16:00:06 <smcginnis> #startmeeting Cinder
16:00:07 <openstack> Meeting started Wed Jun 14 16:00:06 2017 UTC and is due to finish in 60 minutes.  The chair is smcginnis. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:08 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:10 <openstack> The meeting name has been set to 'cinder'
16:00:12 <smcginnis> bswartz: Only just
16:00:13 <Swanson> hello
16:00:16 <_alastor_> o/
16:00:17 <smcginnis> ping: dulek duncant eharney geguileo winston-d e0ne jungleboyj jgriffith thingee smcginnis hemna xyang1 tbarron scottda erlon rhedlind jbernard _alastor_ bluex karthikp_ patrickeast dongwenjuan JaniceLee cFouts Thelo vivekd adrianofr mtanino karlamrhein diablo_rojo jay.xu jgregor lhx_ baumann rajinir wilson-l reduxio wanghao thrawn01 chris_morrell watanabe.isao,tommylikehu mdovgal ildikov wxy
16:00:23 <smcginnis> viks ketonne abishop sivn
16:00:31 <bswartz> .o/
16:00:32 <jgriffith> hola
16:00:34 <xyang1> Hi
16:00:36 <mdovgal> hi
16:00:38 <tommylikehu> hi
16:00:38 <jungleboyj> o/
16:00:43 <jungleboyj> @!
16:00:43 <pewp> jungleboyj ( ^_^)／
16:00:47 <wxy|> hi
16:00:53 <e0ne> hi
16:00:54 <jgriffith> pewp
16:00:54 <rfolco> o/
16:00:55 <scottda> hey
16:01:00 <pots> o/
16:01:13 <jgriffith> errr "poop"
16:01:27 <lhx__> o/
16:01:39 <jungleboyj> @!b jgriffith
16:01:39 <pewp> jungleboyj (╯°□°)╯︵ ┻━┻ ︵ ╯(°□° ╯) jgriffith
16:01:48 <smcginnis> #topic Announcements
16:01:58 <smcginnis> #info P-2 was cut last week
16:02:12 <smcginnis> We should now be focusing on finishing up new features and fixing bugs.
16:03:11 <smcginnis> PSA: tooz has been changed by devstack to now use the etcd driver instead of local file
16:03:20 <smcginnis> So if you see anything odd, keep that in mind.
16:03:40 <smcginnis> _alastor_ found a bug in the case where SERVICE_HOST=127.0.0.1
16:03:54 <smcginnis> I have a patch in to devstack to fix that, but it has not merged yet.
16:03:57 <rajinir> 0/
16:03:58 <xyang1> smcginnis: when was that merged
16:04:07 <smcginnis> xyang1: The change in devstack?
16:04:11 <xyang1> smcginnis: yes
16:04:23 <smcginnis> xyang1: I'm not entirely sure, but I think it was just a few days ago.
16:04:28 <xyang1> Ok
16:05:04 <tbarron> hi
16:05:07 <smcginnis> I wasn't aware of it until things broke, but I suppose it's good to get testing coverage on that if that is the way of the future.
16:05:35 <smcginnis> #info PTG registration is now open.
16:05:38 <smcginnis> #link https://www.eventbrite.com/e/project-teams-gathering-denver-2017-tickets-33219389087
16:05:54 <smcginnis> Real far for jgriffith, but hopefully he can make it. :)
16:06:11 <jgriffith> Yeah, trying to book travel now :)
16:06:21 <e0ne> smcginnis: it's real far for everybody outside US:(
16:06:26 <jgriffith> BMW express
16:06:29 <jungleboyj> Can we all stay at your house jgriffith ?
16:06:41 <jgriffith> sure
16:06:43 <smcginnis> e0ne: Yeah...
16:06:47 <jgriffith> you have to sleep in the barn though
16:06:52 <smcginnis> e0ne: Hopefully the next one is better.
16:06:58 <jgriffith> and it is an hour+ to the venue
16:06:59 <smcginnis> jgriffith: I like horses.
16:07:02 <tommylikehu> lol
16:07:07 <e0ne> :)
16:07:26 <jungleboyj> smcginnis:  :-)
16:07:27 <xyang1> jgriffith: make sure you don't fly out of Denver that week:)
16:07:30 <smcginnis> If I remember right, they wanted to do the first couple in the US to get things started, then start moving the PTG around more.
16:07:50 <Swanson> Denver's officail slogan: We're the counterpoint to all the beauty of the Rocky Mountains.
16:07:56 <Swanson> Official
16:08:04 <smcginnis> Hey now. :)
16:08:16 <jgriffith> Swanson ha!
16:08:21 <jgriffith> that's not too far off
16:08:26 <smcginnis> #topic IBM KVM on Power CI
16:08:32 <jgriffith> but there are worse places... much worse
16:08:35 <jgriffith> in co
16:08:39 <smcginnis> True
16:08:47 <smcginnis> rfolco: Hi.
16:08:56 <rfolco> hi
16:08:57 <smcginnis> rfolco: Usually don't have to ask to run CI, but I appreciate it.
16:08:57 <rfolco> :)
16:09:06 <lhx__> jgriffith, where are you from, guy?
16:09:07 <rfolco> Yeah... In spite of not being a hypervisor sensitive project, we consider cinder as �platform� sensitive.
16:09:15 <smcginnis> rfolco: Looks like this is a little special case. ;)
16:09:33 <rfolco> We want to avoid breakages like https://review.openstack.org/#/c/457233/, fixed later on https://review.openstack.org/#/c/467726/.
16:09:34 <rfolco> There might be other entry points for breaking Power platform, so reporting on cinder would help us take action before the change merges.
16:09:52 <smcginnis> rfolco: So I would recommend running against os-brick patches actually.
16:10:03 <smcginnis> rfolco: To catch those before we release an update to that library.
16:10:35 <rfolco> correct me if I am wrong.... os-brick is just one entry point for breaking the code here
16:10:46 <smcginnis> rfolco: True!
16:11:16 <smcginnis> rfolco: Both os-brick and cinder would be best, but if you need to limit what you can do, I think you will get more bang for your buck testing against os-brick.
16:11:36 <smcginnis> But definitely - we have many places where we can break things. :)
16:11:48 <rfolco> ok, honestly for cinder we have found ZERO bugs so far in 4 years
16:11:59 <rfolco> os-brick was the first case
16:12:16 <smcginnis> Not too bad a track record, I guess.
16:12:35 <jungleboyj> Cinder is quite perfect in every way.
16:12:47 <smcginnis> rfolco: So I guess to officially answer your agenda question - permission granted. :)
16:13:10 <rfolco> does this include os-brick and cinder right ?
16:13:34 <smcginnis> rfolco: Yeah, one or the other or both is fine with me.
16:14:00 <rfolco> awesome, we commit to monitor and not to produce much noise in case of failures, we just disable it.
16:14:04 <rfolco> thanks !!
16:14:16 <smcginnis> rfolco: Perfect, thanks!
16:14:35 <smcginnis> #topic Open discussion
16:14:49 <smcginnis> jgriffith: Anything to bring up with attach_v2?
16:14:52 <jungleboyj> rfolco: Thanks for wanting to add CI.  :-)
16:15:05 <jgriffith> smcginnis yes please
16:15:10 <eharney> Ceph CI should be functional again now
16:15:19 <smcginnis> eharney: +1
16:15:31 <lhx__> what that means?
16:15:59 <jungleboyj> eharney:  Good news!
16:16:21 <smcginnis> lhx__: Which part?
16:16:43 <lhx__> smcginnis, Ceph CI
16:17:22 <smcginnis> lhx__: There were problems with tempest tests causing the ceph CI to always fail.
16:17:35 <smcginnis> lhx__: They've been fixed now so patches are not blocked by it.
16:17:46 <smcginnis> By "fixed" I guess we mean "worked around"/
16:18:15 <lhx__> smcginnis, okay for that :)
16:19:32 <jgriffith> ready?
16:19:35 <smcginnis> jgriffith: Yes please.
16:19:45 <jgriffith> #topic attach v2 status
16:19:54 <jgriffith> bah who cares
16:19:56 <jgriffith> So...
16:20:04 <jgriffith> we're getting pretty close on the V2 attach stuff
16:20:14 <smcginnis> #topic Attach v2 changes
16:20:17 <jgriffith> I just pushed a patch to deal with the live-migrate stuff
16:20:20 <jgriffith> smcginnis show off!
16:20:28 <smcginnis> #chair jgriffith
16:20:29 * jgriffith has forgotten all of that foo
16:20:30 <openstack> Current chairs: jgriffith smcginnis
16:20:33 <smcginnis> :P
16:20:52 <jgriffith> There's one sticky thing we still have to figure out.. and that's the whole "shared connection" mess
16:20:55 <jgriffith> https://review.openstack.org/#/c/472796/
16:21:04 <ildikov> https://bugs.launchpad.net/cinder/+bug/1697008
16:21:05 <openstack> Launchpad bug 1697008 in Cinder "attach_delete removes export when multiple connections exist" [Undecided,In progress] - Assigned to Steve Noyes (steve-noyes)
16:21:22 <smcginnis> ildikov: Midnight?
16:21:35 <ildikov> smcginnis: a bit after :)
16:21:36 <jgriffith> So the solution initially was that a driver would get called on term connection and it would be up to that driver to respond if it was safe for brick to kill/delete that connection or not
16:22:11 <smcginnis> jgriffith: We will have multiple attachment records in the DB even if they all share the same connection, right?
16:22:15 <jgriffith> The problem is that I don't have access to a backend that doe this sort of thing or cares about it
16:22:21 <jgriffith> smcginnis correct
16:22:52 <jgriffith> I need some help from somebody with one of those backends to help verify a good solution here that will also work with multi-attach
16:22:53 <smcginnis> jgriffith: Can't we just pass in that count to the driver? Or use that in the manager to recognize that there are multiple?
16:22:59 <jungleboyj> jgriffith: Who's backend does that?
16:23:02 <ildikov> don't we have the problem or some version of it even in case of using LVM?
16:23:05 <jgriffith> I know 3par and pure fall into those categories
16:23:07 * smcginnis looks at Swanson
16:23:15 <jgriffith> patrickeast ^^
16:23:16 <jgriffith> ahh... Swanson !!
16:23:20 <jgriffith> You are now my target!
16:23:36 <smcginnis> mwaaahaahaaa
16:24:08 <jgriffith> so I need some help figuring those details out
16:24:27 <jgriffith> if folks ignore me and their driver breaks... well; sorry charlie :)
16:24:28 <patrickeast> Ah yea, I'll take a look
16:24:42 <ildikov> jgriffith: :)
16:24:50 <ildikov> patrickeast: tnx!
16:24:52 <jgriffith> If you could review the migration patch too that would be great (you==everyone)
16:25:03 <jgriffith> It's ugly, but I can only put so much lipstick on that pig
16:25:34 <Swanson> Which patch?
16:25:42 <jgriffith> or the other anology, "don't try to make a pig sing, it sounds awful and the pig just gets annoyed"
16:25:53 <Swanson> the hell?
16:25:57 <jgriffith> Swanson https://review.openstack.org/#/c/472786/
16:26:13 * jungleboyj needed the laugh.  Thanks jgriffith
16:26:26 <Swanson> thx
16:26:38 <jgriffith> If you haven't experienced the migration code "enjoy"
16:26:44 <smcginnis> hah
16:26:50 <jgriffith> I tried to add a bunch of comments so hopefully it will help
16:27:10 <Swanson> Oh, I've peeked at it.
16:27:15 <Swanson> Then went blind.
16:27:35 <jgriffith> That's what I've got for now.  Just going to need some active involvement from folks here in the final few days of this saga
16:27:49 <smcginnis> jgriffith: Thanks
16:28:02 <smcginnis> Looks like some good progress on the Nova side too.
16:28:09 <ildikov> we would love to get the Nova support in place for Pike
16:28:10 <jungleboyj> Yay!
16:28:18 <jungleboyj> ildikov:  +2
16:28:30 <ildikov> so pretty please, sugar or lipstick or whatever on top! :)
16:28:36 <smcginnis> :D
16:29:15 <smcginnis> Cool, anything else today? Or can we end early?
16:29:22 <smcginnis> (that usually ends up meaning we go to the end)
16:29:37 <jungleboyj> Bite your tongue!
16:29:40 <Swanson> So, repl v3....
16:29:49 <smcginnis> \ban Swanson
16:29:56 <Swanson> I kid I kid
16:30:10 * jungleboyj sends Swanson to the corner
16:30:19 <smcginnis> Swanson: Go figure out migration. :)
16:30:28 <tommylikehu> lol
16:30:33 <lhx__> lol
16:30:34 <bswartz> I think it's /mode +b Swanson
16:30:47 <smcginnis> bswartz: Hah, thanks for the tip.
16:31:01 <smcginnis> Alrighty, let's wrap up early this week. Thanks everyone.
16:31:07 <tommylikehu> thanks
16:31:11 <jungleboyj> Thanks!
16:31:19 <smcginnis> #endmeeting