22:02:20 <rockyg> #startmeeting RefStack
22:02:21 <openstack> Meeting started Thu Mar  6 22:02:20 2014 UTC and is due to finish in 60 minutes.  The chair is rockyg. Information about MeetBot at http://wiki.debian.org/MeetBot.
22:02:22 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
22:02:25 <openstack> The meeting name has been set to 'refstack'
22:03:04 <zehicle_at_dell> o/
22:03:37 <catherine|2> here
22:03:38 <waiman> hi
22:03:55 <rockyg> o/
22:04:27 <rockyg> rollcall for those who haven't typed in since meeting start...
22:06:21 <rockyg> agenda:  Update from zehicle_at_dell on refstack demo and defcore discussion at board; project status; selection of next sprint stories; schedule?; open discussion
22:07:24 <rockyg> hi tedchang_
22:08:05 <rockyg> Let's start with report from zehicle_at_dell about board meeting decisions/discussions
22:08:24 <tedchang_> Hi :)
22:08:37 <zehicle_at_dell> ok
22:08:45 <zehicle_at_dell> lots to report
22:09:00 <sarob_> yo
22:09:18 <zehicle_at_dell> we reported about the progres  of the program overall
22:09:36 <zehicle_at_dell> and got board OK to keep going with expanding the community involvement in the process
22:10:05 <zehicle_at_dell> we did NOT demo TCUP but asked the board members to identify people who would run it as a alpha user group
22:10:17 <zehicle_at_dell> we said that we would start that program next week
22:10:44 <zehicle_at_dell> We worked out an agreement with the driver testing team from Mirantis
22:10:54 <zehicle_at_dell> that they would bring that code into RefStack
22:11:19 <zehicle_at_dell> and we also talked to Monty from the Infra team to start the process of getting a FOO.openstack.org domain
22:11:27 <zehicle_at_dell> running the RefStack code started
22:11:32 <zehicle_at_dell> FOO = TBD
22:12:14 <zehicle_at_dell> outside of the board meeting, I also talked to Boris about considering if TCUP could be used as an alternate method of driver testing
22:12:20 <zehicle_at_dell> that will take some community discussion
22:12:44 <zehicle_at_dell> I think those are the the highlights
22:13:03 <zehicle_at_dell> question?
22:13:50 <malini> Would like to comment from my attendance at the Operator Summit March 3 at eBay.
22:14:00 <rockyg> kewl
22:14:15 <catherine|2> For the alpha user group, do they stand up their own RefStack?
22:14:21 <malini> Their action item was to get involved in filing bugs, joing blueprint discussions and stability stability"
22:14:27 <malini> soo provide test use cases
22:14:55 <rockyg> Excellent
22:15:06 <malini> may be we should get them to help rank defcore tests/weights
22:15:14 <zehicle_at_dell> we talked about the operator summit at the board meeting too - I was wondering if the operators would find the capabilities chart we are making helpful
22:15:30 <zehicle_at_dell> malini, yes, that's what we want from refstack
22:15:45 <zehicle_at_dell> the plan is to get the capabilities up on the website and start collecting results from TCUP
22:15:53 <malini> I think so, there was also a sense of being excluded, free pass, credit for reviews/bugs
22:16:04 <rockyg> we are thingking of having a voting method for identifying the most wanted/needed capabilities
22:16:12 <zehicle_at_dell> then we will allow people to vote +/- on their preference for that capability
22:16:16 <sarob_> stefano and tom will work on the pass thing
22:16:28 <zehicle_at_dell> how we score that will be interesting - will need a dedicated meeting to discuss
22:16:36 <malini> nice -- voting :-)
22:16:45 <sarob_> i joined the storyboard meet this morning
22:17:13 <sarob_> providing more blueprint information and operator feedback is part of their plans
22:17:30 <rockyg> It would also be great to get some alphas from the ops community.
22:17:49 <zehicle_at_dell> I'm excited to see malini and sarob_ on!  anyone else missed in the roll call?
22:18:16 <malini> :-) thank you zehicle_at_dell !
22:18:30 * sarob_ yip yip
22:18:48 <rockyg> I think a lot of the user community finds it difficult to follow blueprints beyond the blueprint and included links.  Going off to reviews gets very confusing.  We need a better feedback/discussion mechanism on bps
22:19:24 <zehicle_at_dell> as a use-case, it would be great to link capabilities (which are from tests) back to blueprints
22:19:54 * zehicle_at_dell lasso we're a bit off topic
22:20:09 * zehicle_at_dell sets_expectations that's likely for the next cycle after the summit
22:20:45 <rockyg> ok.  let's get back on topic.  last call for board meeting discussion
22:21:18 <rockyg> Oops..I forgot to set topic.  Dang.
22:21:33 <catherine|2> About tests by alpha group, will they set up their own refstack?
22:21:47 <rockyg> #topic project status
22:22:14 <zehicle_at_dell> alpha?
22:22:56 <catherine|2> we talk about have some group to test and collect results  ..
22:23:00 <rockyg> we need to provide either source or some sort of distro (zipped tar?) that will give them tcup and gui.
22:24:16 <catherine|2> They can download refstack from github ..
22:24:27 <rockyg> So let's see where we are on having a tcup we can hand out and the simple gui you guys put together.  I know there was code check in but I haven't reviewed it yet.
22:24:27 <catherine|2> to their environment and test ..
22:25:32 <rockyg> download will work as long as we can provide build instructions.  Seems to me we want dev-ops, rather than just ops folks for alpha.
22:26:16 <zehicle_at_dell> Josh worked out a a way to inject the dockerfile
22:26:28 <zehicle_at_dell> and we have a kick.py to start it off
22:26:37 <zehicle_at_dell> we made a specific design decision on Monday:
22:26:40 <rockyg> Let's hear where we are then discuss what needs to get done to get a usable download.
22:26:45 <zehicle_at_dell> 1) you clone the git repo
22:26:54 <zehicle_at_dell> 0) install docker (I always forget 0)
22:27:08 <zehicle_at_dell> 2) run the kick.py script
22:27:24 <zehicle_at_dell> 1.5) source openrc.sh
22:27:34 <zehicle_at_dell> need coffee, I'm getting out of order
22:27:54 <rockyg> zehicle_at_dell: also test results are in csv format
22:28:22 <rockyg> I'll clean up the order numbers when I put on in the wiki ;-)
22:28:26 <zehicle_at_dell> thanks
22:28:37 <davidlenwell> sorry im late guys
22:28:49 <davidlenwell> reading scroll back now
22:28:50 <rockyg> Where do the test come from? tempest repo?
22:29:10 <sarob_> let me know where to start, i can be an alpha tester
22:29:17 <zehicle_at_dell> I think that we have some control over that
22:29:50 <zehicle_at_dell> but I think the thought is to prefer a repo
22:29:52 <rockyg> And we have to remember they are havana tests
22:31:06 <davidlenwell> caught up
22:32:05 <davidlenwell> so it sounds like we are done with this topic
22:32:15 <rockyg> The other question that will come up a lot is whether more packages will need to be loaded to the test vm to run refstack
22:32:34 <zehicle_at_dell> expect when we are going to have it
22:32:34 <zehicle_at_dell> so, a little more time
22:32:37 <rockyg> do you have any status to add?  API or db schema?
22:32:48 <zehicle_at_dell> we're still missing the config step
22:33:00 <zehicle_at_dell> so the docker file gets the env variables
22:33:03 <rockyg> Ah, yes.
22:33:11 <zehicle_at_dell> and can run tempest, but we do not build the tempest config yet
22:33:24 <zehicle_at_dell> the good news is that we don't have to do that in the container, we can just test it
22:33:38 <zehicle_at_dell> and the container will read it out of the local diretory because of file mappings
22:33:39 <rockyg> how?
22:33:44 <davidlenwell> zehicle_at_dell: have you looked at the code that has merged today?
22:33:51 <zehicle_at_dell> actually, it would create it from the code
22:33:55 <zehicle_at_dell> no!  joy?
22:34:00 <davidlenwell> much joy
22:34:09 <zehicle_at_dell> I'm behind then - can you give an update?
22:34:10 <rockyg> Nope.  Kewler.
22:34:15 * zehicle_at_dell catching up with day job
22:34:21 <davidlenwell> we've merged most of the changes from catherines team
22:34:46 <davidlenwell> two are hung up and need rebase before they can pass tests.. but they have been fantastic about our nit picky changes
22:35:15 <rockyg> +1
22:35:31 <davidlenwell> I'm hoping we can handle those today
22:35:32 <catherine|2> not all nit picky .. many are great suggestion ..
22:35:47 <davidlenwell> yeah .. we had 3 piston folks code reviewing
22:36:15 <davidlenwell> anyways if you guys will talk to me after in #refstack we'll get the rest in
22:36:31 <davidlenwell> then josh has a patch that uses env variables and runs as root and is more secure
22:36:52 <catherine|2> we will talk to you after this ... David
22:36:57 <davidlenwell> excelent
22:37:18 <zehicle_at_dell> davidlenwell, so did we get those two parts glued together too?  what's the delta once the patches land?
22:37:44 <davidlenwell> zehicle_at_dell: not just yet .. will have to look at it after josh submits its for review
22:37:50 <zehicle_at_dell> kk
22:37:51 <davidlenwell> he was just telling me he had it working this morning
22:38:15 <zehicle_at_dell> I've been trying to get to it - maybe tomorrow
22:38:16 <davidlenwell> So I have three open tasks..
22:38:26 <zehicle_at_dell> but that would let me start testing it
22:38:51 <davidlenwell> one. update db schema to what was discussed a few weeks ago in our f2f
22:39:20 <davidlenwell> two. write code for parsing results into score cards and indexable data
22:39:52 <davidlenwell> three. redeploy the new version of refstack on refstack.org and set up ci for it .. so that it updates when stuff merges without me having to ssh in and do it manually
22:40:06 <davidlenwell> I'm moving to the bay area next week and am packing like a mad man this week ..
22:40:23 <rockyg> 1.5 get me the diagrams/schemas for archtecture and db
22:40:30 <rockyg> Welcome!
22:40:34 <davidlenwell> added to my list rockyg
22:41:07 <davidlenwell> I am running a bit behind on a lot of things as a result .. but I am still putting in at least 6 hours a day and should get through at least the first one on my list before the weekend
22:41:21 <rockyg> If we can get one and 2, can we run and get test results in csv form?
22:41:28 <malini> davidlenwell welcome! other than the real estate prices, i like it here, moved 20 months back
22:42:06 <davidlenwell> rockyg: im sure thats possible .. it might actually be easier than I am thinking as I think I might have already written that code and forgotten to commit it
22:42:09 <rockyg> Great! we have critical mass for not just dev f2f but ops/training/docs f2f!
22:42:31 <rockyg> +2!
22:42:48 * zehicle_at_dell agenda_parking_lot driver testing UI work in RefStack
22:43:28 <davidlenwell> I just need to get my hands dirty on it tonight and I think its handled
22:43:36 <rockyg> OK.  Can we motve to next sprint work discussion?
22:43:38 <zehicle_at_dell> so, tomorrow we should be able to do an integrated TCUP test
22:43:49 <zehicle_at_dell> I'll block time for it tomorrow and hang out on #refstack
22:43:54 <davidlenwell> oh another thing on my plate refstack related is that for some reason stackalitics stopped tracking our project .. need to fix that
22:44:08 <davidlenwell> zehicle_at_dell: I think thats possible yes
22:44:15 <zehicle_at_dell> coolio
22:44:23 <davidlenwell> I'll make sure im around at least the first half of the day ..
22:44:27 <rockyg> I'll be there if I remember.
22:44:37 <zehicle_at_dell> stackalitics should be fixable considering Mirantis team coming over to play
22:44:38 <davidlenwell> also .. worth mentioning
22:44:45 <zehicle_at_dell> anyone from the Driver Test group?
22:44:48 <davidlenwell> zehicle_at_dell: indeed
22:45:04 <zehicle_at_dell> we should make sure to topic that
22:45:23 <zehicle_at_dell> it related to how we need to collect/track the capabilities -> tests map (was a brief topic from Monday)
22:45:30 <rockyg> #topic work for sprint2
22:46:33 <rockyg> Let's move davdlenwell's 3 to sprint 2
22:46:41 <davidlenwell> so I just laid out 4 tasks on my plate
22:46:52 <davidlenwell> if we count stackalitics
22:47:14 <rockyg> If we get the the integrated tcup test running tomorrow, what are the next steps?
22:47:18 <davidlenwell> rockyg: I'll make 1.5 part of 1 .. since I have a thing that will generate a diagram from sql
22:47:19 <zehicle_at_dell> is anyone from Mirantis on the call?  can someone take Stackalitics off davidlenwell plate?
22:47:39 <davidlenwell> zehicle_at_dell: that would be ideal
22:47:58 <rockyg> I can possibly get some help from Jerry here at Huawei.  He did some stackalytics fixes for Compass
22:47:59 <zehicle_at_dell> hold off on that then and see if we can expand it - not urgent
22:48:15 <zehicle_at_dell> rockyg, thanks.
22:48:30 <davidlenwell> altho it will retroactively track stuff .. so its not like we need it in now
22:48:45 <rockyg> I've also reviewed some of the code.  Pretty straightforward.
22:49:08 <rockyg> Or I can jsut file a bug;-)
22:49:31 <davidlenwell> yeah .. I am about to commit to it anyways to put some more piston people in the right files for tracking our stuff ..
22:49:35 <zehicle_at_dell> what other work is in the queue to drop?
22:49:41 <davidlenwell> so its not the end of the world if I just deal with it
22:49:48 <rockyg> So.  We have working integrated tcup.  Next development needs?
22:49:50 <davidlenwell> we need a remote execution thing
22:50:02 <davidlenwell> right now tcup gets launched on the server thats running refstack
22:50:10 <davidlenwell> which is fine for dev .. but won't scale
22:50:24 <zehicle_at_dell> can we get the upload post working first?
22:50:31 <rockyg> Ah.
22:50:33 <davidlenwell> of course
22:50:37 <davidlenwell> thats one of my list of three
22:50:41 <davidlenwell> this is for after that
22:50:49 <rockyg> 1. get post working
22:50:49 <zehicle_at_dell> also, who's working w/ infra team on hosting the site?
22:50:57 <davidlenwell> i fix database
22:51:03 <zehicle_at_dell> (oh, we need to take poll for naming
22:51:03 <davidlenwell> 2 get post working
22:51:04 <zehicle_at_dell> )
22:51:13 <davidlenwell> naming what ?
22:51:15 <rockyg> hosting refstack.org?
22:51:26 <davidlenwell> zehicle_at_dell: right now im hosting it
22:51:30 <zehicle_at_dell> refstack.openstack.org
22:51:32 <davidlenwell> until there is governance
22:51:40 <zehicle_at_dell> ok, that's enough for now -> let's move on
22:51:55 <davidlenwell> so back to remote execution
22:52:05 <rockyg> I can make some inquiries.  what about Catherine and her team?  What are they gonna do for sprint2?
22:52:16 <zehicle_at_dell> FWIW, with the addition of the driver test, they want to accelerate getting it hosted by infra and a base name
22:52:45 <davidlenwell> zehicle_at_dell: I'm all for that .. however until we're a little further along they aren't going to be happy about hosting us
22:52:54 <davidlenwell> which brings me to something I should share
22:53:01 <rockyg> Lemme look into the infra hosting, too.
22:53:03 <davidlenwell> we sort of got bashed last night in the infra room
22:53:21 <davidlenwell> rockyg: I'm in close contact with them and have good relationships with the team
22:53:28 <rockyg> OK.
22:53:32 <davidlenwell> rockyg: I've already broached the subject of them hosting it ..
22:53:39 <davidlenwell> I don't think it will be a problem
22:53:43 <zehicle_at_dell> sounds like a topic
22:53:45 <davidlenwell> but right now we aren't built right
22:54:10 <davidlenwell> so they were bashing us for our use of docker.. one of them tolt me we arne't very openstacky
22:54:26 <davidlenwell> #topic infra hosting us
22:54:47 <davidlenwell> we need a scaleable testing model
22:55:06 <davidlenwell> I plan to use gearman for that .. run a pool of workers to run tests
22:55:18 <davidlenwell> its pretty much how they are used to doing things
22:55:31 <rockyg> It's one of the selling points of refstack-- you don't have to build a new stack to run it against your openstack....
22:56:04 <davidlenwell> we'd also have to write a bunch of puppet stuff in order for them to be able to host it
22:56:10 <davidlenwell> everything they deploy is puppet
22:56:28 <davidlenwell> piston is close to hiring someone mostly to work for the infra team ..
22:56:37 <davidlenwell> once we do I can make that his first assignment
22:57:06 <davidlenwell> I think we can get there in the next two sprints .. but we shouldn't spend a lot of time bothering them about it until we are ready
22:57:37 <zehicle_at_dell> +1
22:57:42 <rockyg> +1
22:58:07 <davidlenwell> in the mean time..  I can talk to catherine|2 after this about what tasks her team has time for ..
22:58:36 <catherine|2> davidlenwell: yes we are here ...
22:58:43 <zehicle_at_dell> last call for topics....
22:58:46 <davidlenwell> they aren't assigned to this project full time so we might loose 1 or all of them anytime now .. so I'd like to figure out how to use them as best we can
22:58:55 <catherine|2> really want to get the last 2 commits in ..
22:58:59 <rockyg> I'd like to take some of this discussion over to #refstack (with malini if possible) to get more detailed planning.
22:59:08 <zehicle_at_dell> we'll have to carry integrating Driver Test & DefCore Reports to next meeting
22:59:21 <davidlenwell> sounds good
22:59:42 <zehicle_at_dell> oh, for the record.  I think we decided to create a config file (type to be decided) that has the test -> capabliitiy mapping
22:59:51 <rockyg> yup.
22:59:55 <zehicle_at_dell> that would allow us to manage/version it in the repo very easily
23:00:11 <davidlenwell> exactly
23:00:16 <zehicle_at_dell> we may need to have different ones per OSrelease
23:00:26 <zehicle_at_dell> since they are going to drift per release
23:00:27 <davidlenwell> zehicle_at_dell: I think we will
23:00:33 <zehicle_at_dell> ok, +1
23:00:36 <davidlenwell> or a json structure in the file that allows lots of versions
23:00:41 <rockyg> I'd like to publicize the bps, the launchpad site and the wiki for the ops folks interested
23:00:48 <zehicle_at_dell> since there are >1000 tests already
23:01:00 <zehicle_at_dell> I'd rather not create an endlessly growing file
23:01:05 <zehicle_at_dell> would be harder to maintain
23:01:09 <davidlenwell> zehicle_at_dell: thats fine
23:01:10 <davidlenwell> we
23:01:18 <davidlenwell> 'we'll come up with a clean way of naming them
23:01:28 <rockyg> #link https://wiki.openstack.org/wiki/RefStack
23:01:31 <davidlenwell> okay .. lets end and move ofer to #refstack
23:01:49 <zehicle_at_dell> I've got to sign off - I'll join there tomorrow
23:02:12 <rockyg> #link https://launchpad.net/refstack
23:02:17 <zehicle_at_dell> great meeting everyone!  progress is great
23:02:27 <rockyg> #endmeeting