35:32 Fortunately https://media.giphy.com/media/1cv7Gwf0Utdeg/giphy.gif so should be able to do it shortly. I have two patches started, one for mainline Neutron and one for LBaaS which will be a little more tricky. 16:35:48 I wish I could help, but I'm tied up all this week with projects in Austin. 16:36:49 hogepodge: Thanks. It's been "one of those weeks". Just glad I wasn't traveling when I caught it this time. =p 16:37:04 Anyhow, I expect to finish that tomorrow 16:37:53 get better! 16:38:24 On to Swift 16:38:29 #link https://review.openstack.org/#/c/453453/ Swift scoring 16:38:57 THis was a minor change in scoring with no new capabilities 16:39:42 Pretty uncontentious I think. =) 16:40:02 So unless there's something folks think was ommitted that needs to be there, let's skip on to Keystone 16:40:47 #link https://review.openstack.org/#/c/456774/ Keystone scoring 16:41:13 Again, a pretty light change here...just a scoring adjustment 16:41:54 We did have a note about writing a test for standardization in catalog keys...hogepodge I think you were on that, but assuming it'll keep for next time? 16:42:12 I need to write that test 16:42:21 I may suggest luzc include a note about it in the sheet just to remind us. 16:42:32 I have notes on it, and could possibly get something in next week 16:42:54 we could add a placeholder for advisory, and then fill in for the next cycle on testing 16:43:10 Ok, we're pretty low on time so if you can do it, great, if not we'll note it. 16:43:11 let me see if I can knock it out on Monday or Tuesday 16:43:31 that would be great 16:43:34 'k thanks 16:43:42 Ok, anything else on scoring? 16:44:21 #topic Name change 16:44:38 I pinged infra about the unreviewed name change patch 16:44:56 Basically they're aware of it and just waiting to round up a few other things for a maintenance window 16:45:12 In short, sounds like there's nothing we need to do other than hurry up and wait. =) 16:45:23 (which is good news) 16:45:56 Once that's done I think we'll be mostly complete 16:46:13 Anything else on the name change this week? 16:46:59 #topic Summit 16:47:21 We have an etherpad to solicit topics here: https://etherpad.openstack.org/p/InteropWGBostonSummit 16:47:34 PLease add yours. I'll start dumping in a few things from backlog shortly 16:47:47 There is a forum session for the program expansion too, so keep that in mind for topic planning 16:47:58 ++ 16:48:09 * markvoelker was just about to mention that 16:48:14 cool. I was gonna add that.... 16:48:42 * markvoelker is glad to note that the telepathy thing appears to be working fine despite us being all over the map this week 16:48:57 put in the place holder 16:49:21 There are also some other interop-related sessions in this week's etherpad for your scheduling enjoyment 16:49:50 Anything else on Summit? 16:50:50 kind of related 16:50:57 but also a different topic 16:50:59 hogepodge: go for it 16:51:01 v2.0 schema 16:51:15 I'll get some slides together and we can talk about the scoring talk, markvoelker 16:51:23 catherineD: was saying she needs something approximating a final version soon 16:51:48 hogepodge: thx for mentioning that 16:52:00 hogepodge: Ok, has refstack looked into the existing patch? 16:52:09 we have a summit session tiltled " Does my app or tool work on OpenStack Cloud X? Use RefStack to find out ... " 16:52:19 https://www.openstack.org/summit/boston-2017/summit-schedule/global-search?t=refstack 16:52:50 that was target to discuss the work around vertical and program certification usong schema 2.0 and RefStack ... 16:53:05 Basically I think the simplification I was investigating boils down to "we may be able to remove the components layer" but otherwise it's pretty similar 16:53:22 without schema 2.0 being merged this talk is not practical 16:53:44 and re schema 2.0, will Heat and Trove and others be one of the "addon" features for 2017.08 guideline ? 16:54:14 zhipeng_: to me that is the vertial thoughts with schema 2.0 16:54:34 markvoelker: hogepodge: am I close to correct to say so? :-) 16:55:18 catherineD yes it is, like do we have a timeline to look at for the vertical or the addons ? 16:55:30 an ther usage with schema 2.0 and RefStack is OPNFV 16:55:46 Well, there have been a few discussions about those...Heat is viewed by many as a potential addition to the existing Powered program. Trove was a candiate for add-on or a new component, but Tesora's acquisition kinda removed some of the impetus for that 16:56:07 NFV is a target for a vertical program, yes 16:56:40 Trove might yet be a good add-on candidate. THe other was Designate IIRC. 16:56:47 markvoelker, we should ask amrith about trove. Hes talking about putting it in maintenance mode 16:57:04 The OPNFV certification team is thinking about working with us to reuse the tool-chain that we already have 16:57:22 * markvoelker issues 3-minute warning 16:57:50 when we submited the session we were thiinking of these use case ... but I think we are ahead of our time with where we are now 16:57:59 yes, add ons for things like heat ant trove and designate 16:58:08 so do we have a timeline for these things to happen ? :) Are we looking at 2017.08 or later ones ? 16:58:20 that is my question ... 16:58:34 not sure whether cancel the session is an option .. 16:58:42 I mean the speaker session 16:59:21 catherineD, the foudation usually has alternative sessions for most tracks in case there's an issue with a chosen one 16:59:35 Probably later ones at this point, but we had planned to put together at least a skeleton to show to the BoD 17:00:03 Ok, looks like we're out of time--switch over to #openstack-interop? 17:00:12 #endmeeting