17:01:54 <davidlenwell> #startmeeting refstack
17:01:54 <openstack> Meeting started Thu May  8 17:01:54 2014 UTC and is due to finish in 60 minutes.  The chair is davidlenwell. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:01:55 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:01:57 <openstack> The meeting name has been set to 'refstack'
17:02:03 <zehicle_> o/
17:02:10 <davidlenwell> roll call
17:02:13 <catherineD> Here
17:03:00 <davidlenwell> agenda: simplified tester, TCUP, api->api authentication, communication channels, refstack meetups at summit. summit sessions
17:03:30 <zehicle_> cool
17:03:40 <davidlenwell> zehicle_: can you think of anything else ?
17:03:42 <rockyg> o/
17:03:56 <davidlenwell> rockyg: you missed this : agenda: simplified tester, TCUP, api->api authentication, communication channels, refstack meetups at summit. summit sessions
17:03:57 <zehicle_> capabilities page?
17:04:10 <rockyg> tnx
17:04:14 <davidlenwell> okay we'll just cover that one first.
17:04:50 <davidlenwell> #topic capabilities page
17:05:05 <davidlenwell> you have the floor zehicle_
17:05:10 <zehicle_> thanks
17:05:29 <zehicle_> the DefCore committee has put out a draft of the capabilities
17:05:44 <zehicle_> and I wanted to be able to put it out as a simple render.
17:05:48 <davidlenwell> I did get your email .. but it seems like troy and I were both unsure who you meant should put that up some place
17:05:49 <zehicle_> we have a JSON format that looks OK
17:06:00 <zehicle_> sorry about that  -I was thinking restack.org
17:06:05 <zehicle_> should have been specific
17:06:10 <davidlenwell> oh.. yeah .. no biggy ..
17:06:25 <zehicle_> if the format of the json is OK then I'll do a patch w/ the HTML page
17:06:32 <davidlenwell> why don't you commit to the repo with those in the static files folder and link to it on the home page .. I'll pretty up the html if you need me to
17:06:39 <zehicle_> and also update the capabilities counts so they are not all 0
17:06:48 <zehicle_> yes please > pretty++
17:06:58 <davidlenwell> okay go ahead and do that today ..
17:06:59 <rockyg> Yeah.  the first pass is good enough to approve and get in.
17:07:12 <zehicle_> thanks
17:07:15 <davidlenwell> zehicle_: please make sure you don't have dependanies on your commit
17:07:27 <zehicle_> can do once I get out from behind the firewall
17:07:27 <davidlenwell> which reminds me of another agenda item we'll jump to next
17:07:28 <rockyg> ++
17:07:36 <zehicle_> will do, may ping you to make sure
17:07:37 <zehicle_> offline
17:07:51 <davidlenwell> #topic code reviews
17:08:24 <davidlenwell> specifically
17:08:25 <davidlenwell> https://review.openstack.org/#/c/91640/
17:08:29 <davidlenwell> needs to be rebased
17:08:43 <davidlenwell> I'd like to be able to merge it
17:08:56 * zehicle_ done
17:09:02 <davidlenwell> thank you
17:09:37 <davidlenwell> there is also a new draft of the v1 api specs I'd like to merge if everyone can review it today.
17:09:38 <davidlenwell> https://review.openstack.org/#/c/90044/
17:09:55 * zehicle_ can do
17:10:00 <rockyg> will do
17:10:02 <davidlenwell> Thank you
17:10:21 <davidlenwell> okay this leads nicely into the next topic
17:10:37 <davidlenwell> #topic simplified tester
17:10:58 <davidlenwell> you may have noticed if you were watching that we merged this last night ..
17:10:58 <davidlenwell> https://review.openstack.org/#/c/92552/
17:11:54 <davidlenwell> its actually not finished yet .. I have another patch that will hit it probably tomorrow afternoon that will complete the circle .. but there is enough of the tool finished to prep the env .. and start tests on it .. should be very useful for tcup
17:12:03 <rockyg> missed that.  saw stuff fly by related.
17:12:38 <davidlenwell> I have a diagram and some slides for it in the deep dive session
17:13:27 <rockyg> we gonna sync f2f at summit on tha beforehand?
17:13:27 <zehicle_> the sooner the better - we're down to the wire
17:13:40 <davidlenwell> rockyg: we'll get to that in a minute
17:13:47 <davidlenwell> its an agenda item
17:14:00 <rockyg> ping us all when code is ready for us to try.
17:14:00 <davidlenwell> zehicle_: you should be able to test with it now
17:14:15 <zehicle_> ok, I'll be on #refstack and look for help
17:14:21 <davidlenwell> the tests won't pass because the cloud prep piece is just a stub
17:14:22 <rockyg> Sounds like we can try now.
17:14:26 <zehicle_> will make my day
17:14:38 <davidlenwell> but it preps and launches tests..
17:14:43 <davidlenwell> then sends them home
17:14:52 <davidlenwell> and reports status to the api when it fails ..
17:15:05 <zehicle_> I just need to be able to launch - we can iterate in parallel after that
17:15:13 <davidlenwell> and interactively asks you for any values it needs in env that its missing
17:15:32 <davidlenwell> so its actually a useful tester more in line with our original specs
17:15:32 <praveen_dell> let me know if you want me to test any code
17:15:51 <davidlenwell> praveen_dell: you can test it now
17:16:05 <zehicle_> we may need to document the steps
17:16:10 <praveen_dell> sure,will do it
17:16:11 <rockyg> Everyone who has cloud access should try against their cloud.
17:16:19 <zehicle_> +1
17:16:19 <davidlenwell> some q/a on the parts that are done would be very useful for me to refine it into a finished state
17:16:35 <davidlenwell> note* it only works in havana right now
17:16:50 <rockyg> And file bugs once they get it working for missing parts, etc.
17:17:02 <davidlenwell> rockyg: ++
17:17:10 <rockyg> Can we run afainst others since the test don't run anyway?
17:17:22 <davidlenwell> it does run tests
17:17:28 <rockyg> kewl.
17:17:32 <davidlenwell> they'll just fail because the config isn't complete
17:17:53 <davidlenwell> I also put in a stub for scrubing results before sending them home
17:17:55 <catherineD> rockyg: File bug against Tempest Havana
17:18:01 <catherineD> ?
17:18:15 <davidlenwell> file bugs on the refstack lp page for the code I just put up
17:18:22 <davidlenwell> the tester is a central part of refstack
17:18:32 <davidlenwell> and filed bugs will help me make it stronger
17:18:32 <rockyg> what david said.
17:19:04 <davidlenwell> we're at a stage were we'll have to do some q/a passes
17:19:24 <davidlenwell> I can call it "finished" but it really isn't until we've worked out the kinks .. and that means lots of different people need to use it.
17:19:48 <davidlenwell> and give feed back
17:20:01 <catherineD> OK I will get off line with you on how to file bug against Refstack
17:20:09 <davidlenwell> I know of some kinks off hand .. thats why I'm saying its still a wip
17:20:23 <davidlenwell> catherineD: just go to launchpad.com/refstack and click file bug
17:20:29 <catherineD> ok
17:20:45 <rockyg> Question:  I've never seen a way to idenitfy the version in the various openstack projects.  Is ther?
17:21:34 <davidlenwell> im sure we can figure that out later
17:21:40 <davidlenwell> offline that one
17:21:53 <davidlenwell> we should move into our next topic .. TCUP
17:21:59 <zehicle_> +1
17:22:11 <davidlenwell> #topic TCUP
17:22:37 <davidlenwell> So with the tester in its current state .. we should be able to get the tcup code in better working order.
17:22:38 <zehicle_> basicaslly, nothing since last meeting.  I've been blocked on the launcher.
17:22:48 <zehicle_> I think we should be able to jump ahead now
17:23:07 <zehicle_> are all the pending patches looking good to go?  it will help w/ debug and instructions
17:23:07 <davidlenwell> excelent
17:23:29 <rockyg> for catherineD, last topic:  #link https://bugs.launchpad.net/refstack
17:23:31 <zehicle_> I can also make sure to hand off to praveen_dell
17:23:33 <davidlenwell> zehicle_: I'll look more closely offline and ping you if I need you to take any action
17:23:37 <zehicle_> so he can get an early jump on it
17:23:50 <zehicle_> thanks - I'm available this afternoon for TCUP
17:23:53 <catherineD> Thx Rocku
17:24:16 <davidlenwell> I will be hyper viligent with code reviews leading up into the summit
17:24:23 <zehicle_> praveen_dell, you have time tomorrow AM to try out the latest?
17:24:29 <praveen_dell> yeah
17:24:33 <praveen_dell> i will try tomorrow
17:24:39 <praveen_dell> latest tcup? right?
17:24:43 <zehicle_> davidlenwell, we should be in good shape.
17:24:49 <davidlenwell> good ..
17:25:00 <davidlenwell> we should move on
17:25:04 <davidlenwell> #topic api->api authentication
17:25:10 <zehicle_> yes - I may have to give you directions because the patch may not be accepted tomorrow
17:25:35 <praveen_dell> <zehicle_> ok please ping me tomorrow or send me an email a
17:25:48 <zehicle_> praveen_dell, will do!
17:25:53 <praveen_dell> <zehicle_> thanks
17:26:29 <davidlenwell> catherineD: rockyg and zehicle_ and I were having a discussion on irc the other day about how to handle auth for the api ->api sync feature
17:26:46 <davidlenwell> and I think that as  an openstack project we should eat our own dogfood and use keystone for this
17:27:12 <davidlenwell> it ends up openstack already built a tool for just this sort of thing and I think we should use it..
17:27:20 <rockyg> The big question for me is whether vendors consider it secure enough to accept it.
17:27:24 <davidlenwell> So asap I will put a running keystone server on refstack.org
17:27:34 <davidlenwell> rockyg: as I said the other day .. I don't care what they think
17:27:48 <davidlenwell> if they are openstack vendors they can trust keystone
17:27:52 <rockyg> We will if they won't send test results.
17:28:07 <davidlenwell> if they don't trust keystone they should help fix whatever they think is broken in it
17:28:18 <davidlenwell> rockyg: im not aware of any vendors who wouldn't trust it.
17:28:30 <davidlenwell> please inform me thusly if I am wrong
17:28:30 <zehicle_> it's good for a first pass
17:28:38 <zehicle_> if there are issues, we can address as they come up
17:28:41 <rockyg> Yes.  Good first pass.
17:28:49 <davidlenwell> dissagree
17:28:53 <zehicle_> ok
17:28:55 <davidlenwell> I think its the long term solution
17:29:02 <zehicle_> davidlenwell, perhaps it is
17:29:09 * zehicle_ wants it to be
17:29:24 <davidlenwell> I just don't understand why it would be up for debate
17:29:29 <rockyg> But, whether problems arise with it or not, we have a working solution to demo.
17:30:01 <zehicle_> davidlenwell, I don't have issues.  was mainly asking for us to hold any issues for later to see if they emerge
17:30:23 <davidlenwell> there is a team that works on it full time .. they understand the problem space and are constantly working to keep it secure.. that is way more than we can bring to the problem.. and I won't participate in re-inventing wheels in the openstack ecosystem.
17:31:02 <catherineD> Maybe this is a topic for the face-2-face.  Using Keystone practically we will have 2 Keystone one on refstack.or and the other one on local?  How do we sync? Why not using lainchpad? Why do we use launchpad to log in and keystone for others?
17:31:07 <davidlenwell> period.. so if someone thinks there is a good reason why Vendors woldn't trust there keys in keystone on a foundation run server.. please speak up ..
17:31:28 <davidlenwell> catherineD: not at all.. the farside will just have the client libs
17:31:53 <zehicle_> +1 catherineD on discussion f2f not here
17:32:18 <davidlenwell> catherineD: lets talk online .. its actually a topic that should be in a written up spec .. which we are supposed to have in progress for this topic already.
17:32:38 <catherineD> davidlenwell: +1.  Better face-2-face next week
17:32:44 <davidlenwell> that is why this process is in place .. to refine ideas.. lets use them
17:33:10 <davidlenwell> this leads me nicely into the next topic of discussion
17:33:13 <davidlenwell> #topic communication channels
17:33:42 <rockyg> David, whether or not ther is an issue, I always prepare for the worst.  Having an awareness that we might need to refactor something means we are thinking ahead and may have better solutions, or ways to strengthen current solution to answer concerns.
17:34:41 <davidlenwell> rockyg: I think its the solution.. its how openstack projects authenticate between client and sever .. its ludicris for us to discuss doing it any other way.
17:34:45 <davidlenwell> discussion over.
17:34:51 <davidlenwell> we'll work out the how later
17:35:08 <rockyg> coom channels
17:35:11 <rockyg> comm
17:35:35 <davidlenwell> Over a month ago I implimented a new policy called specs
17:36:07 <davidlenwell> in general we've started to head down the right path with them. working out the how before writing code and having to work in a lot of refactors.
17:36:37 <davidlenwell> I know its an od idea.. but this is how projects in the ecosystem are dealing with distributed teams .. and for nova and others it has been very effective.
17:37:03 <davidlenwell> I believe it can be for us.. But we all need to get on board with the work flow.
17:37:18 <davidlenwell> it is very tempting to just want to work things out in f2f meetings..
17:37:26 <davidlenwell> the problem is our team is likely about to grow..
17:37:59 <davidlenwell> and almost certainly the people we talk into jumping into the stream with us won't live in the bay area.
17:38:08 <rockyg> The problem is that until we have a good foundation, we need to determine directions brefore we can drill down to the deail level of specs.
17:38:13 <davidlenwell> So we cannot rely on those meetings in order for progress to happen
17:38:18 <davidlenwell> rockyg: .. yes you can
17:38:33 <davidlenwell> write down how you see it and submit it for review
17:38:41 <davidlenwell> if you are wrong you will be corrected
17:38:43 <rockyg> No.  You casnt'.  Until we know the requirements, specs don't mean much because req
17:38:49 <rockyg> s are the foundation.
17:39:06 <davidlenwell> the very first part of the spec should be a list of requirements
17:39:15 <davidlenwell> the problem is this process needs to work async
17:39:22 <davidlenwell> otherwise we can't be productive ..
17:39:37 <zehicle_> IMHO, there's a balance here and part of the reason for having f2f summits is to work out the top level issues and then use specs to make sure we have agreement
17:39:49 <rockyg> ++
17:39:58 <davidlenwell> zehicle_: rockyg.. the specs work without the agreement
17:40:03 <catherineD> ++
17:40:12 <zehicle_> davidlenwell, yes but not as quickly
17:40:26 <rockyg> How many different specs do we have for TCUP?
17:40:26 <davidlenwell> zehicle_: we can't always rely on face to face or phone calls
17:40:35 <davidlenwell> I hope a lot
17:40:40 <davidlenwell> thats the point of it
17:40:47 <rockyg> I mean alreday?
17:40:54 <davidlenwell> its called a "process"
17:41:03 <davidlenwell> one spec has been written .. and its been merged
17:41:16 <zehicle_> davidlenwell, I agree!  I think with the summit looming it's easy to say next week.
17:41:23 <davidlenwell> rob didn't need us to tell him what to write in the first version.. he wronte it
17:41:33 <davidlenwell> then we all jumped in and commented on it
17:41:37 <davidlenwell> and then he refined it
17:41:49 <rockyg> And it got trashed.  And written aggain.  And trashed.  And never checked in.
17:41:55 <davidlenwell> we all have full time jobs and we live all over the world
17:41:56 * zehicle_ <3 being the positive role mode *blush*
17:41:59 <davidlenwell> rockyg:  its merged now
17:42:05 <davidlenwell> it did get checked in
17:42:14 <rockyg> We are not working effectively because we don't have enough info to do so.
17:42:44 <davidlenwell> rather it will be because of how zehicle_ submitted it we have to merge something else first that wasn't rebased the right way .. now thats fixed and its in
17:43:13 <davidlenwell> if used propperly this allows the team to work async and distributed .. which is how openstack projects work
17:43:18 <davidlenwell> we are not all under the same roof
17:43:30 <davidlenwell> we all have different time zones and schedules and personal responsibilites
17:43:45 <davidlenwell> we can't always sysnc up on the phone or in person and we cannot rely on that to make progress
17:43:53 <davidlenwell> that is what this process was designed to solve
17:44:03 <rockyg> I underrstand thaat and have actually worked with teams that weay before.  I've managed off shore teams and had them work with teams here.  We were very productive.
17:44:28 <davidlenwell> where there 12 differnet employers involved in 8 time zones ?
17:44:37 <catherineD> We only have about 15 mins left.  I think we should discuss summit meet up and prep for the session ... who is presenting and at the center chair etc ...
17:44:39 <davidlenwell> thats how openstack projects go
17:44:45 <davidlenwell> I am
17:44:46 <rockyg> Becaue we were all on the same page.  To get there, we had a foundational understanding of what we were doing.
17:44:59 <davidlenwell> rockyg: lets talk about this offline ..
17:45:11 <davidlenwell> catherine is right it is time to move on
17:45:23 <davidlenwell> refstack meetups at summit
17:45:27 <davidlenwell> #topic refstack meetups at summit
17:45:44 <rockyg> I'm going to call the f2f working session at11:30am Monday in the infra pod.
17:45:56 <davidlenwell> I won't be there
17:46:04 <davidlenwell> flight doesn't land until 3pm on monday
17:46:08 <rockyg> K.
17:46:36 <rockyg> There's lots we can work on without you.  Not the pres. unless you give us direction for thaat.
17:46:53 <davidlenwell> rockyg: I'd preffer to be present for meetups regarding this project
17:46:55 <rockyg> We can get supporting stuff together if you give us pointers, though.
17:47:01 <davidlenwell> its coutner productive otherwise
17:47:11 <rockyg> No it's not.  we can work on specs.
17:47:12 <davidlenwell> I've got the session it self planned..
17:47:30 <rockyg> We can research keystone.
17:47:43 <davidlenwell> you could do that without a f2f
17:47:54 <catherineD> davidlenwell: That is good ... so you will be at the center chair and we will be on first row ...
17:48:02 <davidlenwell> yes
17:48:13 <davidlenwell> I have slides and diagrams and a whole thing planned .. you'll love it
17:48:17 <davidlenwell> then we'll do q/a
17:48:23 <davidlenwell> q and a
17:48:25 <rockyg> Do you need anything beyond your lapop and a projector for the pres?
17:48:30 <davidlenwell> no
17:48:32 <davidlenwell> I do not
17:48:34 <catherineD> and you have stuff (slide , lauchpad ..) ready ...
17:48:40 <davidlenwell> yes I do
17:48:44 <catherineD> that is good ....
17:48:44 <rockyg> kewl.
17:48:59 <davidlenwell> still putting images and stuff together for it and refining diagrams
17:49:12 <rockyg> Praveen, will you be at summit?
17:49:19 <davidlenwell> as far as meetups are concerned
17:49:36 <davidlenwell> I'd be open to a team breakfast .. or us eating lunch together on tuesday before the session
17:49:46 <zehicle_> +1
17:49:50 <davidlenwell> we also have the session it self to use fot planning and discussion .. that is what it is for
17:49:52 <rockyg> +1
17:50:12 <zehicle_> Breakfast w/ the board is Wed.  would be cool to have you (all) there too
17:50:13 <davidlenwell> also you should all note that at 5:30 pm there is another session that concerns us
17:50:26 <rockyg> We need to try to get meeting with QA so that we get better comm channels re; tempest
17:50:31 <catherineD> +1 say we meet for lunch how do we connect and know the table location ?
17:50:47 <davidlenwell> catherineD: we'll jump into irc and figure it out
17:51:00 <davidlenwell> since none of us know the venue now its silly to pick a meeting place here
17:51:21 <davidlenwell> the conf has wifi .. we should all be prepared to use irc while there and probably follow each other on twitter
17:51:26 <davidlenwell> @davidlenwell is me
17:51:30 <praveen_dell> ] <rockyg>you didn't approve me for summit
17:51:48 <praveen_dell> ] <rockyg> sorry my boss didnt approve me for summit :)
17:51:56 <davidlenwell> boo zehicle_
17:52:17 <zehicle_> was not me
17:52:32 <rockyg> The pods.  Infr has one.  We can try to sign up for it Tuesday lunch time on Monday.
17:52:34 <davidlenwell> well maybe he can come to paris
17:52:40 * zehicle_ activities != Dell activities
17:52:42 <davidlenwell> rockyg: we don't need a pod
17:52:54 <rockyg> It's a lovation.
17:53:02 <davidlenwell> I don't know how much time you ahve spent at a summit
17:53:03 <rockyg> location.  Just sayin
17:53:07 <davidlenwell> but finding meeting space is really easy
17:53:20 <davidlenwell> we designed most of tripple o sitting in a hall way in san diego
17:53:36 <rockyg> Which session 5:30 Tues?
17:53:39 <davidlenwell> there are hotel lobby's and couches and eating tales all over
17:53:55 <rockyg> Test natrices and feature parity?
17:54:04 <davidlenwell> http://junodesignsummit.sched.org/event/fd84ec7ddc3252270fb73e8e9e09cfba#.U2vEt61dUa0
17:54:06 <davidlenwell> yes
17:54:14 <davidlenwell> #link http://sched.co/1h3Xca9
17:54:28 <rockyg> Already in personal schedule ;-)
17:54:33 <davidlenwell> mine too
17:54:36 <davidlenwell> I'll see you there
17:55:10 <catherineD> So no meeting on Monday?
17:55:11 <rockyg> With lots of questions in hand.  That may spur the conversattion to better sync us to temprst
17:55:21 <davidlenwell> I won't be in town until the afternoon ..
17:55:38 <davidlenwell> rockyg: we'll just see when we are there
17:55:45 <rockyg> We can still do Monday.  We can work on getting specs cleaned up and filed.  Or teesting and filing bugs.
17:55:51 <davidlenwell> our long term goals are to be an infra project
17:56:09 <zehicle_> I need to build a sched but I never manage to get where I'm going
17:56:14 <davidlenwell> you guys are free to meet up and work without me.. wouldn't consider them meetings
17:56:37 <zehicle_> so it's like a tasty desert that I never get to eat
17:56:39 <davidlenwell> zehicle_: I'll rangle you around on tueaday .. covers most of the refstack things
17:56:52 <rockyg> So that fits meeting should be on tuesday or later.  Tuesday is tight so, Wednesday?
17:57:03 <zehicle_> once I make it to the design summit, I can usually seek in
17:57:35 <davidlenwell> what fits meeting ?
17:57:46 <rockyg> Or we could just announce we'll be hanfging out somewhere for some timeperiod for q/a
17:57:52 <davidlenwell> lets move offline .. we're out of time in here
17:57:58 <davidlenwell> #endmeeting