16:57:18 <davidlenwell> #startmeeting refstack
16:57:18 <openstack> Meeting started Thu May 29 16:57:18 2014 UTC and is due to finish in 60 minutes.  The chair is davidlenwell. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:57:19 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:57:22 <openstack> The meeting name has been set to 'refstack'
16:57:47 <davidlenwell> Roll call
16:59:49 <davidlenwell> roll call?
16:59:55 <zehicle_at_dell> o/
16:59:59 <rockyg> o/
17:00:04 <fcarpenter> o/
17:00:18 <davidlenwell> Do we have catherine|2?
17:00:31 <catherine|2> yes
17:00:37 <davidlenwell> excelent
17:00:38 <davidlenwell> agenda: storyboard, specs, f2f, open discussion
17:00:45 <davidlenwell> #topic storyboard
17:01:07 <davidlenwell> So I spent some time in the last week getting to know storyboard.
17:01:09 <zehicle_at_dell> could we also talk about Havana vs Icehouse tempest?
17:01:13 <davidlenwell> sure
17:01:25 <davidlenwell> we do that before specs
17:01:28 <davidlenwell> so next
17:02:01 <davidlenwell> while storyboard is an incomplete product.. I feel its as usable as launchpad.. and the storyboard team is very open to our feedback.
17:02:10 <zehicle_at_dell> +1
17:02:15 <davidlenwell> I'll probably even commit some code it to my self.
17:02:26 <zehicle_at_dell> would like to have names instead of IDs in the URL, FWIW
17:02:35 <davidlenwell> yep
17:02:41 <davidlenwell> I filed a story for that on friday
17:02:46 <davidlenwell> they want it to
17:02:48 <davidlenwell> too
17:03:28 <davidlenwell> So make sure when you are writing specs and making commits to reference your storyboard link.
17:03:57 <rockyg> good point.
17:04:07 <davidlenwell> for those of you who didn't catch my emails to fits  .. storyboard is at storyboard.openstack.org
17:04:22 <zehicle_at_dell> are all the blueprints migrated to storyboard at this pont?
17:04:24 <davidlenwell> if you have not already .. go sign in for the first time so I can assign you work ;)
17:04:24 <fcarpenter> will you update restack.org to include that link?
17:04:33 <davidlenwell> zehicle_at_dell: yes
17:04:43 <rockyg> i've filed the story and will file spec later today
17:04:46 <davidlenwell> fcarpenter: yes.. I will do that today.. there is actually a story that includes that task
17:05:50 <davidlenwell> okay .. so thats that .. everyone on board with storyboard?
17:06:04 <davidlenwell> any objections will need to be voiced now ;)
17:06:06 <catherine|2> yes
17:06:17 <davidlenwell> awesome.. moving right along..
17:06:23 <zehicle_at_dell> +1
17:06:30 <davidlenwell> #topic icehouse vs havana tempest
17:06:40 <sarob> It looks good
17:06:42 <davidlenwell> so this should be short
17:06:52 <davidlenwell> use trunk tempest always .. moving along
17:07:11 <zehicle_at_dell> +1 but want to talk about timing
17:07:12 <davidlenwell> branchless tempest is what we want.. when do we want it?  now
17:07:29 <zehicle_at_dell> since I know that we have to update TCUP
17:07:34 <davidlenwell> I was actually surprised not to see a commit updating the version of tempest the tester is using
17:07:44 <davidlenwell> but I guess I can do that since I wrote that
17:08:21 <davidlenwell> zehicle_at_dell: I'll update that as soon as I finish writing my specs
17:08:27 <zehicle_at_dell> thanks
17:08:28 <davidlenwell> which I plan to have done in the next 24 hours
17:08:42 <zehicle_at_dell> if we can get it running outside of tcyup that would be a good gate
17:08:51 <zehicle_at_dell> before I work on making it work inside of it
17:08:51 <catherine|2> there are about 600 testcase in for API in Havana... In Icehouse this number is about 2000+ ...
17:09:05 <davidlenwell> catherine|2: yes .. is that a problem?
17:09:38 <catherine|2> no except for test time and extra configuration needed
17:10:22 <catherine|2> I am trying to use master Tempest to test my Havana clouds ...
17:10:25 <davidlenwell> catherine|2: yes .. thats an issue.. but its an issue we have to solve refardless
17:10:36 <zehicle_at_dell> are the test IDs the same?
17:10:59 <zehicle_at_dell> it would be a problem for defcor if not
17:11:02 <catherine|2> test IDs --> full test case name ?
17:11:08 <zehicle_at_dell> since we need havana results for havana clouds
17:11:09 <zehicle_at_dell> yy
17:11:38 <catherine|2> yes I think it is more a problem for DefCore ..
17:12:08 <zehicle_at_dell> it should be OK unless the test really changes funciton but keeps the same name
17:12:15 <catherine|2> I am trying to take a inventory to see whether 1) full test name are retain ..2) what additional
17:12:19 <zehicle_at_dell> which I think would have much more serious implications
17:12:34 <davidlenwell> catherine|2: good thinking
17:12:46 <catherine|2> I don't think so but I am trying to verify with Havana cloud before going to Icehouse cloud ..
17:13:06 <davidlenwell> So there is no icehouse tempest .. just to be clear
17:13:23 <davidlenwell> after havana its branchless
17:13:23 <zehicle_at_dell> davidlenwell, I think moving to trunk tempest is the right thing but it's a bit of scope creep for Havana
17:13:38 <catherine|2> just using the sme master tempest for both Havana and Icehouse cloud ...
17:13:40 <zehicle_at_dell> totally able to justify, we just need to be transparent
17:13:54 <davidlenwell> of course
17:14:02 * zehicle_at_dell is thinking about impact on reports and happy to have single tempest
17:14:07 <catherine|2> since I have lots of data from testing havana tempest with havana cloud
17:14:08 <davidlenwell> I don't want to keep focusing on havana tempest
17:14:17 <catherine|2> I can make tht comparison ..
17:14:19 <zehicle_at_dell> davidlenwell, +1
17:14:31 <zehicle_at_dell> It seems to me that we
17:14:37 <zehicle_at_dell> are ok as long as the IDs are consistent
17:14:48 <zehicle_at_dell> since that's the ultimate handle we're using, not the source
17:14:49 <davidlenwell> even if they are not .. we have to find a way to migrate
17:15:10 <davidlenwell> branched tempest is a thing of the past
17:15:11 <zehicle_at_dell> if they are not, then it's even more urgent to move to the consinstent one
17:15:18 <davidlenwell> +1
17:15:25 <davidlenwell> okay .. moving on
17:15:29 <zehicle_at_dell> kk
17:15:30 <davidlenwell> #topic specs
17:15:51 <catherine|2> I am more concern about the consistency of test name/ID for core and capability test list
17:16:06 <davidlenwell> so specs are rolling in.. catherine|2 I started to review the one you posted last night but my eyes didn't want to read anymore.. so I'll review it this morning.
17:16:21 <davidlenwell> zehicle_at_dell: what is the status of your specs?
17:16:41 <zehicle_at_dell> I'm working on the designs for both
17:16:49 <zehicle_at_dell> making some progress on layout and approach
17:17:03 <davidlenwell> Do you have an eta?
17:17:08 <zehicle_at_dell> I'd be happy to upload partials if people want to see the problem statement
17:17:10 <zehicle_at_dell> tomorrow
17:17:24 <rockyg> I think partials would be informative
17:17:26 <zehicle_at_dell> I think I've cracked the design for the comparison one
17:17:29 <davidlenwell> I'd always like to see partials if you are into feedback
17:17:35 <zehicle_at_dell> happy to do that
17:17:40 <zehicle_at_dell> will upload by EOD for them
17:17:45 <davidlenwell> excelent
17:17:52 <davidlenwell> what about your specs rockyg
17:17:54 <davidlenwell> ?
17:18:08 <rockyg> Matrix will be in today.
17:18:18 <davidlenwell> yay!
17:18:23 <rockyg> Looks like you got the example unit test cooking?
17:18:26 <davidlenwell> and catherine|2 has one more riht ?
17:18:30 <zehicle_at_dell> I had a question about catherine|2 spec https://review.openstack.org/#/c/96311/
17:18:32 <davidlenwell> rockyg:  I did
17:18:35 <catherine|2> yes working on it
17:18:45 <zehicle_at_dell> it looks like it overlaps w/ the other API upload spec
17:18:59 <zehicle_at_dell> I called that out in review, but figured we could talk here if that helps
17:19:03 <rockyg> davidlenwell: great on example unit test
17:19:22 <davidlenwell> ha ha test_nothing is a good start
17:20:03 <rockyg> Dow we need a spec on breaking out the config into the three sections?  If so, we should do it cross project with storyboard.
17:20:21 <davidlenwell> rockyg: not yet
17:20:37 <rockyg> Can storyboard do cross project?
17:20:42 <davidlenwell> for our imidiate porposes we are going to consider config a blackbox
17:20:47 <davidlenwell> rockyg: yes it can
17:20:54 <rockyg> kewl x2
17:21:01 <zehicle_at_dell> one of the items I
17:21:13 <zehicle_at_dell> noticed in review was that we may want to have a client library for the API
17:21:23 <zehicle_at_dell> so that we could use it from multiple places in the same way
17:21:26 <davidlenwell> zehicle_at_dell: indeed
17:21:35 <rockyg> sounds right
17:21:42 <davidlenwell> lets add a story for that
17:22:02 <davidlenwell> assign it to me .. made the first task write a spec
17:22:17 <rockyg> +1
17:22:44 <davidlenwell> next topic
17:22:48 <davidlenwell> #topic f2f
17:22:49 <catherine|2> wait
17:23:05 <davidlenwell> yes catherine|2?
17:23:35 <catherine|2> zehicle_at_dell:suggest we should merge the api--api-sync with an API upload spec?
17:24:03 <catherine|2> I am open for that ..
17:24:08 <zehicle_at_dell> catherine|2, there are parts of it that overlap
17:24:16 <zehicle_at_dell> but I think you also have use-cases that are different
17:24:28 <zehicle_at_dell> I'd suggest that we pull the API spec out and use the current one
17:24:40 <zehicle_at_dell> but leave in the refstack to refstack synch part
17:24:54 <zehicle_at_dell> because that was distinct.
17:25:01 <catherine|2> davidlenwell: I avouid the sync word ...
17:25:09 <zehicle_at_dell> +1
17:25:27 <zehicle_at_dell> could we call it "private refstack to public refstack push" ?
17:25:29 <catherine|2> because I don't think we sync the data because both sides do not have identical data
17:25:35 <zehicle_at_dell> I think that's the primary use case
17:25:47 <catherine|2> zehicle_at_dell: +1
17:25:56 <rockyg> good point.  push sounds good to me.
17:26:10 <davidlenwell> +1
17:26:23 <davidlenwell> "private refstack to public refstack push" seems like a better title
17:26:27 <zehicle_at_dell> that would make it easier for me to review.  I'll save additional comments for the revision
17:26:40 <zehicle_at_dell> let's put "push" first now that I'm reading it back
17:26:44 <catherine|2> so we change the story?
17:26:51 <catherine|2> story title
17:26:53 <davidlenwell> yep
17:26:57 <catherine|2> +1
17:26:59 <davidlenwell> okay.. so moving on to face to face
17:27:54 <davidlenwell> zehicle_at_dell: will be in town in june for a few days.. thought we should take the time and have a f2f
17:28:16 <rockyg> yup
17:28:22 <catherine|2> +1
17:28:28 <davidlenwell> zehicle_at_dell: are your dates final ?
17:29:10 <zehicle_at_dell> I have some flex if needed
17:29:16 <zehicle_at_dell> but they are getting pretty locked
17:29:29 <davidlenwell> so I think the original plan was for us to get together at the piston office on the 11th
17:29:34 <zehicle_at_dell> yy
17:29:38 <davidlenwell> or am I off base?
17:29:47 <zehicle_at_dell> that's right.  defcore f2f is 12th
17:30:07 <zehicle_at_dell> I was planning to be at CF summit on 10th
17:30:26 <davidlenwell> excelent .. fcarpenter can you reserve us a conf room that can hold 7 people comfortably?
17:30:48 <fcarpenter> on 6/11?
17:30:51 <davidlenwell> yes
17:30:56 <fcarpenter> time of day,?
17:31:05 <davidlenwell> I'd say noon to 6pm
17:31:41 <fcarpenter> yes I can probably get panama
17:31:42 <rockyg> you'll send out new address for us?
17:31:52 <fcarpenter> 126 Post St 5th Floor San Francisco CA 94108
17:31:56 <fcarpenter> but we'll send it out, too
17:32:05 <davidlenwell> yeah .. we'll email fits with the info
17:32:12 <davidlenwell> panama would be great
17:32:14 <zehicle_at_dell> do you want an etherpad for an agenda?
17:32:20 <davidlenwell> zehicle_at_dell: yes
17:32:34 * zehicle_at_dell making pad
17:32:57 <davidlenwell> okay.. so im going to sneak in another topic here
17:33:05 <davidlenwell> #topic swagger
17:33:14 <davidlenwell> #link https://helloreverb.com/developers/swagger
17:33:28 <zehicle_at_dell> #link https://etherpad.openstack.org/p/refstack.2014-06-11
17:33:32 <davidlenwell> this is swagger.. I won't go into describing it because .. well they have done a good job of that
17:34:35 <davidlenwell> it does do client generation .. but not currently in python.. however I'm told that is in the works.
17:35:25 <rockyg> interesting....
17:35:34 <davidlenwell> yes it is
17:36:03 <davidlenwell> So I'm going to put some serious thought into if and how we can use it.. will probably update the v1 api spec to include it if I think we need to use it
17:36:36 <rockyg> if it does what it says, it is tes cool.
17:36:43 <davidlenwell> indeed
17:36:48 <davidlenwell> we shall see
17:36:52 <rockyg> tres
17:36:56 <davidlenwell> so moving along.. one more sneak in topic
17:37:06 <davidlenwell> #topic py26 gate is no more
17:37:07 <zehicle_at_dell> I've got some parking lots....
17:37:59 <davidlenwell> anyone who submitted a patch to refstack and payed attention to its testing might have noticed that I took py26 testing out of the gate.
17:38:08 <davidlenwell> as I said I would a few months ago
17:38:40 <zehicle_at_dell> I've noticed that we're faiing patches because we have no tests
17:38:43 <rockyg> how does this work with havana refstack, or will the advisory status of H refstack suffice to let us do away with it?
17:39:06 <davidlenwell> zehicle_at_dell: that is no longer true
17:39:07 <zehicle_at_dell> rockyg, yes.  that's a general hall pass
17:39:15 <zehicle_at_dell> davidlenwell, cool!
17:39:28 <davidlenwell> I posted an empty test yesterday .. if you rebase your patch it will make it through
17:39:37 <davidlenwell> I did send you a message on skype saying this
17:39:44 * zehicle_at_dell ok, will do
17:39:54 <zehicle_at_dell> yes, I saw and did rebase
17:40:03 <zehicle_at_dell> I know that praveen had the same issue
17:40:14 <davidlenwell> yes .. its because infra made a change on the 9th
17:40:26 <davidlenwell> 0 tests is no longer acceptable
17:40:34 <davidlenwell> so now we have 1 tests
17:40:36 <davidlenwell> teset
17:40:37 <davidlenwell> test
17:40:43 <davidlenwell> typing is hard
17:41:05 <rockyg> especially when low on caffeine
17:41:12 <davidlenwell> #topic open discussion
17:41:27 <zehicle_at_dell> I've got a pending change to the coretest.json file
17:41:43 <zehicle_at_dell> based on DefCore discussion, it needed some more info
17:41:52 <zehicle_at_dell> specifically, the criteria description
17:41:54 <rockyg> right.
17:42:04 <catherine|2> About the spec that only send passing data
17:42:04 <zehicle_at_dell> also needed to open it up a little so we could add scores
17:42:19 <davidlenwell> yes catherine|2
17:42:20 <davidlenwell> ?
17:42:23 <zehicle_at_dell> we need to enable people to post changes to single test criteria via gerrit for review
17:42:26 <catherine|2> davidlenwell: is that just for TCUP .
17:42:26 <zehicle_at_dell> does that make sense?
17:42:35 <davidlenwell> catherine|2: no
17:42:37 <catherine|2> I mean for data collection to determine core
17:42:45 <zehicle_at_dell> catherine|2, I think that we want the public refstack to only take pass results
17:42:49 <davidlenwell> catherine|2: that is for everything
17:43:16 <zehicle_at_dell> we may want private refstacks to also have fail tests
17:43:18 <davidlenwell> which is why it includes an option for not scrubbing the data ..so you can use it inhouse without only submitting pass.
17:43:35 <zehicle_at_dell> davidlenwell, +1
17:43:41 <rockyg> All test info needs to get scrubbed before sent to/added to refstack.org
17:43:50 <davidlenwell> I did include that in the spec.. thanks to zehicle_at_dell and rockyg feedback
17:44:17 <catherine|2> so from refstack.org point of view we only save passing dta no logs or raw data for debug ...
17:44:25 <davidlenwell> catherine|2: so you will also need to add that to your api-api sync
17:44:27 <zehicle_at_dell> rockyg, if someone does send the wrong data, then I suspect we should scrub it again just in case
17:44:28 <rockyg> but local copies not officially on refstack.org can have everything
17:44:49 <davidlenwell> yes
17:44:59 <zehicle_at_dell> perhaps we should 500 on uploads that have non-pass info
17:45:09 <zehicle_at_dell> IMHO, that data could be toxic
17:45:24 <catherine|2> so the scrub data ill be in a format that refstack define?  It is no longer a subunit defined format?
17:45:28 <rockyg> Right.  So, scrubbed in remote refstack and tcup  before upload then verified/rescrubbed on arrival at refstack.org
17:45:55 <davidlenwell> catherine|2: we should be able to keep it in subunit format
17:46:01 <rockyg> I think it's still subunit, but no tests left that fail or are skipped.
17:46:05 <zehicle_at_dell> I'm backing away from rescrubbed.  I'd think to just bounce it
17:46:16 <davidlenwell> zehicle_at_dell: I agree
17:46:25 <zehicle_at_dell> we don't want that information hanging out on public servers anywhere
17:46:25 <davidlenwell> the client submitting it will have the tools to scrub
17:46:28 <rockyg> agreed.
17:46:31 <davidlenwell> we don't want to liability
17:46:50 <catherine|2> davidlenwell: that would be a big job ... essentially we are writing a different version ofr subunit output data ...
17:46:55 <davidlenwell> last thing I want is storys of how someones cloud was comprimised only after submitting data to refstack
17:46:56 <rockyg> but it still means we need to verify it's prescrubbed.
17:47:13 <davidlenwell> rockyg: thats easy enough
17:47:20 <rockyg> cool
17:47:21 <davidlenwell> if it includes a fail .. its not scrubbed
17:47:30 <rockyg> or skipped
17:47:43 <davidlenwell> skip == fail from our perspective
17:47:44 <zehicle_at_dell> +1
17:48:12 <rockyg> can we insert a tag or some metadata in the subunit that will indicate the file has passed through our scrubber?
17:48:13 <davidlenwell> catherine|2: lets discuss that offline
17:48:20 <catherine|2> yup
17:48:40 <zehicle_at_dell> different topic > does anyone have a preference for wireframe tools?
17:48:48 <davidlenwell> balsamiq
17:48:54 <zehicle_at_dell> kk, will check that you.
17:50:12 <davidlenwell> So everyone.. re: storyboard..
17:50:33 <davidlenwell> as you are using it and you percieve an limitation .. please discuss it with them.. they are very open
17:50:42 <davidlenwell> either file a story .. or chat with them in #storyboard
17:50:51 <davidlenwell> we are an early adoptor
17:50:58 <davidlenwell> and our feedback is very valuable to them
17:50:58 <krotscheck> ping me
17:51:05 <krotscheck> VERY valuable
17:51:09 <krotscheck> I love feedback.
17:51:18 <davidlenwell> hi krotscheck
17:51:22 <krotscheck> Also, we JUST got our second +2, so things should move faster now.
17:51:31 <davidlenwell> awesome news!
17:51:45 <davidlenwell> let the patches flow!
17:51:54 <rockyg> kk.  You know me, I'm not shy;-) you'll hear frim me.
17:51:54 <krotscheck> …that sounds like a sewage metaphor.
17:52:06 <davidlenwell> krotscheck: it did didn't it
17:52:20 <krotscheck> davidlenwell: It really did.
17:52:34 <davidlenwell> I'll be careful with that in the future
17:53:00 <davidlenwell> anything else for open discussion?
17:53:12 <rockyg> I got nuthin
17:53:15 <zehicle_at_dell> we've got some backlog
17:53:18 <zehicle_at_dell> of patches
17:53:32 <davidlenwell> zehicle_at_dell: ping me in an hour or so and we'll sift through them
17:53:41 <davidlenwell> now that patches can land ..
17:53:42 <zehicle_at_dell> ok
17:53:50 <zehicle_at_dell> reasonable
17:53:59 * sarob_ lurking still
17:54:13 <zehicle_at_dell> them I'm good
17:54:15 <davidlenwell> sarob_: can you remind me of your real name?
17:54:26 <rockyg> sean roberts
17:54:32 <sarob_> Sean Roberts
17:54:36 <davidlenwell> oh thanks .. thats right
17:54:57 <davidlenwell> sarob_: did you have any more thoughts on tempest config that we didn't already talk about at the summit?
17:55:12 <sarob_> Not yet
17:55:24 <davidlenwell> okay .. well don't be shy when youd o
17:55:31 <sarob_> I've got too much plate
17:55:46 <sarob_> Or small plate
17:55:50 <davidlenwell> sarob_: understood .. thanks for paying attention
17:55:51 <sarob_> Either one
17:55:54 <rockyg> is the meetup in SV tonight?
17:55:59 <sarob_> Yup
17:56:06 <davidlenwell> meetup?
17:56:08 <rockyg> Ill be there.
17:56:11 <sarob_> Bldg e training room 8
17:56:15 <sarob_> Cool
17:56:27 <rockyg> OpenStack "hackathon"
17:56:35 <rockyg> recruiting tool
17:56:40 <sarob_> I run the sfbay Openstack user group
17:56:47 <davidlenwell> oh cool
17:57:02 <davidlenwell> is there a link to more info?
17:57:21 <rockyg> I can email you the reminder....
17:57:23 <sarob_> #link meetup.com/Openstack
17:57:31 <davidlenwell> awesome .. thanks
17:57:31 <rockyg> better
17:57:33 <sarob_> That works too
17:57:50 <davidlenwell> okay everyone.. good meeting .. thanks for showing up and participating :)
17:58:04 <rockyg> very productive today.
17:58:10 <sarob_> #link http://meetup.com/Openstack
17:58:11 <zehicle_at_dell> +1
17:58:14 <rockyg> ttfn
17:58:16 <sarob_> Thx
17:58:20 <davidlenwell> #endmeeting