22:00:33 <rockyg> #startmeeting RefStack
22:00:34 <openstack> Meeting started Thu Mar 13 22:00:33 2014 UTC and is due to finish in 60 minutes.  The chair is rockyg. Information about MeetBot at http://wiki.debian.org/MeetBot.
22:00:35 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
22:00:37 <openstack> The meeting name has been set to 'refstack'
22:01:03 <rockyg> Roll call
22:01:22 <davidlenwell> oh did the time shift with day light savings?
22:01:28 <rockyg> Yup.
22:01:30 <davidlenwell> we've all been talking in #refstack for about an hour
22:01:50 <fcarpenter> that makes sense
22:02:08 <davidlenwell> yeah .. its all utc
22:02:17 <davidlenwell> emerica!
22:02:24 <davidlenwell> here!
22:02:29 <rockyg> I caught the tail.  I'm fighting a network driver incompatibility in Ubuntu.  Frustrating and getting me to that obsessive, kill all mode
22:02:47 <catherineD> here
22:02:47 <davidlenwell> rockyg: drivers in ubuntu can have that effect
22:02:52 <fcarpenter> here
22:03:08 <tedchang> hrer
22:03:11 <tedchang> here
22:03:19 <rockyg> Except for the kernel.  The driver hangs, goes zombie to the point that the machine needs a hard boot to get it back.  Won't soft reboot
22:03:25 <catherineD> Raymond is also here
22:03:31 <davidlenwell> I think rob is on vacation so its just us ..
22:04:02 <davidlenwell> agenda:  status, goals
22:04:06 <rockyg> Cool.  Forgot to come up with an agenda as usual, but I think a current stat of cod would be a good start.
22:04:12 <rockyg> #topic Status
22:04:27 <davidlenwell> progress!!!
22:04:41 <davidlenwell> we've landed a few important commits in the last week
22:04:43 <rockyg> Yay!  I've seen a number of check ins
22:04:56 <rockyg> and read the DB schema checkin
22:05:20 <davidlenwell> also the api public function for recieving results
22:05:25 <RaymondWong> i am going to check in that "read the DB" code today.
22:05:25 <rockyg> I see the hardcoding is giving way to better methods.
22:05:45 <davidlenwell> that part hasn't fully landed yet
22:06:09 <rockyg> Yeah.  But some great learning and education about python best practices going on.
22:06:13 <davidlenwell> catherineD, tedchang and raymod have put up some  blue prints and some initial commits .. but we are getting close
22:06:31 <davidlenwell> thanks to fcarpenter
22:06:43 <davidlenwell> who is an excelent code reviewer
22:06:56 <fcarpenter> i do my best
22:06:57 <rockyg> and probably knows how to spell excellent
22:07:03 <rockyg> ;-)
22:07:10 <fcarpenter> e-x-c-something something
22:07:37 <davidlenwell> ha.. my irc is in a terminal and does not have the spell check that macos has made me rely on so heavily
22:08:10 <davidlenwell> moving along
22:08:20 <davidlenwell> goals before next week?
22:08:30 <rockyg> It's Thursday and lots has been happening, but my question is: how close are we to me being able to run a refstack test against a devstack cloud?
22:08:52 <davidlenwell> I'd say by next thursday you'll be able to do that and see an html score card
22:09:07 <rockyg> YYEESSSSSS!
22:09:16 <davidlenwell> rockyg: at that time you and I should plan some facetime to pimp out our docs and wikis
22:09:43 <davidlenwell> from there we'll need to load test and set some goals for getting us managed by infra
22:10:06 <rockyg> So, I think we have our goals identified.  next question is:  from RefStack.org gui or from package I install?
22:10:24 <davidlenwell> not yet .. will have that going by monday
22:10:40 <davidlenwell> both
22:10:52 <rockyg> I'm really looking to get deeper doc stuff.  I am a top down and bottom up kinda engineer, so the middle out is the hardest for me.
22:11:00 <davidlenwell> by that I mean that on monday I'll have things publishing live as they merge
22:11:25 <davidlenwell> so that when code reviews happen and things merge they are on refstack.org shorly after
22:11:38 <rockyg> Fantastic.
22:12:08 <davidlenwell> the goal all along is to have both remote and internal official testing
22:12:41 <rockyg> Another part of status:  the fits mailing list and should we send an email to the dev list to let folks know who/where/what we are?  Sounds like next Friday or Monday for the dev email?
22:13:38 <rockyg> fcarpenter: is there a python way to generate a db schema doc from the code?
22:14:12 <davidlenwell> spynx
22:14:31 <rockyg> OK.  I'll spend time digging deeper into it.
22:14:36 <davidlenwell> and that should be another engineering task .. setup auto generation of docs
22:15:12 <davidlenwell> it can actually be easily done in the merge process with a commit to the infra-config repo .. I think .. < fungi??
22:15:55 <davidlenwell> I think that puts docs in the correct place on openstack.org or something unless I just made that up ..
22:16:22 <rockyg> :-)
22:16:53 <davidlenwell> I did attend an infra boot camp . but I've slept since then
22:17:11 <rockyg> I want to link to it from the wiki, but will need it to be a live link that changes with the merges.
22:17:27 <fungi> well, if you're talking about generating/updating code documentation on each commit, we do that currently for api docs
22:17:29 <rockyg> Sounds like that's what I need. Infra bootcamp.
22:17:43 <davidlenwell> fungi: yes .. that is what I am reffering to
22:17:50 <rockyg> DB schema docs, too.
22:18:39 <fungi> the docs below http://api.openstack.org/api-ref.html
22:18:54 <fungi> like http://api.openstack.org/api-ref-compute.html
22:19:04 <rockyg> OK.  anything else on status?  Runnable local and remote refstack working by 3/20 CoB?
22:19:15 <fungi> those are dynamically generated using sphinx from jenkins jobs run after each commit merges
22:19:49 <davidlenwell> rockyg: sounds right
22:20:20 <rockyg> So we just need to provide location for the api file (pretty obvious) and the DB file (less obvious, but we can workit out)
22:20:57 <davidlenwell> do you mean a url?
22:21:05 <davidlenwell> oh .. that
22:21:07 <davidlenwell> yeah
22:21:19 <davidlenwell> we need to move all the api calls into the api file .. right now they are in web.py
22:21:20 <davidlenwell> I can do that
22:21:40 <davidlenwell> and catherineD can you guys please put any new api calls into that file ?
22:21:46 <rockyg> OK. That covers status and short term goals.  What about long term  davidlenwell:  yeah. URLs for the auto-genned docs.
22:21:49 <davidlenwell> I'll make sure they get loaded correctly
22:22:13 <catherineD> sure will do
22:22:19 <catherineD> for the new API
22:22:31 <davidlenwell> interface
22:22:41 <davidlenwell> rockyg: .. we have a lot of work to do on the ui
22:23:03 <davidlenwell> we should have a ftf ux planning event and work out how this app will look behave and provide
22:23:20 <rockyg> Do we want a default of some small number of tests for the refstack runs?  Do we already have that?
22:23:24 <davidlenwell> rob has a lot of opinions about that so we'll need to coordinate with him
22:23:39 <davidlenwell> by default we want refstack to always run all tests
22:23:52 <davidlenwell> right now we are limiting it because we want to be able to debug faster
22:23:56 <rockyg> Right, but this is for trial run purposes
22:24:00 <davidlenwell> correct
22:24:34 <rockyg> Maybe the gui defaults to show trial and you could select full as an alternative?
22:24:57 <rockyg> Until we are farther along.
22:25:04 <davidlenwell> rockyg: thats certaily an option.. want to file it as a blue print and put some stuff in the whiteboard
22:25:26 <davidlenwell> we'll define it more clearly and get it into the pipeline
22:25:42 <rockyg> Sure.  also, have we figured out how we are going to let users select test sets?
22:26:01 <davidlenwell> im not sure we're going to
22:26:12 <RaymondWong> if they are running local copy, they can set the test sets in the config file.
22:26:38 <davidlenwell> I was speaking for refstack.org
22:27:08 <rockyg> Maybe submit a config file to refstack.org?
22:27:21 <davidlenwell> I wouldn't want to do that
22:27:29 <davidlenwell> if they want custom config they can run a local copy
22:27:38 <davidlenwell> refstack.org will be for official testing only
22:27:42 <rockyg> OK.
22:27:50 <davidlenwell> by verified vendors
22:28:08 <davidlenwell> vendors who want an official openstack seal on their website
22:28:22 <davidlenwell> that will link into a report card on refstack
22:28:41 <davidlenwell> but we still want all the results from everyone
22:28:58 <davidlenwell> then we can built interesting interop maps and really give users fun tools that are useful
22:28:59 <rockyg> Then users and ops would run the package instead.
22:29:50 <davidlenwell> yes
22:30:08 <rockyg> #action    I'm gonna make sure this gets captured in the requirements
22:30:15 <davidlenwell> rockyg: thank you
22:31:02 <davidlenwell> anything else?
22:31:06 <rockyg> #topic open discussion
22:32:25 <rockyg> Don't forget to sign up to fits@lists.openstack.org
22:32:36 <davidlenwell> I think I did that the day they brought it up
22:32:43 <catherineD> thx rockyg:
22:33:01 <rockyg> :-)
22:33:17 <catherineD> any status about Rally?
22:33:38 <catherineD> are we (RefStack and Rally) working together
22:34:48 <rockyg> Have not seen any dev list postings.  I heard that Mirantis is going to join us in force soon.  That's why I want to make sure we're using the mailing list instead of our internal collected emails.  but, I haven't seen many new people jump on board.
22:34:54 <davidlenwell> nobody from rally has talked to me directly
22:35:33 <davidlenwell> I would encourage them to do so if we are supposed to work together
22:36:12 <rockyg> Oh, I have a request for davidlenwell and/or fcarpenter:  could one/both of you write a simple unit test so that catherine and team has a good example to start from?
22:36:42 <davidlenwell> probably won't have the head space for it until wednesday .. but I can make that happen
22:36:45 <davidlenwell> file a blue print
22:36:52 <davidlenwell> that is what I am going to work from
22:37:14 <rockyg> Knowing the system makes it easier to write more by others.  And, I will file a blueprint;-)
22:37:25 <davidlenwell> rockyg: agreed
22:37:50 <davidlenwell> there is a stub of a test in there.. needed it to get into stackforge .. but its just a stub
22:39:12 <rockyg> #action  two blueprints:  unit test and auto gen docs and ??  I think there's one more.
22:39:45 <davidlenwell> actually if we are going to do this right we should have unit tests updated for every merge
22:40:22 <davidlenwell> shouldn't pass review if it doesn't have propper test coverage.. it can just be labor intensive and I am horrible about keeping upt hat habit my self
22:41:00 <davidlenwell> but we'll reach a critical mass of code where we'll need to rely on those tests to make sure things aren't broken
22:41:18 <davidlenwell> and I'd rather start now than have to rush it all in later
22:41:21 <rockyg> Well, once the working demo is landed, hopefully there will be a little more time to write tests.
22:41:29 <davidlenwell> but this should probably be a goal for the next sprint
22:41:51 <davidlenwell> I digress
22:41:59 <rockyg> Especially now that we are moving from hardcoded to configurable.
22:42:06 <fcarpenter> educate the new guy: what's the sprint schedule for refstack (begin/end)?
22:42:16 <rockyg> topic is open discussion.
22:42:31 <davidlenwell> its two weeks fcarpenter.. we are midway now
22:42:36 <fcarpenter> thank you
22:42:46 <rockyg> mid term goal is RefStack for Havana by Icehouse summit
22:42:47 <davidlenwell> should line up with piston
22:43:53 <rockyg> But, we're not great on getting our stories straight;-)  Right now, it's tasks.
22:44:11 <davidlenwell> rockyg: correct
22:44:20 <davidlenwell> infra is working on something called storyboard
22:44:30 <davidlenwell> its going to give openstack more real scrum than lp
22:45:05 <rockyg> Yeah.  Attended the IRC this morning.  but I thought it was a replacement for BPs and bugs and project management....
22:45:27 <rockyg> It sounds like HP is hiring a UI guy for the effort.
22:46:15 <davidlenwell> rockyg: yaeh .. they're all about it
22:46:48 <davidlenwell> I hope they build in a migration path from pivotal tracker and piston can use them .. "hope"
22:47:15 <davidlenwell> okay . im good with ending now . have to get back to packing
22:47:30 <rockyg> Considering I think LP BPs and bugs system is one of the worst I've had to navigate in quite a while, I was hoping they would do something like use Pivitol's tools.
22:47:44 <tedchang__> good luck :)
22:47:47 <davidlenwell> its not open source
22:47:59 <davidlenwell> rockyg: infra won't adopt anything new that isn't open source .. period
22:48:06 <davidlenwell> and apache lic .. I think
22:48:09 <rockyg> When you make it up here, consider dropping by for some wine in my backyard on some Saturday.  goes for everyone here.
22:48:32 <davidlenwell> are you in south bay rocky?
22:48:52 <rockyg> Yup.
22:49:13 <rockyg> Easy access from highways.  It's supposed to get to 78 Saturday.
22:49:24 <fcarpenter> davidlenwell: I can drive you ;)
22:50:00 <rockyg> Know where you're going to rent yet?  Pacifica?
22:50:40 <rockyg> #endmeeting  See you guys on refstack.