22:00:07 <adrian_otto> #startmeeting Solum Team Meeting
22:00:08 <openstack> Meeting started Tue Jul  8 22:00:07 2014 UTC and is due to finish in 60 minutes.  The chair is adrian_otto. Information about MeetBot at http://wiki.debian.org/MeetBot.
22:00:09 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
22:00:11 <openstack> The meeting name has been set to 'solum_team_meeting'
22:00:19 <adrian_otto> #link https://wiki.openstack.org/wiki/Meetings/Solum Our Agenda
22:00:23 <adrian_otto> #topic Roll Call
22:00:25 <adrian_otto> Adrian Otto
22:00:29 <paulmo_> Paul Montgomery
22:00:32 <gpilz> Gilbert Pilz
22:00:35 <ravips> Ravi Sankar Penta
22:00:38 <PaulCzar_> Paul Czarkowski
22:00:46 <tomblank> Tom Blankenship
22:00:52 <peoplemerge> Dave Thomas
22:01:09 <james_li> James Li
22:01:41 <muralia> murali allada
22:01:42 <aratim> Arati Mahimane
22:02:07 <adrian_otto> welcome back gpilz (returned from vacation)
22:02:22 <paulmo_> Just in time for your turn :)
22:02:25 <datsun180b> Ed Cranford
22:02:26 <adrian_otto> :-)
22:02:29 <asalkeld> o/
22:02:29 <datsun180b> right before the buzzer
22:03:10 <adrian_otto> welcome everyone. Feel free to chime in at any time to be recorded in attendance today
22:03:17 <devkulkarni> Devdatta Kulkarni
22:03:22 <anish_karmarkar> Anish Karmarkar
22:03:22 <adrian_otto> #topic Announcements
22:03:39 <adrian_otto> Reminder: adrian_otto will be on vacation 2014-07-11 to 2014-07-24. Tom Blankenship (tomblank) has agreed to act as interim chair for 2014-07-15 and 2014-07-22 meetings.
22:03:53 <adrian_otto> I will be sending tom my tips so it should run smoothly
22:04:03 <tomblank> hopefully :)
22:04:03 <roshanagr> Roshan Agrawal
22:04:26 <adrian_otto> there is a remote chance that I may be able to attend one or two of them, but I am going to have Tom query you for agenda items
22:04:53 <adrian_otto> so I had an announcement for "Mistral Log Jam Cleared" which turned out to still be jammed
22:05:26 <adrian_otto> we have a +A patch that should unjam us: https://review.openstack.org/105484
22:05:54 <adrian_otto> long story short, we removed stackforge/solum from the list of projects that mirror requirements with OpenStack to allow us to use python-mistralclient
22:06:16 <adrian_otto> I am hopeful that at a later time we can rejoin that list, but without a job system we were left little choice
22:06:42 <adrian_otto> OpenStack *did* move to a full pypi mirror, but the requirements enforcement changes did not happen
22:06:49 <adrian_otto> so we had to opt out of that.
22:06:54 <adrian_otto> any question on this?
22:07:21 <devkulkarni> good to have all the options discussed on this point
22:07:31 <datsun180b> makes sense for now
22:07:32 <devkulkarni> thanks for working on this adrian_otto
22:07:39 <adrian_otto> we really tried every possible alternative
22:07:49 <devkulkarni> yeah, I agree
22:07:59 <datsun180b> we can still do our best to adhere to global-reqs even if it's not mechanically enforced
22:08:12 <adrian_otto> datsun180b: yes, I plan to submit a bug tocket…
22:08:15 <adrian_otto> ticket
22:08:32 <adrian_otto> that calls for an automated job to submit patches just like the openstack bot does
22:08:45 <adrian_otto> so we can stay as close as possible but still use what we need
22:09:05 <adrian_otto> so anyone interested in that, let me know and I will shoot you the bug number as soon as I open it
22:09:23 <adrian_otto> #topic Review Action Items
22:09:26 <adrian_otto> (none)
22:09:36 <adrian_otto> #topic Code Review Guidelines
22:09:44 <adrian_otto> Discussion of code reviews, and suggestions for reviewers to help keep quality and throughput up on our review queue.
22:09:51 <adrian_otto> #link https://wiki.openstack.org/wiki/Solum/Contributing#Review_Guidelines New Review Guidelines
22:10:03 <adrian_otto> many of us have already seen this and begun to apply the new guidance
22:10:24 <adrian_otto> if you ahve not seen it yet, please take a moment now to read it (it's short)
22:10:45 <adrian_otto> some of it is informative
22:11:00 <adrian_otto> and some of it is new guidance, particularly the part about how we should deal with votes and questions
22:11:25 <adrian_otto> you ahve a chance to use your "0" vote option to ask questions that should not interfere with the merge process
22:11:37 <adrian_otto> feel free to use your best judgement on when to use which option
22:11:41 <adrian_otto> discussion on this?
22:12:33 <adrian_otto> any suggestions for improving this further?
22:12:53 <adrian_otto> noorul!
22:13:05 <adrian_otto> isn't it an absurd time of day for you?
22:13:07 <asalkeld> also please don't -1 if jenkins -1
22:13:08 <devkulkarni> adrian_otto: you should contribute this to wider openstack
22:13:42 <adrian_otto> devkulkarni: good idea. Let's test it out and see how well it works for us.
22:14:05 <asalkeld> (no point in duplicating jenkins)
22:14:14 <adrian_otto> asalkeld: good point. We should probably "0" fote with pointers to why jenkins fails rather than a -1. That's a really good suggestion.
22:14:22 <noorul> watching football semi finals
22:14:25 <adrian_otto> *vote
22:14:26 <asalkeld> also no point in that
22:14:40 <asalkeld> just leave until ready to go
22:14:43 <adrian_otto> ok, I'll add that in
22:14:54 <adrian_otto> any other thoughts?
22:14:57 <asalkeld> https://review.openstack.org/#/q/status:open+NOT+label:Code-Review%253E%253D0%252Cself+label:Verified%253E%253D1%252Cjenkins+NOT+label:Code-Review%253C%253D-1+NOT+label:Workflow%253C%253D-1+is:watched,n,z
22:15:01 <asalkeld> try that ^
22:15:02 <gpilz> should we add Angus' point to the guidelines?
22:15:16 <asalkeld> to see what to review
22:15:28 <adrian_otto> wow, that's a real hum dinger
22:15:39 <adrian_otto> I thought of one additional improvement we could make
22:15:46 <datsun180b> do comments count as +0, or are they a separate mechanism from reviews altogether?
22:15:56 <adrian_otto> when we have a WIP review, use the workflow-1 flag
22:16:03 <datsun180b> because we have a comment button in addition to the review button
22:16:29 <adrian_otto> datsun180b: comments allow you to select the vote score
22:16:45 <adrian_otto> comments without votes can be added after a patch merges
22:16:52 <adrian_otto> votes can only be applied before merge
22:16:57 <datsun180b> gotcha
22:17:41 <adrian_otto> ok, good suggestions on this, thanks!
22:17:44 <PaulCzar_> I’m okay with people adding comments into why a jenkins job failed …  as long as its not just parrotting the failure ( although sometimes the reason for failure can be hard to fine, even if its in the logs )
22:18:17 <adrian_otto> I have tried to offer a link to the actual failure
22:18:26 <adrian_otto> (probably right most of the time, I hope)
22:18:44 <datsun180b> sometimes those suckers are three logfiles deep
22:18:56 <asalkeld> maybe helpfully if non-obvious
22:19:01 <adrian_otto> right, which if you have found it, could help the contributor resolve it more quickly
22:19:08 <datsun180b> like when devstack fails to set up or a freak accident means pypi.openstack is unreachable
22:19:42 <adrian_otto> #topic Blueprint/Task Review
22:19:50 <adrian_otto> #link https://launchpad.net/solum/+milestone/juno-2 Juno Development Tasks
22:19:53 <adrian_otto> Pipelines
22:20:13 <asalkeld> so going well
22:20:19 <adrian_otto> I already touched on what I think is a resolution to our mistral dependency issue
22:20:25 <asalkeld> exactly
22:20:32 <adrian_otto> there is definitely plenty of code waiting for review there
22:20:42 <adrian_otto> I'd encourage everyone to take a good look at that
22:21:14 <asalkeld> I'll keep working on it
22:21:40 <asalkeld> when it gets in more people can get stuck in
22:22:25 <asalkeld> need any more info?
22:22:31 <adrian_otto> ok, we can advanct to the next
22:22:41 <adrian_otto> unless others have questions or concerns?
22:22:52 <adrian_otto> Build Farm
22:23:02 <adrian_otto> link https://review.openstack.org/100539 Build Farm Spec
22:23:07 <adrian_otto> #link https://review.openstack.org/100539 Build Farm Spec
22:23:22 <devkulkarni> Paris might be asleep by now
22:23:26 <asalkeld> yeah
22:23:30 <adrian_otto> I saw more work on this from Julien
22:23:59 <adrian_otto> so I get the sense this is advancing nicely as well
22:24:14 <adrian_otto> Private git repo integration (ravips)
22:24:23 <adrian_otto> #link https://blueprints.launchpad.net/solum/+spec/support-private-github-repos Private Repo Blueprint
22:24:29 <adrian_otto> #link https://bugs.launchpad.net/solum/+bug/1319604 Private Repo Feature Task
22:24:33 <uvirtbot> Launchpad bug 1319604 in solum "Improvement: Add support for private GitHub repos" [Wishlist,In progress]
22:24:46 <ravips> I got good inputs from the team last week, updated BP...new WIP code: https://review.openstack.org/#/c/105605/
22:24:59 <devkulkarni> thanks ravips. will take a look
22:25:11 <adrian_otto> so ravips, last week you responded quickly to the idea of using a WIP review.
22:25:21 <adrian_otto> do you feel that was a helpful approach?
22:25:28 <ravips> adrian_otto: yeah, WIP review was very helpful
22:25:58 <adrian_otto> ok, good. I noticed a change in direction within about an hour of posting that
22:26:24 <devkulkarni> yeah. it helped that we all were discussing the WIP as soon as it got posted
22:26:25 <adrian_otto> thanks for your continued work on this feature. This will help us reach those who are using GitHub but not for opensource projects
22:27:30 <adrian_otto> Chained Trusts (julienvey, asalkeld)
22:27:38 <adrian_otto> #link https://review.openstack.org/99908 Spec for Trusts Redelegation
22:27:47 <adrian_otto> anything new on this?
22:27:51 <asalkeld> no
22:27:58 <asalkeld> we can work around it
22:28:18 <asalkeld> create an empty heat stack
22:28:26 <asalkeld> at pipeline create time
22:28:36 <asalkeld> then mistral passes the trust token about
22:28:44 <asalkeld> should all be good
22:28:57 <adrian_otto> ok
22:29:11 <adrian_otto> Mysterious Failures in Gate Tests
22:29:16 <adrian_otto> this is more of an FYI
22:29:24 <adrian_otto> #link https://review.openstack.org/100869 Example of review with apparently unrelated gate failures.
22:29:33 <adrian_otto> I wanted to flag this for input from team members
22:29:48 <adrian_otto> the gate failure does not *appear* to be related to the code in the patch
22:29:59 <adrian_otto> and it's not the first time we have seen this
22:30:06 <asalkeld> could just be a slow node
22:30:10 <adrian_otto> does anyony know what's actually going on?
22:30:16 <asalkeld> it waits for the deployer to delete the stack
22:30:34 <asalkeld> if not it doesn't delete the assembly
22:30:43 <asalkeld> and all plan deletes fail
22:30:58 <asalkeld> the tests seem to use the same plan id
22:31:07 <asalkeld> and assembly id
22:31:14 <adrian_otto> oh, that's a bug!
22:31:29 <asalkeld> well that is bad tests
22:31:35 <asalkeld> not a function bug
22:31:40 <asalkeld> functional
22:31:43 <adrian_otto> a bug in func tests
22:31:52 <asalkeld> sure
22:32:04 <adrian_otto> ok, that makes *much* more sense now
22:32:34 <adrian_otto> ok, next topic
22:32:42 <adrian_otto> #topic DIB-as-a-service
22:33:04 <adrian_otto> this came up while reviewing the custom language pack spec
22:33:07 <adrian_otto> Should we consider making a REST API and service for producing VM disk images. If so, how should it work?
22:33:18 <asalkeld> we have it already
22:33:26 <asalkeld> solum-builder-api
22:33:38 <asalkeld> it accepts dib
22:33:40 <PaulCzar_> adrian_otto: I think we can table DIB, it maybe become irrelevent after we switch to using coreos+docker for VMs
22:34:08 <adrian_otto> fair enough
22:34:32 <peoplemerge> good to know
22:34:36 <PaulCzar_> so until we want to tackle windows … DIB is probably a distraction.   I think.
22:34:52 <adrian_otto> paulmo_: good point
22:34:56 <adrian_otto> hi peoplemerge
22:35:11 <devkulkarni> asalkeld: solum-builder-api accepts dib, as in it accepts different elements via REST api?
22:35:31 <asalkeld> it sees dib as a type of app
22:35:40 <asalkeld> needs to be in a git repo
22:35:52 <peoplemerge> hi adrian_otto we were just talking about that
22:36:04 <devkulkarni> how does it handle different elements?
22:36:19 <devkulkarni> in git repo, we will have different elements?
22:36:26 <asalkeld> so you provide the element
22:36:40 <asalkeld> and depend on elements that dib provides
22:36:56 <asalkeld> not tested much
22:37:00 <devkulkarni> will look into it..
22:37:03 <devkulkarni> okay
22:37:06 <adrian_otto> A change was merged to stackforge/solum: Adding projects.txt check for devstack  https://review.openstack.org/105484
22:37:13 <asalkeld> but we were planing to use that for custom images a some point
22:37:20 <asalkeld> woot ^
22:37:29 <noorul> isnt dib hypervisor dependent?
22:37:33 <adrian_otto> whoot x 10^6
22:37:48 <devkulkarni> right. and that's what came up in the custom-lang-pack spec
22:38:00 <PaulCzar_> noorul: yes!  it seems to only really support qcow2 images … maybe raw .img
22:38:25 <asalkeld> it's the nature of the tool
22:38:34 <asalkeld> imo just support different tools
22:38:45 <asalkeld> so users can use what they want
22:38:53 <datsun180b> after projects.txt merged, next in that chain is 'Add mistral client'
22:40:34 <adrian_otto> I am ordering rechecks now
22:41:41 <noorul> shouldn't you rebase
22:41:53 <asalkeld> yeah, I'll do that
22:42:02 <adrian_otto> tx
22:42:47 <asalkeld> done
22:42:51 <adrian_otto> #topic Open DIscussion
22:42:57 <adrian_otto> I have one topic for here
22:43:11 <adrian_otto> that I thought of earlier in the meeting when we touched on weak func tests
22:43:27 <adrian_otto> what's the best way for us to wrap a timeout around a func test?
22:43:46 <adrian_otto> so if something is running slow we see that in the fail logs as "Test timed Out" or something?
22:44:36 <adrian_otto> does tempest offer anything for this?
22:44:41 <asalkeld> adrian_otto, I think make a bug and let's see what is really the problem
22:45:07 <asalkeld> I am not sure if the wait is in tests or assembly
22:45:45 <adrian_otto> ok.
22:45:52 <adrian_otto> other topics?
22:46:06 <asalkeld> https://github.com/stackforge/solum/blob/master/solum/deployer/handlers/heat.py#L97
22:46:18 <asalkeld> the wait is in deployer
22:46:27 <asalkeld> tests could wait  a bit longer
22:46:53 <devkulkarni> adrian_otto: any word on mid-cycle meetup?
22:47:06 <devkulkarni> there was some discussion about this a while back
22:47:24 <adrian_otto> devkulkarni: we will revisit that next week. Tom will lead that pursuit.
22:47:33 <devkulkarni> okay
22:47:36 <asalkeld> do we need one? are we enough on-track
22:47:55 <devkulkarni> asalkeld: I had similar thoughts
22:47:55 <adrian_otto> asalkeld: good question.
22:48:10 <adrian_otto> I'm happy to skip it if we think we can be as effective without one.
22:48:13 <asalkeld> it is quite expensive
22:48:41 <tomblank> asakleld: are you saying you don't think we really need one for this cycle?
22:48:43 <asalkeld> it's good at resolving big issues
22:48:47 <adrian_otto> for those of us who would travel to Ausin to attend, how do you feel about it?
22:48:48 <devkulkarni> one possibility is combine it with some other team (heat, mistral, nova, etc.)
22:49:12 <asalkeld> there is an infra/heat meetup in germany
22:49:17 <asalkeld> bit late for that
22:49:18 <gpilz> it's easy for people already in north america - expensive and exhausting for those who aren't
22:49:19 <asalkeld> and far
22:49:53 <PaulCzar_> skipping it could be good for those on tight budgets (money or time) to justify Paris
22:49:56 <gpilz> have we tried google meetups (or similar) in lieu of physical meetings?
22:50:32 <asalkeld> gpilz, i think TZ is out enemy not tech
22:50:44 <asalkeld> s/out/our
22:51:05 <asalkeld> but we could try more at that
22:51:09 <adrian_otto> gpilz: Yes, aslakeld is right, we are in too many timezones to let that work smoothly.
22:51:09 <noorul> also it has limit on users
22:51:35 <adrian_otto> noorul: Asusme we could use a tool so the limit were a non-issue
22:51:37 <asalkeld> we need some good thoughts on plans and infra/pipeline integration
22:51:51 <asalkeld> also bringing back the concept of services
22:52:06 <asalkeld> possibly merging services and infra
22:52:10 <devkulkarni> asalkeld: imo, we need to take a hard look at how things like CLI are going to be affected with the the new pipeline stuff
22:52:17 <noorul> i thought google is tge only choice
22:52:26 <asalkeld> skype?
22:52:38 <asalkeld> does it do groups??
22:52:52 <PaulCzar_> I think google is okay,  those of us in groups can find a way to share a connection
22:52:53 <adrian_otto> noorul: Hangouts is the easiest, but there are other things to try as well.
22:53:08 <adrian_otto> asalkeld: yes, Skype will allow groups.
22:53:10 <datsun180b> we've been banging our head against the ceiling of hangouts though
22:53:25 <adrian_otto> WebEx has it, and all the similar meeting services, like the ones from Level3.
22:53:42 <PaulCzar_> also `If you activate Google+ premium features, the limit increases to 15 participants for Hangouts created from Google Calendar events.`
22:53:44 <adrian_otto> so we can begin using one of those tools.
22:53:58 <adrian_otto> GoToMeeting is another
22:54:02 <gpilz> i can look into using our WebEx
22:54:21 <asalkeld> maybe try something crazy like non live video
22:54:27 <clarkb> you can always go the "simple" route and use pbx.openstack.org and screenshare with something like vlc
22:54:29 <gpilz> hehe
22:54:40 <adrian_otto> asalkeld: you are smoking something awesome. Pass some to me!
22:54:40 <clarkb> er vnc
22:54:54 <clarkb> though I wonder if vlc does that too. that would be neat
22:55:02 <clarkb> (it probably does)
22:55:06 <asalkeld> adrian_otto, facebook "leave a video message"?
22:55:07 <adrian_otto> clarkb: oh, neat
22:55:21 <datsun180b> it worked for Professor O'Blivion in Videodrome
22:55:28 <adrian_otto> clarkb: "Firefox can't establish a connection to the server at pbx.openstack.org."
22:55:40 <clarkb> adrian_otto: you don't connect to it with firefox :) it isn't a webserver
22:55:49 <adrian_otto> oh, a SIP server?
22:55:53 <clarkb> yes
22:56:00 <clarkb> https://wiki.openstack.org/wiki/Infrastructure/Conferencing
22:56:07 <adrian_otto> oh, my.
22:56:31 <gpilz> i thought we were talking about video
22:56:43 <clarkb> gpilz: we are, see v*c comment
22:56:52 <clarkb> I don't actually know hopw well that would work as I have never done it
22:56:56 <clarkb> but ya
22:57:22 <clarkb> http://www.videolan.org/doc/streaming-howto/en/
22:58:09 <adrian_otto> clarkb: that's really pretty cool
22:58:39 <adrian_otto> ok, we are closing in on our scheduled end time now
22:58:44 <adrian_otto> any more parting thoughts?
22:58:57 <adrian_otto> anyone else who would like to chime in to be recorded in attendance today?
22:59:38 <datsun180b> "Any last words?" "Yes, just three."
22:59:58 <devkulkarni> :)
23:00:01 <adrian_otto> thanks everyone for attending today. I'll see you back in three weeks. Tom will chair next time.
23:00:04 <adrian_otto> #endmeeting