16:00:03 <thingee> #startmeeting Cinder 16:00:04 <openstack> Meeting started Wed Jul 15 16:00:03 2015 UTC and is due to finish in 60 minutes. The chair is thingee. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:05 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:07 <smcginnis> Howdy 16:00:08 <openstack> The meeting name has been set to 'cinder' 16:00:10 <geguileo> Hi! 16:00:14 <thingee> hi everyone 16:00:18 <DuncanT> Hi 16:00:20 <vincent_hou> Hi folks. 16:00:20 <winston-d_> o/ 16:00:20 <jseiler> hi 16:00:25 <scottda> hi 16:00:28 <thingee> #topic announcements 16:00:30 <eharney> hi 16:00:32 <Fdaisuke> Hello! 16:00:40 <dannywilson> hello 16:00:47 <srikanth_poolla> Hello everyone.. 16:00:54 <Swanson> Hello 16:00:59 <rhe00_> hi 16:01:02 <patrickeast> hi 16:01:08 <janonymous_> o/ 16:01:14 <thingee> #info hemna is continuing with the brick switch in nova, please help and review! 16:01:17 <thingee> #link https://review.openstack.org/#/c/175569/ 16:01:30 <jungleboyj> o/ 16:01:35 <cebruns> Hello 16:01:47 <thingee> #info Nova patches in flight for multi-attach support, please help and review! 16:01:48 <e0ne> hi 16:01:50 <thingee> #link https://etherpad.openstack.org/p/liberty-volume-multiattach 16:01:51 <tbarron> hi 16:02:00 <erlon> hi 16:02:02 <xyang1> hi 16:02:03 <flip214> hi 16:02:17 <thingee> #info Propose your topics for the Cinder Midcycle meetup! 16:02:19 <thingee> #link https://etherpad.openstack.org/p/cinder-liberty-midcycle-meetup 16:02:27 <thingee> you should see the header with "topics" 16:02:36 <ameade> o/ 16:02:46 <thingee> #action thingee to mention this in the mailing list 16:02:48 <hemna> yo 16:03:00 <jgriffith> o/ 16:03:03 <kmartin> o/ 16:03:07 <mtanino> o/ 16:03:21 <thingee> #info apache/nginx support in cinder api will be approved assuming no objections this week 16:03:23 <thingee> #link https://review.openstack.org/#/c/192683/6 16:03:43 <thingee> #info clone consistency group spec will be approved this week assuming no objections 16:03:45 <thingee> #link https://review.openstack.org/#/c/194834/4 16:03:50 <e0ne> thingee: thank you 16:04:13 <thingee> #info incremental backup enhancement spec will be approved this week assuming no objections 16:04:17 <thingee> #link https://review.openstack.org/#/c/187926/8 16:04:22 <xyang1> thingee: thanks 16:04:42 <thingee> #info os-brick 0.3.0 will be released this week 16:05:04 <smcginnis> +1 16:05:14 <hemna> nice 16:05:16 <vincent_hou> Cool 16:05:17 <thingee> #info cinderclient 1.3.1 is released with version discovery removed 16:05:18 <hemna> thingee, thanks 16:05:29 <thingee> alright lets get started! 16:05:35 <smcginnis> One other announcement in case folks don't keep up on the mailing list: M release == Mitaka now. 16:05:36 <thingee> Agenda for today https://wiki.openstack.org/wiki/CinderMeetings#Next_meeting 16:05:45 <thingee> smcginnis: yes thank you :) 16:06:00 <vincent_hou> Wow. 16:06:19 * jgriffith starts a pool on whether it changes again :) 16:06:20 <thingee> so I moved scottda's topic up because he has to cut out early 16:06:27 <thingee> #topic Cinder <-> Nova API fixes Update 16:06:29 <thingee> scottda: hi 16:06:32 <scottda> HI 16:06:32 <jungleboyj> smcginnis: Was so hoping for Musashi! 16:06:33 <vincent_hou> Sounds Japanese. 16:06:37 <thingee> #link https://etherpad.openstack.org/p/CinderNovaAPI 16:06:51 <Swanson> I vote for Miku. 16:06:54 <smcginnis> jungleboyj: Might get there yet. ;) 16:07:03 <scottda> So, I'm trying to corral all the info, bugs, design ideas, work in this etherpad... 16:07:19 <vincent_hou> Actually they all sound the same to me. :-) 16:07:28 <geguileo> jungleboyj: Me too 16:07:35 <hemna> scottda, I still have some work on my TODO list for this, mostly regarding live migration 16:07:38 <scottda> I've put some goals, just my ideas. Please comment if they seem crazy or if you can think of other goals that involve fixing cinder-nova 16:07:49 <scottda> I've listed ~20 bugs. 16:07:51 <hemna> scottda, I'm going to document the drivers that have potential problems with initialize_connection 16:08:05 <scottda> Not an exhaustive list, just that I got exhausted from searching. 16:08:11 <smcginnis> hemna: Add me to your list. :) 16:08:19 <scottda> I've put some design ideas and questions.... 16:08:28 <scottda> Please add and comment to that if you are interested. 16:08:36 <thingee> #action scottda is corral all the bugs, design ideas in etherpad https://etherpad.openstack.org/p/CinderNovaAPI 16:08:36 <winston-d_> scottda: appreciate your efforts 16:08:37 <hemna> smcginnis, heh ok. yah live migration is still a mess due to several things, but mostly because of how nova is calling cinder in some cases. 16:08:44 <jgriffith> scottda: wondering if we could take a small step back maybe. 16:08:47 <scottda> I think I'll put out a survey on design ideas prior to the mid-cycle 16:08:49 <jgriffith> scottda: so first, awesome job!!! 16:08:51 <scottda> sure jgriffith 16:09:00 <jgriffith> scottda: really great details and info here 16:09:17 <jgriffith> scottda: but I twinge a little when I see things like "assumptions are wrong" 16:09:21 <jgriffith> scottda: about "xyz" 16:09:43 <jgriffith> scottda: the reason being is that we have never clearly defined these things to say who's right/wrong 16:09:43 * scottda is looking for that part 16:09:55 <scottda> makes sense jgriffith 16:10:10 <scottda> I think one goal is to clearly define the API 16:10:20 <jgriffith> scottda: I'm wondering if we should start by proposing what the flow "should" be form Cinder's point of view (and in some cases why) 16:10:21 <hemna> having spent a great deal of time looking at this myself, nova is making some assumptions about cinder, what it can call and when, that just happen to not work for everyone. 16:10:32 <jgriffith> hemna: and there it is again :) 16:10:43 <hemna> yes, there it is again. 16:10:54 <hemna> I've been spending a great deal of time on this the last few weeks. 16:11:07 <scottda> That's Goal #4 jgriffith , but I don't necessarily care about the ordering. 16:11:16 <hemna> scottda, +1 16:11:19 <jgriffith> hemna: My point is rather than continuing to just say "nova makes assumptions", lets define what we want these interfaces to do/mean 16:11:30 <scottda> jgriffith: +1 16:11:34 <jgriffith> scottda: yes... so if I can finish :) 16:11:36 <hemna> jgriffith, sure, that's always been our plan 16:11:42 <thingee> scottda: if possible I would like to have a rough idea of a timeline based on volunteers what we can commit to in liberty 16:11:49 <hemna> but I think what we had to do first is understand what nova is doing right now and why it doesn't work. 16:11:59 <jgriffith> scottda: I think it would be ideal and much easier for all of us to help out here if we started by defining the interface/s 16:12:01 <scottda> first, jgriffith please finish.. 16:12:14 <thingee> scottda: of course not this entire list, but some movement instead of us coming with yet another etherpad on this subject. 16:12:27 <jgriffith> thingee: LOL 16:12:55 <jgriffith> scottda: so what I was trying to ask.... do *you* see any benefit in starting out by laying out the definition of the calls/interface first? 16:13:01 <thingee> scottda: that wasn't meant to come across what you're doing is a waste of time. I just want some people to commit to some work here. 16:13:04 <jgriffith> scottda: then proceeding from there? 16:13:22 <scottda> Sure. I figured that after gathering input we would discuss this at the mid-cycle. 16:13:34 <jgriffith> scottda: fi we did that, then I *think* it would be fairly easy to go forth and conquer 16:13:57 <hemna> you have to understand the problem points first before you can propose solutions. 16:13:59 <scottda> I imagine that there will be enough contention on things that we won't be able to come to consensus without being (mostly) face-to-face 16:14:02 <jgriffith> scottda: ahh.. yeah, I suppose that's not so far into the future at this point 16:14:12 <jgriffith> hemna: You have to have a definition to start IMHO 16:14:12 <thingee> scottda: I'm assuming we have communicated anything in the nova meetings yet because of how much yet has been defined? 16:14:26 <scottda> thingee: correct 16:14:31 <thingee> johnthetubaguy: ^ 16:14:50 <scottda> I had hoped that post-midcycle we'd have some concrete plan and consensus to take to Nova 16:15:24 <johnthetubaguy> so do reach out to me, if/when you want to get things reviewed in Nova, will do my best to make that happen for you 16:15:24 <jgriffith> scottda: Ok, that makes sense. Honestly it might be late for L to make changes in Nova... but I'd be curious depending on what we come up with how much we could just fix up on the Cinder side anyway 16:15:38 <scottda> johnthetubaguy: Great, I/we will do. 16:15:39 <jgriffith> scottda: but anyhow, that's a completely different disucssion 16:15:52 <jgriffith> scottda: Thanks for listening to me 16:15:56 <thingee> alright so who is going to lay out the the definitions of the calls/interfaces? 16:15:59 <smcginnis> jgriffith: +1, see what we can do, then prep for M in nova. 16:16:04 <scottda> I think it will take >= 2 cycles for all this to change (nova and cinder) 16:16:18 <hemna> I think it's 2 separate things at this point. 1) what can we do short term to make stuff work for L and 2) what do we do long term to fix the API and interaction. 16:16:23 <thingee> #info it maybe too late to get changes into Nova at this point, may prep for M 16:16:24 <jgriffith> thingee: I'm happy to do a first draft.. then hemna and scottda or anybody else can jump in and shred it up :) 16:16:37 <scottda> I tried to capture long and short goals under Goals 16:16:46 <scottda> jgriffith: that would be great 16:16:54 <thingee> jgriffith: can you commit to that be ready to show at the midcycle? 16:16:56 <jgriffith> thingee: I can start it based on what we did in nova-volume when most of those calls were created 16:16:58 <johnthetubaguy> trying to get a backlog nova-spec merged before the summit sounds like a good goal for that stuff (and getting a summit session if we don't agree in time) 16:17:03 <jgriffith> thingee: yes 16:17:07 <thingee> jgriffith: thank you! 16:17:14 <jgriffith> thingee: wait.. when is the mid-cycle? 16:17:16 <scottda> johnthetubaguy: +1 to summit session 16:17:17 * jgriffith jokes 16:17:18 <thingee> jgriffith: haha 16:17:41 <smcginnis> jgriffith: Make sure you get your travel arranged. :) 16:17:46 <johnthetubaguy> scottda: in a happy world, we will have agreed the plan by then, but if not, yeah, a joint summit session might be a good plan 16:17:55 <scottda> hemna: short term seems like bug fixes, long term is discussion, docs, and mid-cycle wrangling over issue IMO 16:18:01 <hemna> scottda, +1 16:18:05 <thingee> #action jgriffith will write a first draft of defining call/interface 16:18:05 <jgriffith> smcginnis: LOL... funny thing is I do have to do that... I'll be traveling all the week before ;) 16:18:09 <tbarron> jgriffith will travel by horse 16:18:20 <smcginnis> :) 16:18:23 <hemna> scottda, leeantho and I have been working hard on the bug fix side for live migration at least. it's a mess. 16:18:31 <jgriffith> scottda: that sounds good 16:18:33 <thingee> tbarron: +1 16:18:58 <tbarron> just tie him to his saddle after he visits microbreweries ... 16:19:00 <thingee> scottda: great anything else? 16:19:03 <scottda> Cool. Please comment wildly in the etherpad on the big stuff, i.e goals and design 16:19:09 <scottda> that's it. Thanks all. 16:19:21 <thingee> #topic Volume Migration Improvement has made some progress 16:19:24 <thingee> vincent_hou: hi 16:19:26 <asselin> o/ 16:19:29 <vincent_hou> Hi 16:19:44 <vincent_hou> I have a quick update with all cinder folks. 16:19:44 <thingee> #info volume status management is waiting for comments 16:19:47 <thingee> #link https://review.openstack.org/#/c/186312/ 16:19:58 <vincent_hou> I made some progress in the volume migration for L. 16:20:02 <thingee> #info migration status management waiting for comments 16:20:04 <thingee> #link https://review.openstack.org/#/c/189614/ 16:20:25 <vincent_hou> Have got 2 patches submitted for cinder . one for client. one for tempest. 16:20:29 <thingee> #info tempest tests 16:20:29 <thingee> #link https://review.openstack.org/#/c/195443/ 16:20:38 <thingee> #info cinder client changes 16:20:38 <thingee> #link https://review.openstack.org/#/c/189547/ 16:20:39 <vincent_hou> You are welcome to download them and try them. 16:21:02 <vincent_hou> I love you folks' comments BTW. 16:21:39 <vincent_hou> The volume has a better status and migration status management, with the migration progress. 16:21:45 <jungleboyj> :-) 16:22:00 <erlon> vincent_hou: great! 16:22:17 <vincent_hou> Tempest can check if the retype/migration work among any drivers, if you configure two backends. Any back-end! 16:23:05 <vincent_hou> Cinder client can help you to check the migration status and progress if available. 16:23:26 <thingee> vincent_hou: great thanks 16:23:35 <mtanino> vincent_hou: great for your much work. 16:23:36 <thingee> anyone have objections they'd like to raise with some of this progress? 16:23:59 <smcginnis> vincent_hou: Thanks for your work on this! 16:24:08 <jungleboyj> vincent_hou: Thank you very much! 16:24:39 <thingee> ok with that then... 16:24:51 <thingee> #topic Add Fujitsu ETERNUS DX Volume Driver (again) 16:24:59 <thingee> Fdaisuke: hi 16:25:07 <thingee> #link https://review.openstack.org/#/c/201500/ 16:25:13 <Fdaisuke> I'm here 16:25:27 <Fdaisuke> thingee: hi 16:25:34 <thingee> Fdaisuke: are you aware of previous openstack mailing list posts about this subject? 16:26:07 <smcginnis> My question is, if it was dropped in Kilo, why wasn't this addressed prior to L1 instead of waiting until now? 16:26:26 <jgriffith> smcginnis: maybe they didn't know it was dropped :) 16:26:26 <e0ne> smcginnis: +1 on your question 16:26:27 <Fdaisuke> yes, I understand the deadline was already past. 16:26:45 <smcginnis> jgriffith: :) We should probably talk about that one too. 16:27:07 <jgriffith> smcginnis: indeed :) 16:27:16 <thingee> Fdaisuke: we also have other vendors with drivers removed in K, that didn't make the cut in L-1. It wouldn't really be fair for us to make an exception here. 16:27:25 <thingee> Fdaisuke: but this was communicated well ahead of time http://lists.openstack.org/pipermail/openstack-dev/2015-May/064072.html 16:27:42 <Fdaisuke> so, we planned re-merging of our driver for L, assuming the deadline would be l-3 as it was k-3. 16:28:23 <hemna> thingee, when was the driver removed? K ? 16:28:34 <Fdaisuke> We couldn't fit our schedule to it as there're a lots of stuff to do such as purchasing new servers, assign people, etc. 16:28:43 <smcginnis> hemna: Yep 16:29:03 <hemna> smcginnis, then I'd say, they missed the L deadline. M isn't that far away. 16:29:05 <thingee> Fdaisuke: so I'm going to tell you what I tell everyone in this situation. You have priorities, and so does the community. Vendors don't define the timelines here, the community does. I absolutely welcome you to join the community and participate in these discussions so they're not a surprise. 16:29:15 <smcginnis> hemna: +1 16:29:43 <jungleboyj> thingee: ++ 16:29:53 <e0ne> thingee: +1 16:29:58 <jgriffith> hemna: thingee +1 to both of you 16:31:56 <thingee> Fdaisuke: the cinder community working towards this effort has been known for almost two years now. Things really being defined in a long email i sent in January 2015. This being a timing issue at this point is not really a good excuse. 16:32:42 <thingee> Fdaisuke: http://lists.openstack.org/pipermail/openstack-dev/2015-January/054614.html 16:33:08 <vincent_hou> Fdaisuke: Try to stay more with the cinder community. IRC, Mail list, etc. 16:33:52 <erlon> thingee: do you have those links on your copy area? :P impossible to find that so fast 16:34:13 <thingee> erlon: I do certain finds from the archive page of the mailing list. 16:34:26 <winston-d_> erlon: trust search engine and your memory ;) 16:34:48 <thingee> Fdaisuke: I'm going to move on from this topic, but please start making preparations for M. 16:34:54 <jungleboyj> thingee: Would be good to spend a few minutes talking at the meetup about how to find stuff in the mailing list. :-) 16:35:10 <erlon> winston-d_: I would have to rely on the search engine, my memory is a disaster 16:35:12 <thingee> #topic Move cinderclient-dsvm-functional job to (non)voting queue 16:35:15 <thingee> e0ne: hi 16:35:22 <e0ne> thingee: hi 16:35:28 <thingee> so mtreinish explained I'm crazy here for proposing this for a meeting topic 16:35:31 <thingee> I agree 16:35:41 <thingee> so any strong objections with this? 16:35:47 <jgriffith> LOL 16:35:56 <thingee> heh 16:35:56 <e0ne> quick topic:) 16:36:02 <jgriffith> thingee: the part about the job, or you being crazy? 16:36:03 <jgriffith> :) 16:36:04 <mtreinish> thingee: heh, I think we can all agree to blame jgriffith and move on :) 16:36:11 <thingee> #info job stats 16:36:13 <thingee> #link 16:36:13 <thingee> http://graphite.openstack.org/render/?width=877&height=548&_salt=1436533755.887&from=00%3A00_20150524&until=23%3A59_20150715&target=stats.zuul.pipeline.experimental.job.check-cinderclient-dsvm-functional.FAILURE&target=stats.zuul.pipeline.experimental.job.check-cinderclient-dsvm-functional.SUCCESS&target=stats.zuul.pipeline.experimental.openstack.python-cind 16:36:13 <thingee> erclient.total_changes&title=check-cinderclient-dsvm-functional 16:36:14 <DuncanT> +A on both motions 16:36:17 <thingee> dang it 16:36:30 <thingee> #info review request 16:36:32 <thingee> #link https://review.openstack.org/#/c/200522/ 16:36:42 <smcginnis> +1 16:36:50 <jgriffith> thingee: rebase an I'm good with it for sure!! 16:36:54 <jgriffith> err... e0ne 16:36:57 <thingee> excellent! 16:37:06 <thingee> #agreed rebase and merge it! 16:37:12 <e0ne> i've updated change request 16:37:17 * jgriffith was just waiting for rebase 16:37:34 <jungleboyj> DuncanT: ++ 16:37:34 <e0ne> jgriffith: done 16:37:42 <jgriffith> e0ne: me too :) 16:37:46 <thingee> e0ne: anything else? 16:37:55 <e0ne> thingee: no, thank you 16:37:58 <flip214> thingee: got that link from me? ;P 16:38:10 <e0ne> it was too quickly for me 16:38:12 <flip214> it's a bit long, though... 16:38:12 <thingee> flip214: link? 16:38:17 <thingee> oh 16:38:20 <flip214> graphite.openstack.org/render... 16:38:27 <thingee> flip214: https://wiki.openstack.org/wiki/CinderMeetings#Next_meeting 16:38:30 <thingee> flip214: last item 16:38:38 <thingee> #topic open discussion 16:38:42 <flip214> no problem 16:38:57 <Swanson> Blueprints for driver changes? 16:39:00 <jgriffith> thingee: I've got a couple things :) 16:39:11 <thingee> jgriffith: you have the floor 16:39:20 <jgriffith> thingee: thanks! 16:39:24 <jgriffith> so first.. replication V2 16:39:31 <erlon> me too, id like to discus about our HBSD2 CI status 16:39:44 <jgriffith> I put a POC up to try and give a clearer picture of what was in the spec 16:39:55 * jungleboyj is excited! 16:39:56 <jgriffith> https://goo.gl/eTNPrb 16:40:20 <thingee> #info replication v2 16:40:24 <thingee> #link https://review.openstack.org/#/c/155644/ 16:40:37 <jgriffith> So most of the comments are detail things... but a couple bigger ones 16:40:57 <jgriffith> well... not really big ones; the auto-failover thing 16:41:04 <jgriffith> everybody agree on just leaving that out for now? 16:41:19 <patrickeast> yea 16:41:22 <dannywilson> +1 16:41:28 <patrickeast> i think the initial version of this doesn’t really need it, right? 16:41:30 <jungleboyj> jgriffith: +1 16:41:32 <kmartin> +1 16:41:33 <jgriffith> and anybody have any glaring objections to the current direction? 16:41:37 <vincent_hou> y 16:41:49 <jgriffith> One question that came up fro me when doing this was (well couple questions): 16:41:53 <hemna> jgriffith, +1 16:41:53 <vilobhmm> +1 16:42:17 <jgriffith> 1. The existing V1 stuff... just replace it's API calls and reuse them, or keep them and make it a config option to fall back for a deprecation period? 16:42:17 <jungleboyj> jgriffith: Still need to actually review closely. 16:42:20 <thingee> #agreed auto-failover isn't needed in the initial version 16:42:28 <thingee> jgriffith: please reference this when needed ^ 16:42:32 <thingee> :) 16:42:39 <j_king> can I close https://bugs.launchpad.net/cinder/+bug/1298135 ? anyone affected by that? 16:42:39 <openstack> Launchpad bug 1298135 in Cinder "Cinder should handle token expiration for long ops" [Medium,Confirmed] - Assigned to j_king (james-agentultra) 16:42:39 <jgriffith> thingee: :) 16:43:02 <jgriffith> the other question I had.... 16:43:18 <Swanson> Haven't read that spec since it first came out. Will take another look. 16:43:39 <jgriffith> Config versus just using update_targets? 16:43:44 <DuncanT> j_king: We're affected by it, not looked at the implied trusts stuff yet 16:43:56 <jgriffith> Swanson: sorry.. forget the spec.... just look at this: https://review.openstack.org/#/c/200269/ 16:44:09 <jgriffith> IMO it's easier to get an idea from that than the spec 16:44:23 <jgriffith> on opinions on the two questions there? 16:44:34 <Swanson> jgriffith: thanks! 16:44:35 <e0ne> j_king: we met this issue few times 16:44:38 <thingee> jgriffith: sorry not familiar to comment 16:44:38 <jgriffith> Yayyy! /me has full liberty :) 16:44:54 <dannywilson> jgriffith: more info on config vs. update_targets? 16:45:02 <dannywilson> or is it too much detail to go into? 16:45:12 <jungleboyj> jgriffith: As far as leaving an option to go back to V1 and having an option. I think we want to just move forward with V2. I know Tao and Ronen are looking at the impact on the existing code right now but their first impression is that we can move forward pretty easily. 16:45:14 <jgriffith> dannywilson: so initially I thought "use config file to specify replication targets to a driver" 16:45:23 <jungleboyj> dannywilson: ++ 16:45:30 <j_king> DuncanT: I have some ideas/code I can put into a WIP review but it looks like it might not work without some help from keystone folks. 16:45:33 <jgriffith> dannywilson: but then I wrote a little code that introduced an "add_target" method to the api 16:45:43 <jgriffith> making the config file seem somewhat pointless 16:45:56 <jgriffith> j_king: DuncanT do you mind waiting till I finish this topic? 16:46:01 <j_king> sure. 16:46:06 <DuncanT> sure 16:46:06 <jgriffith> or have your side conversation elsewhere? 16:46:31 <jgriffith> dannywilson: so using the API is trivial IMO 16:46:35 <dannywilson> jgriffith: add_target would be called on a per volume basis? 16:46:47 <jgriffith> dannywilson: no, it would be called for a backend 16:47:04 <jgriffith> dannywilson: specify valid secondary configs via that call 16:47:20 <jgriffith> dannywilson: then it's up to you to set up which (or all) via type specification 16:47:31 <jgriffith> dannywilson: which can be auto-handled by driver, or explicit or whatever you want 16:47:47 <jgriffith> dannywilson: but my main goal here is to remove as much of the vendor unique stuff from the main code 16:47:48 <dannywilson> jgriffith: gotcha 16:47:56 <patrickeast> so is the thought here that there would also be api’s to change the targets, list them, etc? 16:48:02 <jgriffith> and leave most of the burden on the driver, rather than the proper cinder code 16:48:06 <Fdaisuke> thingee: Thanks for the comments. We might sounded like we're not aware of all the announcements in the past, but we're :) Since our driver is not completely new one, we thought we could find a chance to discuss it for Liberty inclusion. 16:48:14 <jgriffith> patrickeast: yes... admin API's 16:48:21 <patrickeast> gotcha 16:48:31 <jgriffith> patrickeast: and only two... "list" and "update" 16:49:10 <jgriffith> patrickeast: dannywilson please see: https://review.openstack.org/#/c/200269/1/cinder/volume/manager.py L#2737 16:49:24 <dannywilson> jgriffith: +1 for apis, makes it easier to change as well (though that probably doesn't happen often) 16:49:26 <jgriffith> patrickeast: I don't have the list method in there though 16:49:53 <jgriffith> dannywilson: Hope not, but regardless at least there' s no downtime or restart of services to make a change 16:50:12 <dannywilson> jgriffith: right 16:50:21 <jgriffith> I guess nobody has looked at this yet, so we can table the conversation for now 16:50:37 <jgriffith> DuncanT: j_king all yours 16:50:47 <jgriffith> Oh.. no, wait 16:50:49 <j_king> jgriffith: solved it out of band. :) 16:50:49 <jgriffith> one other thing :) 16:50:56 <jgriffith> j_king: cool! 16:51:04 <Swanson> Blueprints for driver changes? Like, say, mine? https://review.openstack.org/#/c/199631/ 16:51:25 * jgriffith puts on flame retardant suit! 16:51:28 <jgriffith> https://review.openstack.org/#/c/201812/ 16:51:30 <smcginnis> Swanson: Added it as a midcycle meeting topic to discuss bps and bugs. 16:51:30 <jungleboyj> jgriffith: Will be on it ASAP. 16:51:55 <jgriffith> I don't want to take up meeting time arguing and trying to justify, but happy to have a conversation in channel 16:52:14 <jgriffith> and explain the rationale in more detail as well as pros/cons etc 16:52:32 <jgriffith> which could lead to an updated commit message 16:53:02 <erlon> Can we talk about the HBSD2 CI status? 16:53:10 <erlon> real quick 16:53:19 <thingee> erlon: sure 16:53:52 <erlon> so, our CI was disabled, we had it turn off due a bug we where working 16:54:30 <erlon> after a power failure, we bring up the machine, and it stated to post failure status into gerrit 16:55:08 <erlon> as, the change has just being merged, we needed it enabled again to test against the new patch 16:55:42 <smcginnis> erlon: I do think you were disabled too quickly. There are certainly others that deserve it more. 16:56:14 <erlon> anteaya: Anita asked to talk to you guys to ask her to re-enable it 16:56:40 <erlon> smcginnis: yep, surely, I not even notice that 16:56:45 <thingee> erlon: I'll look at the stats 16:57:17 <erlon> thingee: you mean the stats of this CI? 16:57:52 <thingee> erlon: I'll look at the announce third party list 16:57:57 <thingee> anything else from anyone? 16:58:02 <erlon> thingee: ok 16:58:05 <smcginnis> 2 minute warning. 16:58:18 <thingee> #endmeeting