15:00:31 <sigmavirus> #startmeeting craton
15:00:31 <openstack> Meeting started Mon Dec  5 15:00:31 2016 UTC and is due to finish in 60 minutes.  The chair is sigmavirus. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:32 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:34 <openstack> The meeting name has been set to 'craton'
15:00:35 <sigmavirus> #chair jimbaker
15:00:36 <openstack> Current chairs: jimbaker sigmavirus
15:00:53 <sigmavirus> #link https://etherpad.openstack.org/p/craton-meetings
15:01:31 <sigmavirus> #topic Roll Call
15:01:32 <sigmavirus> o/
15:03:07 <sulo> o/
15:04:04 <sigmavirus> Looks like it's just the two of us sulo
15:04:10 <sulo> :(
15:05:16 * sigmavirus shrugs
15:05:20 <sigmavirus> #topic Updates
15:05:27 <sigmavirus> Any updates on your end sulo ?
15:05:43 <sulo> not much other than i've update thew wf blueprint
15:05:56 <sulo> https://blueprints.launchpad.net/craton/+spec/craton-workflow-engine
15:08:10 <sulo> so i am working mosting on that
15:08:18 <sulo> most of it i have it in some form in code
15:08:29 <sulo> but watned to put that up for discussion
15:09:45 <jimbaker> o/
15:09:49 <jimbaker> sigmavirus, sulo, sorry, overslept here :)
15:10:04 <sulo> jimbaker: np, we just started
15:10:04 * sigmavirus is just back from vacation
15:10:06 <jimbaker> sulo, thanks for that blueprint
15:10:15 <sigmavirus> I see tojuvone and toan are joining us today
15:10:43 <sigmavirus> jimbaker: any concise updates for us today?
15:11:13 <jimbaker> sigmavirus, right, conciseness is something i could work on...
15:11:59 <jimbaker> i have been working on blueprints about virtualized variables and how secrets are stored in craton
15:12:51 <jimbaker> virtualized variables are fairly straightforward, especially because we have discussed them at length
15:13:25 <jimbaker> secrets are potentially a rabbit hole that go as deep as we want it to go
15:13:52 <sigmavirus> jimbaker: so secrets are back?
15:13:58 <Syed__> o/
15:14:53 <jimbaker> sigmavirus, how else are we able to execute against target nodes? or against remote resources for virtualized variables?
15:15:15 <sigmavirus> jimbaker: a simple yes or no is all I need :)
15:15:21 <jimbaker> sigmavirus, yes
15:15:32 <sigmavirus> (not trying to pick a fight)
15:15:42 <sulo> heh, secrets are b a c k !!
15:15:56 <sigmavirus> or are they in the back? With the cookies? :P
15:16:21 <jimbaker> sigmavirus, i'm just trying to pick a scheme where we have secrets, but really it's deferred to another system
15:16:22 <sulo> sigmavirus: yes yes ;)
15:16:33 <sigmavirus> okay
15:17:23 <sigmavirus> jimbaker: what happened to gathering requirements and rbac?
15:17:27 <jimbaker> anyway, it's a topic we will have to discuss in greater depth
15:17:39 <sigmavirus> jimbaker: let's queue that for tomorrow's video meeting?
15:17:43 <jimbaker> sigmavirus, requirements were further gathered
15:18:21 <sigmavirus> good
15:18:22 <jimbaker> so good discussion with rackspace private cloud about what they need
15:18:29 <sigmavirus> woot
15:18:40 <jimbaker> we will be working on taking that doc and publishing on a public etherpad this week
15:19:25 <jimbaker> sigmavirus, with respect to rbac, i think we need to gather more feedback on the blueprint i put out last week
15:19:57 <sigmavirus> Fair, I didn't have enough time to review it then. It's on my list for today
15:20:03 * sigmavirus crosses fingers hoping to get to it
15:20:03 <Syed__> i believe it would be better to discuss rbac in tomorrow's vidyo
15:20:04 <jimbaker> anyway, vidyo meeting on all three tomorrow sounds good
15:20:07 <sigmavirus> Syed__: any uupdates?
15:20:16 <sigmavirus> or just updates, either way
15:20:29 <Syed__> yeah. been working on gathering information for Rbac
15:20:54 <Syed__> i have one patch in the queue. Today will try uploading two more for networks tests and routes.py tests
15:23:58 <sulo> ok lets move on ?
15:24:03 <jimbaker> +1
15:24:30 <sigmavirus> sulo: move on to?
15:24:36 <sigmavirus> Priority Reviews maybe?
15:24:41 <sigmavirus> #topic Priority Reviews
15:24:57 <sulo> i dont have any reviews pending :)
15:25:21 <sulo> however, not sure if git-harry is in today .. but he has a lot of open review that
15:25:35 <jimbaker> at this point, it's really git-harry
15:26:05 <git-harry> I'm in but need to go in 2 minutes. Saw the -2, will rework (probably go back to my original patch)
15:26:07 <jimbaker> with the bulk of reviews. but i think the root has been -2 by me, so need to revisit
15:26:08 <sulo> yeah so i was goint to say they need review that i've -1'ed
15:26:30 <sulo> but i thik those are teh ones that are in chain and have a -2 on endpoint
15:26:45 <sulo> jimbaker: yeah i think thats right
15:27:13 <jimbaker> git-harry, so we agreed to keep /v1/hosts, /v1/regions, etc
15:27:27 <git-harry> Yeah, I saw that
15:27:30 <jimbaker> i guess no dispute on /v1/regions... anyway
15:27:53 <jimbaker> so keep flat, but actually make it work for querying more generally
15:28:18 <jimbaker> we could start by removing region_id as a required param
15:28:50 <git-harry> Yeah, it's fine.
15:28:51 <jimbaker> which presumably should just work with the python-cratonclient as it is
15:29:00 <git-harry> Anyway, gotta go
15:29:17 <tojuvone> hieulq, yes here, have overlapping meeting. Need to go soon.
15:30:09 <jimbaker> git-harry, np, thanks for looking into this
15:30:49 <jimbaker> sigmavirus, so am i'm correct on python-cratonclient?
15:31:39 <sulo> yeah it should work without change .. atleast the api .. as region_id is something that is passed
15:31:50 <sulo> so should work with  minimal change if any
15:31:56 <sigmavirus> jimbaker: I'm not sure what you're talking about, I'd need a link to brush up on it and confirmd/eny
15:32:17 <jimbaker> yeah, it's just a question of whether the client is specifically checking for this param or not
15:32:34 <jimbaker> but agreed about minimal change (if any)
15:33:07 <jimbaker> going forward, it's certainly easy for the client - just needs to handle pagination as necessary
15:35:54 <jimbaker> that's it for priority reviews - i see that Syed__ just put in something into review a few minutes ago
15:36:29 <Syed__> There is one another for me
15:36:33 <Syed__> For network endpoints
15:36:52 <Syed__> The one for networks is actually dependent for me for routes and networks tests
15:36:53 <Syed__> :)
15:37:06 <jimbaker> Syed__, url ?
15:37:09 <jimbaker> of the review
15:37:29 <jimbaker> is this https://review.openstack.org/#/c/400977/ ?
15:37:43 <Syed__> https://review.openstack.org/#/c/400977/
15:37:45 <Syed__> Yeap
15:37:50 <Syed__> Thats the one jim
15:38:06 <jimbaker> ok, i will test. especially since sulo hasn't actually tested it directly
15:38:33 <Syed__> Sure
15:38:34 <Syed__> Thank you
15:39:12 <jimbaker> Syed__, were you able to add a functional test to this change?
15:39:31 <jimbaker> (now that we have functional testing, we should be using it for future changes....)
15:39:34 <Syed__> No this change doesnt needs one i believe
15:39:46 <Syed__> I will add more functional tests for regions, cells etc this week
15:39:53 <Syed__> Will raise bugs individually for those
15:40:05 <sulo> Syed__: you can leave them .. i have them already
15:40:08 <jimbaker> Syed__, i', pretty sure if there's an endpoint involved, it will require one :)
15:40:20 <Syed__> Ohh i see. I didnt knew that.
15:40:21 <sulo> was waiting for endpoint change to go though
15:40:27 <Syed__> Cool sulo
15:40:28 <Syed__> Thank you
15:40:47 <jimbaker> anyway, usual transition
15:41:14 <jimbaker> the other consideration we will have is how to bring in the python client into the functional testing
15:41:31 <sigmavirus> jimbaker: I don't think the client should be part of functional testing
15:41:39 <sigmavirus> the client and server should have their own individual functional tests
15:41:51 <sulo> jimbaker: you mean from teh client side ?
15:41:56 <sigmavirus> the client shouldn't be used in the server's functional tests
15:42:24 <sigmavirus> if the server changes something that breaks the client then that change becomes dependent on the fix in the client, which is dependent on the change on the server and you have a loop that will never merge
15:42:26 <jimbaker> sigmavirus, yeah, sure, we might call this integration testing. but similar leverage of the functional test framework
15:42:27 <sigmavirus> it's a terrible diea
15:42:39 <jimbaker> sigmavirus, oh i agree it's horribly complicated
15:43:05 <jimbaker> but having something that does this test is still going to be important
15:43:59 <sigmavirus> jimbaker: right, what you want is for the client to have functional tests that run the server
15:44:00 <jimbaker> sigmavirus, i don't know where these tests are currently done in openstack projects. but hopefully the combination of standard client + service is tested on a regular basis, challenges and all
15:44:02 <sigmavirus> I agree, and we'll set that up
15:44:08 <sulo> i am totes confused ... are we talking about putting client in server functional test or doing functional test for client ?
15:44:23 <sigmavirus> sulo: I understood the former, but I"m advocating solely for the latter
15:44:26 <jimbaker> sulo, i'm just saying, do test
15:44:45 <sulo> ok
15:45:11 <jimbaker> the specifics i'm completely open to. it sounds like the client testing will drive, which is absolutely fine for me
15:46:17 <jimbaker> which means the functional testing of the server code itself will purely look at its rest api, as is being done now
15:46:56 <jimbaker> and speculating: the client will probably reuse the server functional test harness to do its own testing
15:48:02 <jimbaker> if the server breaks its contract, the client tests will all start failing on a recheck. but at least we can fix the server, and get out of the pernicious loop that sigmavirus mentioned
15:48:04 <sigmavirus> jimbaker: potentially
15:48:14 <sigmavirus> (reuse the harness)
15:48:39 <jimbaker> sigmavirus, yeah, just throwing something out there
15:50:12 <jimbaker> anything else we should discuss in this meeting?
15:52:08 <jimbaker> i guess that's a *no*
15:52:37 <Syed__> i am done
15:52:44 <Syed__> nothing from my side
15:52:45 <Syed__> thanks jim
15:52:50 <tojuvone> Nothing that special
15:52:58 <jimbaker> let's move additional discussion to #craton
15:53:07 <Syed__> +1
15:53:09 <sigmavirus> #endmeeting