16:00:17 <smcginnis> #startmeeting Cinder
16:00:17 <openstack> Meeting started Wed Mar  2 16:00:17 2016 UTC and is due to finish in 60 minutes.  The chair is smcginnis. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:18 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:21 <openstack> The meeting name has been set to 'cinder'
16:00:23 <smcginnis> Courtesy ping: dulek duncant eharney geguileo winston-d e0ne jungleboyj jgriffith thingee smcginnis hemna xyang tbarron scottda erlon rhedlind jbernard vincent_hou kmartin patrickeast sheel dongwenjuan JaniceLee cFouts Thelo vivekd adrianofr mtanino yuriy_n17
16:00:28 <jungleboyj> o/
16:00:29 <geguileo> smcginnis: Thanks
16:00:29 <eharney> hi
16:00:29 <e0ne> hi
16:00:31 <patrickeast> Hi
16:00:31 <cFouts> hi
16:00:31 <dulek> Hi!
16:00:31 <tbarron> hi
16:00:31 <geguileo> Hi
16:00:32 <yhayashi> hello
16:00:33 <vincent_hou> hi
16:00:33 <mtanino> hi
16:00:33 <smcginnis> Hey all
16:00:34 <akerr> o/
16:00:35 <DuncanT> Hi
16:00:35 <adrianofr_> Hi
16:00:41 <sheel> hello all
16:00:41 <thingee> o/
16:00:41 <scottda> hey
16:00:42 <jgregor> Hello!
16:00:44 <diablo_rojo> Hello :)
16:00:45 <baumann> Hello!
16:00:53 <ntpttr> Hi
16:01:15 <mc_nair> hey
16:01:21 <xyang1> hi
16:01:26 <smcginnis> #topic Release status
16:01:33 <jgriffith> aloha
16:01:40 <Swanson> hi
16:01:42 <smcginnis> Yesterday was our published feature freeze for Cinder.
16:01:59 <smcginnis> I will be cutting M-3 as soon as a couple final patches finish working through gate.
16:02:15 <dulek> Whee! :)
16:02:19 <smcginnis> :)
16:02:44 <smcginnis> We can still accept bug fixes, but we need to start locking down and focusing on final testing and RC prep.
16:02:46 <e0ne> good news!
16:02:55 <smcginnis> So reviewers, please keep that in mind.
16:03:19 <sheel> 👍
16:03:21 <smcginnis> There are a lot of good patches out there right now, but we need to start deciding if it is worth getting in M or safer to wait.
16:03:39 * thingee starts marking reviews accordingly
16:03:45 <smcginnis> At the end of this week is also soft string freeze.
16:04:09 <smcginnis> So even for those bug fixes we want in - please start to pay attention to unnecessary string changes.
16:04:21 <smcginnis> Our translation folks already have an impossible job as it is.
16:04:34 <smcginnis> So if we can help limit that load at all, it will be good.
16:05:05 <smcginnis> OK, that's all I really had to say. Any questions or other things anyone wants to point out?
16:05:31 <dulek> I haven't read a confirmation on RPC version bumps.
16:05:44 <dulek> So these patches will be allowed just before RC-1?
16:05:49 <DuncanT> We said we'd keep accepting cheesecake driver patches through feature freeze, right?
16:05:59 <smcginnis> dulek: I think that's probably safest. You think so?
16:06:17 <smcginnis> DuncanT: Correct. Repl patches should be treated as bug fixes at this point.
16:06:23 <dulek> It won't be end of the world if we end up not merging them, but it would be cool to clean up before the release.
16:06:29 <smcginnis> But the sooner we can help work those through, the better.
16:06:37 <smcginnis> dulek: I agree.
16:06:44 <dulek> smcginnis: I'm okay. I would love some help with testing - if anyone is willing to help.
16:06:46 <jgriffith> dulek: +1
16:07:11 <smcginnis> #topic A lot of TODO tags in cinder code
16:07:13 <smcginnis> sheel: Hey
16:07:20 <sheel> smcginnis: hey
16:07:25 <sheel> hello all
16:07:36 <sheel> actually this is regarding TODO tags present in the cinder code
16:07:52 <sheel> I think its more than 100 places where we could see these TODO
16:07:56 * jgriffith hides in the corner
16:07:57 <smcginnis> I think we did a push at the beginning of Juno IIRC.
16:08:06 <sheel> some are really important to follow
16:08:10 <smcginnis> jgriffith: Hah, there's no hiding for you!
16:08:20 <sheel> but some can be removed
16:08:23 <jgriffith> stupid initials
16:08:23 <sheel> jgriffith: haha
16:08:27 <eharney> really important TODOs should have bugs written
16:08:34 <sheel> eharney: +!
16:08:37 * flip214 has 14 in the one driver, ouch
16:08:46 <geguileo> eharney: What if those TODOs are for future releases?
16:08:48 <sheel> total 110 +
16:09:04 <geguileo> Like things that are marked TODO because they are related to rolling upgrades changes
16:09:08 <sheel> So, we can create bug and add then to outreachy
16:09:14 <geguileo> That require multiple releases changes
16:09:19 <smcginnis> grep -R TODO cinder/* --include="*.py" | grep jdg | wc -l
16:09:21 <smcginnis> Only 10
16:09:36 <dulek> geguileo: I guess we can start creating bugs after the release.
16:09:54 <smcginnis> dulek: +1
16:09:59 <dulek> But most of rolling-upgrades related TODOs are like free commits.
16:10:01 <sheel> 10 are there in dell driver only
16:10:02 <flip214> is there some other tag that should be used for reminders?
16:10:10 <dulek> So I think these will be cleared pretty fast. ;)
16:10:11 * smcginnis now hides
16:10:17 <sheel> smcginnis: hehe
16:10:23 <jgriffith> LOL
16:10:23 <geguileo> dulek: I have a good bunch of TODOs in my job distribution patches
16:10:24 <smcginnis> I blame Swanson
16:10:30 <jungleboyj> Oh how the tide can turn.
16:10:43 <Swanson> Swanson: TODO means not going to happen, right?
16:10:45 <flip214> of course, pushing a commit that removes these comments is easy, but I, for my part, put them in for a reason... ;)
16:11:05 <smcginnis> I do think it's good to review these every once in a while.
16:11:09 <smcginnis> There are a few different types.
16:11:16 <smcginnis> Some that really do need to be addressed.
16:11:30 <smcginnis> Some that are just notes that really we should try to improve something at some point in the future.
16:11:44 <sheel> smcginnis: is it ok to assign respective members and decide if bug creation is required or removal is ok
16:11:48 <smcginnis> And some that are just notes that it should be done, but it's not likely worth the effort to make it happen.
16:11:57 <Swanson> If they've been in there for a couple releases and nothing has happened then it isn't a problem and they can be deleted.
16:11:57 <smcginnis> sheel: Yeah, I think that would be good.
16:12:14 <jgriffith> Swanson: haha.. there are some that I ported from Nova Volume in there still
16:12:17 <sheel> jgriffith: plz stare at me..
16:12:19 <sheel> jgriffith: haha
16:12:29 <smcginnis> Didn't we do an etherpad in Juno for all of the tags to track and make sure they were all at least inspected?
16:12:30 <sheel> *dont stare
16:12:41 <jgriffith> sheel: :)
16:12:48 <e0ne> smcginnis: we did
16:12:56 <dulek> jgriffith: I love the ones from "inital fork from nova". :D
16:13:00 <smcginnis> e0ne: I don't recall - did that work well?
16:13:08 <geguileo> dulek: lol
16:13:13 <jgriffith> dulek: I'm getting to them... :)
16:13:21 <smcginnis> jgriffith: Hah, right!
16:13:32 <e0ne> smcginnis: not sure, because we can't remove all TODO's
16:13:45 <jgriffith> sheel: so if you're interested and nobody (looks at smcginnis ) objects...
16:13:59 <jgriffith> sheel: maybe it would be good to compile a list/ehterpad and start there
16:14:07 <sheel> jgriffith: yes its ok
16:14:14 <jgriffith> I agree with eharney that some should probably be bugs, but I know a bunch are just editorial
16:14:16 <dulek> jgriffith: It's smart you're using different IRC and TODO handles. It took me a while to figure out these weren't added by some guy from Nova.
16:14:24 <smcginnis> jgriffith: I think that would be good.
16:14:38 <sheel> jgriffith: i will be behind you all to get informatino about these TODOs
16:14:46 <jgriffith> dulek: LOL.. that *used* to be my irc handle till some evil doer registered it before me
16:14:49 <sheel> jgriffith: for bug creatino and all
16:15:04 <sheel> hope its ok to go like this
16:15:07 <sheel> ?
16:15:08 <jgriffith> sheel: sounds reasonable to me
16:15:21 <jgriffith> sheel: we can even just mark some as "remove" and be done with it
16:15:23 <sheel> jgriffith: ok 👍
16:15:25 <smcginnis> sheel: Good plan.
16:15:40 <dulek> +1
16:15:43 <smcginnis> sheel: Anything else to discuss here?
16:15:47 <sheel> smcginnis:  :)
16:16:05 <sheel> smcginnis: :)
16:16:26 <smcginnis> We can discuss more in channel as we work through them.
16:16:32 <smcginnis> sheel: Thanks for bringing it up.
16:16:38 <smcginnis> #topic API working group is changing microversion headers
16:16:43 <sheel> smcginnis: no its done i think
16:16:44 <smcginnis> scottda: Hi
16:17:07 <scottda> Hi
16:17:15 <scottda> So, the api working group is about to merge a patch that changes guidance on microversion headers:
16:17:23 <smcginnis> #link https://review.openstack.org/#/c/243414/ API guidelines
16:17:34 <scottda> https://review.openstack.org/#/c/243414/
16:17:44 <sheel> multiple headers
16:17:46 <scottda> I've bug and patch for this:
16:17:58 <scottda> #link https://bugs.launchpad.net/cinder/+bug/1551941
16:17:59 <openstack> Launchpad bug 1551941 in Cinder "api microversion header should conform to new guidlines" [Undecided,New] - Assigned to Scott DAngelo (scott-dangelo)
16:18:06 <scottda> #link https://review.openstack.org/286901
16:18:18 <scottda> Patch should be un-WIP'ed soon. I just want to add a unit test or two.
16:18:28 <scottda> This is a string change, so should be in before string freeze. And certainly before release.
16:18:47 <scottda> So, "Heads Up Reviewers"
16:19:09 <smcginnis> scottda: Thanks. We definitely should follow the api-wg direction.
16:19:09 <sheel> scottda: +1
16:19:23 <jgriffith> scottda: smcginnis just to clarify....
16:19:39 <jgriffith> scottda: smcginnis where does this fall in with M, freeze dates etc etc
16:20:01 <smcginnis> I think it's a bug fix, so we have a little time. Not much though.
16:20:13 <scottda> jgriffith: It should be fixed in M, before we ship with old style header
16:20:25 <scottda> And is a string change, so should be ahead of string freeze.
16:20:37 <jgriffith> Ok, cool
16:20:47 <scottda> That's it. Just something that popped up yesterday. I'm done.
16:20:53 <smcginnis> scottda: Thanks
16:21:33 <smcginnis> #topic Testing Cinder with Latest Oslo code
16:21:43 <smcginnis> jungleboyj: You're up cap'n.
16:22:06 <jungleboyj> smcginnis: Thanks.
16:22:31 <jungleboyj> So, Oslo now has periodic tests they can run with the latest Oslo libraries that makes sure that their changes don't break anything.
16:22:49 <jungleboyj> The reached out to the Liaisons to ask if the projects wanted to add this testing.
16:23:02 <jungleboyj> I don't see any reason not to do it but wanted to make people aware here.
16:23:19 <jungleboyj> Better chance to avoid Oslo merging anything that breaks us in the future.
16:23:31 <smcginnis> jungleboyj: Would be good to have a better chance of finding breaking changes sooner.
16:24:03 <jgriffith> jungleboyj: make it so
16:24:14 <patrickeast> Is there any downside? Seems like we definitely would want to do it
16:24:20 <e0ne> jungleboyj: it's a great idea to have such tests
16:24:34 <jungleboyj> smcginnis: jgriffith Cool.  That was what I figured people would say.
16:24:40 <jungleboyj> e0ne: Not aware of any.
16:24:49 <jungleboyj> I will follow up and make it happen.
16:25:18 <smcginnis> jungleboyj: Thanks for the heads up. I do think it will be a good thing.
16:25:33 <smcginnis> jungleboyj: Anything else on that or oslo in general?
16:25:35 <jungleboyj> smcginnis: Agreed.  I will let you all know when it is happening.
16:25:53 <e0ne> jungleboyj: do you know why https://review.openstack.org/#/c/284150/ is abandoned?
16:25:56 <e0ne> dims: ^^
16:25:59 <jungleboyj> smcginnis: No other Oslo news right now.  They are locked down for Mitaka at this point.
16:26:34 <jungleboyj> e0ne: Not sure.  Would have to follow up with dims_
16:26:37 <smcginnis> jungleboyj: Good, hopefully we start to have a little stability now that library freeze is passed.
16:26:47 * jungleboyj crosses fingers.
16:26:47 <smcginnis> e0ne: I had the impression that was just a test patch.
16:27:03 <dulek> e0ne: Wasn't it a single DO-NOT-MERGE patch to try out the libs?
16:27:05 <smcginnis> #topic Open discussion
16:27:18 <jgriffith> smcginnis: I've got one :)
16:27:25 <e0ne> smcginnis, dulek: thanks, looks you're right
16:27:32 <smcginnis> Anyone other than jgriffith?
16:27:33 <smcginnis> :P
16:27:37 <sheel> smcginnis: I have some questions...
16:27:37 <e0ne> :)
16:27:39 <vivekd> smcginnis: i've got one
16:27:43 <vivekd> :-)
16:27:46 <smcginnis> jgriffith: What's up? :)
16:27:46 <jgriffith> http://lists.openstack.org/pipermail/openstack-dev/2016-March/088007.html   http://lists.openstack.org/pipermail/openstack-dev/2016-March/088007.html and http://docs-draft.openstack.org/81/286981/1/check/gate-cinder-docs/e2a244e//doc/build/html/devref/attach_detach_conventions.html
16:28:09 <smcginnis> vivekd: You're next.
16:28:17 <vivekd> ok
16:28:30 <jgriffith> I know it's not an M thing, but I'd like it if folks could take a look at those things.  In particular the direction I would like to go in this WIP:  https://review.openstack.org/#/c/284867/
16:28:40 <jgriffith> that's all
16:28:53 <smcginnis> Would be very nice if we can get that sorted once and for all in N.
16:28:53 <vivekd> i'll start after jgriffith is done
16:29:04 <jgriffith> vivekd: I'm done :)
16:29:08 <sheel> :)
16:29:15 <scottda> jgriffith: Can I add you to the Interested Persons list https://etherpad.openstack.org/p/cinder-nova-api-changes
16:29:24 <vivekd> i've a proposal on cinder service topics configurability.
16:29:25 <e0ne> jgriffith: thanks for the patch! I'll take a look on it  and test later this week
16:29:28 <jgriffith> scottda: sure
16:29:49 <vivekd> i see that we've volume_topic, backup_topic as config options to let users configure the topics that the respective service nodes listen on.
16:29:58 <jgriffith> e0ne: cool... note that the idea is to just stop the connector passing stuff and just eventually return an attachment ID
16:30:02 <smcginnis> jgriffith: You just filed a bug on the topic thing, right?
16:30:11 <jgriffith> e0ne: sheel the "todo's" in the code should explain it :)
16:30:14 <vivekd> but from dulek comment @ https://review.openstack.org/#/c/286976/
16:30:21 <vivekd> i see that the conf values of topics are never used in the code.
16:30:31 <e0ne> jgriffith: so, I can't -1 for TODO's? :(
16:30:41 <vivekd> and that i think is enough evidence that no one really uses that feature.
16:30:44 <smcginnis> vivekd: Oh yeah, that's the one I was thinking of.
16:30:52 <vivekd> my opinion is that these features are not really adding any value.
16:30:58 <jgriffith> vivekd: :)
16:31:00 <vivekd> so can we remove that feature - "configurability of topics"?
16:31:04 <Swanson> jgriffith,  I'm changing the todo's to note:
16:31:07 <dulek> vivekd: I'm also confused on how to approach everything correctly.
16:31:22 <lixiaoy11> https://bugs.launchpad.net/cinder/+bug/1301888 this one is about topic
16:31:23 <openstack> Launchpad bug 1301888 in Cinder "cinder services doesn't honor _topic setting in config" [Medium,Confirmed] - Assigned to Lisa Li (lisali)
16:31:30 <smcginnis> Anyone with background on topic usage and why we might want it?
16:31:31 <vivekd> that will reduce the confusion
16:31:33 <sheel> Swanson: let me compile, is it ok to do it in one go
16:31:41 <jgriffith> is lixiaoy11 here?
16:31:52 <smcginnis> lixiaoy11: Thanks!
16:31:54 <lixiaoy11> jgriffith: yes
16:31:57 <jgriffith> vivekd: smcginnis I was talking with lixiaoy11 about this last night
16:32:02 <jgriffith> Oh... yeah.. there you are :)
16:32:03 <jgriffith> sorry
16:32:03 <vivekd> i tried to look through git logs but cuoldnt figure out much
16:32:14 <dulek> So my perception when commenting on vivekd's patch was we should fix it.
16:32:17 <geguileo> vivekd: Are you sure it is not being used?
16:32:24 <geguileo> vivekd: I kind of think it is...
16:32:27 <dulek> And let services to connect to configured topics.
16:32:27 <vivekd> yes geguileo
16:32:35 <lixiaoy11> vivekd: there is problem in the codes that we don't hornor topic
16:32:35 <jgriffith> geguileo: so I tried setting it last night and it blows everything up
16:32:37 <dulek> geguileo: World breaks when you change volume_topic. ;)
16:32:40 <Poornima> folks anything I can do to move ahead https://review.openstack.org/#/c/235752/
16:32:45 <jgriffith> geguileo: if you set it you get no rpc
16:33:19 * jgriffith thinks he had a TODO to implement that after moving it from nova-volume :)
16:33:27 <vivekd> :-)
16:33:30 <smcginnis> :)
16:33:33 <geguileo> Shouldn't we fix that?
16:33:36 <dulek> Now if we drop these config options instead of fixing that - is it making things easier for anyone?
16:33:42 <smcginnis> geguileo: Either fix or remove.
16:33:46 <DuncanT> Poornima: I'm looking at a bug in that very code patch now, as it happens, so I'd like to see them run somewhere
16:33:53 <geguileo> I think fixing is the right direction
16:33:58 <smcginnis> I guess at this point, I'm looking for justification why not to just remove.
16:34:02 <jgriffith> dulek: I'd say yes... because it's one lest edge case that we ignore
16:34:13 <smcginnis> It's obviously not being used if no one noticed until now that it's broken.
16:34:18 <DuncanT> If they've never worked, remove them
16:34:20 <mtanino> geguileo: hi, If you have a chance, could you visit the LIO backport patch? https://review.openstack.org/#/c/283085/
16:34:25 <vivekd> +1 DuncanT
16:34:34 <DuncanT> We've got too many already
16:34:48 <smcginnis> DuncanT: That's kinda what I'm thinking.
16:34:57 <dulek> Okay, if anyone will propose a patch removing them I'll comment if I would be able to justify a use case.
16:34:59 <jgriffith> I'd at the very least mark it as deprecated like right *NOW*
16:35:01 <lixiaoy11> DuncanT: at least in liberty, and now it doesn't work,  +1
16:35:03 <Poornima> DuncanT, sure since new to cinder would like to checkin api flow
16:35:10 <dulek> Right now the only one would be having two Cinder deployments on single RabbitMQ.
16:35:15 <vivekd> i can submit a patch to remove them smcginnis dulek
16:35:21 <jgriffith> then consider just removing it altogether, unless we find out that the problem is just that none of us are configuring things correctly :)
16:35:35 <dulek> Well, maybe it's a valid use case, but since it wasn't possible in the past…
16:35:37 <lixiaoy11> vivekd: I think you can take the above bug
16:36:09 <DuncanT> Poornima: I think we need to go through the cinder tempest tests and split out the ones that should gate everybody .v. the ones that should only gate cinder, and make a new gate job for the latter
16:36:21 <smcginnis> Remove, and if someone comes along needing that functionality we can add it then. Once we know someone actually needs/wants it.
16:36:27 <DuncanT> Poornima: It's going to be a Newton timeframe task though
16:36:38 <tbarron> do we know that this never worked? or just that it doesn't work now?
16:36:41 <dulek> What about deprecation period? It wasn't working, so we can skip it?
16:36:47 <vivekd> ok smcginnis
16:37:07 <tbarron> there are lots of customers out there running juno ...
16:37:15 <jgriffith> DuncanT: +1 to further elaborate, my thought was any test you can run with just [cinder, mysql, keystone, rabbit]  and possibly Glance
16:37:30 <DuncanT> Can anybody test it on juno/kilo? See how long it's been broken
16:37:32 <jgriffith> DuncanT: next iteration would be SWIFT I suppose for the backup stuff
16:37:43 <Poornima> DuncanT, alright do we have a bug created for it or ill create a new bug
16:37:53 <smcginnis> dulek: I'm thinking if it really was non-functional it doesn't need to be deprecated.
16:37:55 <DuncanT> jgriffith: For some tests, sure, for this you'd need swift and barbican too
16:38:04 <smcginnis> It kind of implicitly has been in that case.
16:38:11 <jgriffith> DuncanT: like I said... "next" iteration
16:38:12 <jgriffith> :)
16:38:26 <DuncanT> jgriffith: I think there's certainly room for a bunch of tests that run on a very cutdown system, for sure
16:38:27 <dulek> tbarron: Just looked through "initial commit out of nova". No, it wasn't working since Grizzly.
16:38:41 <tbarron> dulek: ty!
16:38:43 <dulek> tbarron: I mean yes, it wasn't working since Grizzly. :P
16:38:50 <DuncanT> dulek: Terminate with prejudice then IMO
16:38:55 <smcginnis> DuncanT: +1
16:39:02 <jgriffith> dulek: I wasn't joking when I said I had a TODO for that
16:39:07 <jgriffith> :(
16:39:10 <tbarron> dulek: yeah, that's how I understood you
16:40:04 <DuncanT> 'git blame' says it is all jgriffith's fault
16:40:10 <dulek> :) Thanks for raising this up vivekd, I should have proposed an agenda item instead of commenting on your patch with solution.
16:40:13 <vivekd> :-)
16:40:17 <jgriffith> damn you GiT!
16:40:22 <vivekd> np dulek
16:40:33 <smcginnis> Hah!
16:40:35 <vivekd> :-)
16:41:19 <smcginnis> Anything else? Or should we get 19 minutes back to review bug fixes and do testing? :)
16:41:25 <sheel> ok guys, if  done, could i go ahead with my queries ?
16:41:28 <mmandel> Can I please get 1 more core to review this change? https://review.openstack.org/#/c/238722/
16:41:28 <sheel> smcginnis: i am there
16:41:29 <lixiaoy11> I have one
16:41:41 <smcginnis> sheel: GO ahead
16:41:57 <sheel> smcginnis: jgriffith  and all other members : do we have any notes for below
16:42:04 <sheel> 1. migration procedure - for testing purpose.
16:42:04 <sheel> 2. how to setup multi backends (multiple cinder nodes), backup nodes, ceph nodes....
16:42:04 <sheel> 3. how can we fake drivers...
16:42:28 <sheel> As I used to have lot of suggestions, I want to propose them to add to trainings
16:42:30 <eharney> mmandel: i was reviewing that patch, i'll look at it again today
16:42:35 <smcginnis> Documentation?!
16:42:36 <smcginnis> :)
16:42:48 <mmandel> eharney: great, thanks very much!
16:42:49 <sheel> smcginnis: yup
16:42:50 <dulek> sheel: #1 is definitely in devref
16:43:03 <sheel> dulek: great, could you give a pointer
16:43:12 <dulek> sheel: http://docs.openstack.org/developer/cinder/devref/migration.html
16:43:29 <sheel> dulek: thanks
16:43:54 <sheel> Also, why we add this to devref and all...why dont we add these topics to openstack trainings ?
16:44:01 <thingee> a little neat program I found if you're a vim user... rst live rendering https://github.com/Rykka/InstantRst
16:44:07 <vincent_hou> For anything regarding migration, you can find me.
16:44:10 <sheel> means any specific agenda behind this
16:44:20 <sheel> thingee: yep, thats cool, i refered it today
16:44:22 <vincent_hou> I wrote http://docs.openstack.org/developer/cinder/devref/migration.html
16:44:25 <jgriffith> thingee: oh, nice!
16:44:30 <smcginnis> thingee: Nice! Missed the dev tools at the midcycle. :)
16:44:59 <thingee> keep in mind that the python server part has a bug, but I made a pullrequest https://github.com/Rykka/instant-rst.py/pull/11
16:45:04 <sheel> vincent_hou: cool, btw I meant db migration
16:45:14 <vincent_hou> sheel: oh.
16:45:30 <sheel> any pointers to below :
16:45:41 <sheel> 2. how to setup multi backends (multiple cinder nodes), backup nodes, ceph nodes....
16:45:41 <sheel> 3. how can we fake drivers...
16:45:56 <dulek> Weren't there more opens to talk through before presenting dev tools? :)
16:46:05 <smcginnis> sheel: What openstack trainings were you referring to?
16:46:21 <sheel> smcginnis: openstack/training-guides
16:46:29 <jungleboyj> sheel: Multi-backend configuration is documented if you google for multi-backend configuration.
16:47:10 <smcginnis> sheel: Would be good to have something in there. I'm not sure who does those though.
16:47:22 <sheel> jungleboyj: i tried...but some misses..could not successed..
16:47:24 <smcginnis> Or if folks even know about them much. They're fairly new IIRC.
16:48:03 <jungleboyj> sheel: Come to our presentation at the Summit.  ;-)
16:48:04 <sheel> I sew Mike Prez added for IRC tool @ https://review.openstack.org/#/c/286941/
16:48:23 <sheel> jungleboyj: I am planning...I voted for your presentation..
16:48:28 <sheel> jungleboyj: :)
16:48:33 <diablo_rojo> sheel :)
16:49:01 <sheel> :)
16:49:11 <jungleboyj> :-)
16:49:26 <sheel> jungleboyj: may be you can leak your presentation even before summit...
16:49:37 <sheel> jungleboyj: haha
16:50:14 <sheel> jungleboyj: I would take that reference for setup :)
16:50:19 <jungleboyj> Talk to diablo_rojo
16:50:32 <sheel> jungleboyj: diablo_rojo: sure
16:50:41 <jungleboyj> :-)
16:50:57 <Poornima> sheel, does those include developer training :)
16:51:40 <sheel> ok, one more point  left
16:51:41 <sheel> how can we fake drivers..
16:51:50 <sheel> Poornima: I will share link
16:52:28 <sheel> Poornima: https://github.com/openstack/training-guides
16:53:05 <sheel> smcginnis: ok, I am asking specifically... is there any way to fake dell driver for testing code on my hp laptop :)
16:53:07 <Poornima> sheel, cool thx
16:53:37 <Swanson> sheel, No, buy a storage center.
16:53:40 <sheel> smcginnis: I am interested in participating in driver code in cinder....so eager to no
16:54:10 <sheel> Swanson: I will think when i will be 60+. will have some money till then at least for buying :)
16:54:26 <sheel> Swanson:  :)
16:54:40 <sheel> Swanson: means there is no other way, right?
16:54:56 <smcginnis> Some vendors supply an array simulator.
16:55:11 <smcginnis> Theoretically you could run that on your laptop and use it as a Cinder backend.
16:55:18 <dulek> xyang1: Don't we have FakeDriver/GateDriver?
16:55:28 <xyang1> dulek: yes
16:56:03 <sheel> smcginnis: ok
16:56:10 <sheel> xyang1: dulek: any pointer
16:56:30 <xyang1> dulek: it is fake_driver under tests
16:56:53 <sheel> *dulek have almost every answer...great
16:57:33 <sheel> xyang1: ohk..
16:57:35 <smcginnis> 3 minutes. Anything else we need to hold the meeting channel for?
16:57:41 <sheel> I am done
16:58:10 <flip214> a free beer would be nie
16:58:12 <flip214> *nice
16:58:16 <smcginnis> Thanks everyone.
16:58:18 <geguileo> Thanks everyone
16:58:19 <smcginnis> flip214: ;)
16:58:25 <jungleboyj> Thanks.
16:58:30 <smcginnis> #endmeeting