16:00:14 <adrian_otto> #startmeeting Solum Team Meeting
16:00:15 <openstack> Meeting started Tue Mar  4 16:00:14 2014 UTC and is due to finish in 60 minutes.  The chair is adrian_otto. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:16 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:19 <openstack> The meeting name has been set to 'solum_team_meeting'
16:00:27 <adrian_otto> https://wiki.openstack.org/wiki/Meetings/Solum Our Agenda for today
16:00:41 <adrian_otto> #topic Roll Call
16:00:52 <devkulkarni> Devdatta Kulkarni
16:00:52 <adrian_otto> hello everyone!
16:00:53 <muralia> murali
16:00:53 <paulmo_> Paul Montgomery
16:00:59 <adrian_otto> Adrian Otto
16:00:59 <noorul> Noorul Islam K M
16:01:01 <paulczar> Paul Czarkowski
16:01:06 <ycombina_> Shaunak Kashyap
16:01:12 <aratim> Arati Mahimane
16:01:38 <alexheneveld> hello
16:01:42 <julienvey_> Julien Vey
16:01:42 <adrian_otto> Alex, welcome!
16:01:55 <julienvey_> won't be able to stay long
16:02:10 <stannie> Pierre Padrixe
16:02:12 <tomblank> tom blankenship
16:02:30 <coolsvap1> Swapnil
16:02:38 <funzo> Chris Alfonso
16:02:51 <adrian_otto> julienvey_: thanks for the heads up. How long will you be with us today?
16:02:53 <alexheneveld> good to be back - and great to be seeing so much activity
16:03:07 <julienvey_> adrian_otto half an hour
16:03:31 <adrian_otto> ok, in that case, I am planning to shuffle the agenda so you can provide input on the Incubation Discussion
16:03:52 <adrian_otto> anyone feel strongly to the contrary?
16:04:12 <datsun180b> argh, missed my cue
16:04:16 <gokrokve> Hi
16:04:17 <datsun180b> Ed Cranford
16:04:23 <adrian_otto> we are still in Roll call, datsun180b
16:04:35 <gokrokve> Georgy Okrokvertskhov
16:04:56 <adrian_otto> ok, so we will do Incubation discussion just before Review Action Items.
16:05:29 <adrian_otto> thanks everyone for attending. Let's proceed to announcements. Feel free to chime in at any time to record your attendance if you have not already.
16:05:38 <adrian_otto> #topic Announcements
16:05:45 <adrian_otto> Solum Summit Registration - Make your travel arrangements
16:06:00 <adrian_otto> Tuesday, March 25, 2014 - Wednesday, March 26, 2014
16:06:07 <adrian_otto> #link https://wiki.openstack.org/wiki/Solum/Summit Find Link to registration here
16:06:29 <adrian_otto> this is a very important event for us. I might argue it's the most important Summit for this project period.
16:06:41 <adrian_otto> so if you are on the fence for whether to attend in person, please attend.
16:07:04 <adrian_otto> next announcement: Solum Demo Session at Atlanta ODS Event
16:07:21 <adrian_otto> please take a moment to vote for our demo session: Vote at: https://www.openstack.org/vote-atlanta/Presentation/introducing-solum
16:07:43 <adrian_otto> instead of submitting a whole mess of sessions, we just want everyone to focus on this, and make it awesome
16:08:07 <adrian_otto> all your votes will make a huge difference
16:08:17 <muralia> Voting was closed on March 3rd. Thats what the website says.
16:08:31 <adrian_otto> oh, well.
16:08:55 <adrian_otto> if you happen to know a member of the committee who reviews these, possibly a mention is justified.
16:09:06 <adrian_otto> as voting is only part of the criteria for selection
16:09:11 <adrian_otto> next announcement:
16:09:13 <adrian_otto> CAMP 15 day Public Review is open. Ends 2014-03-17. Please make comments based on what we have learned in Solum.
16:09:27 <adrian_otto> #link https://www.oasis-open.org/news/announcements/15-day-public-review-for-cloud-application-management-for-platforms-camp-v1-1-end
16:09:53 <adrian_otto> this is a PaaS standard that may still be shaped by what we do in this project.
16:10:05 <adrian_otto> if you care about what lands in the standard, please review and comment
16:10:27 <rajdeep> are we trying to get paas vendors to get involved in the standard?
16:10:51 <adrian_otto> rajdeep: I think all of us should take an interest, and know what's proposed
16:11:25 <adrian_otto> the more consensus reached early, the better it will be
16:11:36 <rajdeep> ok standard will become successful if established players adopt it
16:11:39 <adrian_otto> its' worth an hour or two to look and jot down remarks
16:11:42 <alexheneveld> re CAMP asalkeld (for whom I guess this time slot is not very friendly!) and I had a chat
16:12:01 <alexheneveld> one spot of confusion was whether it should be used for the pipeline (build triggers etc) ... i tend to think *not*
16:12:20 <muralia> +1
16:12:21 <alexheneveld> it is designed to describe and manage application deployments (assemblies)
16:12:28 <adrian_otto> alex, yes. I think we should touch on that again during open discussion today
16:12:42 <alexheneveld> however it could be interesting to give an abstracted api on top of things like build services
16:13:03 <alexheneveld> adrian_otto: cool
16:13:09 <adrian_otto> ok, next topic
16:13:19 <adrian_otto> #topic Incubation Discussion
16:13:21 <adrian_otto> Are we interested in changing our wiki content to adjust language referring to us as "OpenStack Related" to state an intent to file for OpenStack incubation upon achievement of milestones agreed to by the team members?
16:13:58 <adrian_otto> I'm seeking a perspective from as many of you as possible on this question
16:14:15 <funzo> I'm not sure of what the cons are from doing this
16:14:30 <julienvey_> I agreed with that statement, but do we discuss the milestone today ?
16:14:39 <julienvey_> s/agreed/agree
16:14:52 <noorul> We need clarity on m1
16:14:54 <adrian_otto> we can discuss the milestone whenever it makes sense
16:15:26 <adrian_otto> perhaps it's m1, or maybe something different. The question really hinges on whether we want to state an intent to pursue an incubation period in our future.
16:15:40 <adrian_otto> and if that's desired by our contributors, then we should be clear about that intent
16:16:04 <noorul> I think it was our intent from the beginning to apply for incubation
16:16:05 <devkulkarni> the only con I see is, once an intent is stated and if for some reason that does not get followed through, what does that mean?
16:16:09 <julienvey_> it's desired for me :)
16:16:13 <stannie> +1 noorul
16:16:33 <adrian_otto> devkulkarni: if we change our minds, that's fine.
16:16:41 <datsun180b> i don't think there's anything wrong with stating intent and reevaluating later if needs or circumstances change
16:16:49 <adrian_otto> but based on the input I have so far, I think we do have that intent
16:16:50 <datsun180b> that's what intent is
16:17:04 <gokrokve> +1
16:17:14 <devkulkarni> in that case, there is no con as such imo.
16:17:18 <alexheneveld> +1
16:17:25 <gokrokve> The main part will be a clear mission statement to make a proper posiotioning in the community
16:17:28 <tomblank> +1
16:17:29 <devkulkarni> +1 to changing the wiki
16:17:34 <muralia> +1
16:17:34 <stannie> +1
16:17:55 <adrian_otto> gokrokve: exactly, I will begin to offer some draft statements for us to discuss.
16:17:56 <tomblank> gokrokve: yes, i think that is important
16:18:07 <adrian_otto> that's a prerequisite for application
16:18:34 <adrian_otto> ok, this appears to be unanimous. Are there any alternative viewpoints to consider?
16:19:11 <adrian_otto> welcome back rajdeep_
16:19:25 <paulczar> +1
16:19:39 <rajdeep_> Currently are we in the pre incubation phase?
16:19:42 <aratim> +1
16:19:43 <adrian_otto> ok, hearing no objections to the suggestion, I will move to a #agreed on this point,adn we can take this a step further
16:20:18 <adrian_otto> #agreed Solum will adjust it's public facing documentation to state a clear intent to pursue an incubation period for the project.
16:20:35 <adrian_otto> #action adrian_otto to adjust the wiki to show intent for incubation
16:20:54 <adrian_otto> rajdeep_: yes, we are in pre-incubation
16:21:05 <rajdeep_> ah ok
16:21:18 <adrian_otto> so the next logical question is when would be the best time to process an application for incubation?
16:21:40 <adrian_otto> based on the prevailing criteria, once we have a mission statement, we qualify.
16:22:03 <julienvey_> yes
16:22:27 <adrian_otto> one possibility is to pursue this on an immediate basis
16:22:36 <datsun180b> i don't know if there's such thing as too early in this case
16:22:49 <adrian_otto> so that Juno would potentially be our first full cycle in inbubation
16:23:09 <adrian_otto> datsun180b: exactly. That's what incubation is intended for.
16:23:29 <muralia> Yes, we should do it right away. We'll probably attract more contributors that way.
16:23:47 <adrian_otto> after a full cycle in incubation, and doing train based release milestones, we may be at a point suitable for integrated status.
16:24:23 <julienvey_> is there a deadline for incubated projects to get integrated ?
16:24:43 <adrian_otto> julienvey_: good question. You can ally any time.
16:25:04 <adrian_otto> but you must go through a full cycle before you can graduate out of incubated into integrated status
16:25:06 <devkulkarni> I would like to hear gokrokve's views on how distracting it is for the team while the discussion is happening about incubation? what are your experiences with murano?
16:26:02 <gokrokve> devkulkarni: In our case it was pretty clam. There are some discussions with another incubation request for Climate project.
16:26:38 <gokrokve> The main concerns were mission statement and needs of such program\functionality
16:26:50 <gokrokve> The team should be prepared to answer that.
16:26:58 <gokrokve> There are not much technical questions.
16:27:18 <devkulkarni> gokrokve: I see.
16:27:23 <adrian_otto> I expect the technical criteria is more about exiting incubation rather than entering.
16:27:42 <adrian_otto> and we have followed all the guidelines that are normally requested.
16:27:51 <adrian_otto> so I don't expect major rework
16:28:12 <gokrokve> adrian_otto: Sure. There are some requirements for incubation, but as I see they already satisfied in Solum (devstack-gate, requirements check)
16:28:16 <adrian_otto> as our first core croup were almost all in the OpenStack Technical Committee
16:28:48 <adrian_otto> ok, so I am happy to take an action item to offer a mission statement
16:28:50 <rajdeep_> /NICK rajdeep
16:28:57 <gokrokve> Yes. We need to explain why this project should be a part of OpenStack.
16:29:01 <adrian_otto> that we can etherpad on to refine
16:29:15 <adrian_otto> so let
16:29:21 <gokrokve> It will be hard as there are many PaaS solutions which work with OpenStack.
16:29:41 <adrian_otto> us assume for sake of discussion that we are at the point of having a mission we all believe in, and all the rationale needed for an applicaiton
16:29:43 <stannie> we will have to explain how Solum works differently
16:30:08 <adrian_otto> would anyone object to filing right away?
16:30:16 <gokrokve> With the recent CloudFoundry movement to open-source community it will be even harder.
16:30:32 <devkulkarni> gokrokve: Yes. I guess in a way this approach will make us crystallize our thinking around the differences and needs very clearly.
16:30:58 <adrian_otto> gokrokve: I am confident this project is different enough that we can make it clear why these are in fact apples and oranges
16:31:41 <julienvey_> adrian_otto: +1
16:31:46 <noorul> Isn't important, which milestone will bring that difference?
16:31:47 <gokrokve> Sure. We need to prepare couple stron use cases which show where Solum adds value.
16:32:03 <noorul> s/Isn't/Isn't that
16:32:57 <adrian_otto> noorul: that's a good question. Because we are taking an iterative approach, some of the differences will be less obvious in the beginning, but the key differences are in the orientation with the IaaS of OpenStack
16:33:05 <gokrokve> noorul: I think this is important. We need to show that there are benefits right now.
16:33:36 <noorul> That is what I also felt but I think m1 has deviation from earlier plan.
16:33:37 <gokrokve> noorul: I think nobodt will listen stories about benefits in 5 year horizon.
16:34:20 <adrian_otto> we know from our discussions with potential end users, that there are a wide variety of needs and use cases. Solum appeals to some of those, and other existing systems appeal to others.
16:35:08 <adrian_otto> practically all cars have four wheels. Some are SUV's and others are two door coupes. Why do we need both types? Because the consumers have differing interests.
16:35:09 <paulczar> adrian_otto: I agree you should draft up a mission statement as you have a good clear vision about what solum is and why it's important to openstack and different to <insert other thing here>
16:35:22 <adrian_otto> ok, I willt ake that as an action
16:35:50 <adrian_otto> #action adrian_otto to draft mission statement and solicit input form interested contributors
16:36:18 <adrian_otto> if you would like to help, just let me know and I will be certain to include you (even if you are reading this as a transcript)
16:36:45 <julienvey_> I have to go, thanks adrian_otto for bringing this up early, interesting discussion
16:37:07 <adrian_otto> julienvey_: thanks! See you next time!
16:37:11 <alexheneveld> being integrated with heat and openstack is a big differentiator
16:37:27 <adrian_otto> ok, any more thoughts to consider on the incubation subject?
16:37:44 <alexheneveld> cf has a nice approach based on bosh creating "stem cell nodes" but the "build on stem cells" approach won't fit everything.  having standards of heat and openstack underneath allows nice alternatives.
16:38:18 <adrian_otto> alexheneveld: +1
16:38:47 <adrian_otto> ok, let's proceed to action items
16:38:56 <adrian_otto> #topic Review Action Items
16:39:06 <adrian_otto> adrian_otto to support muralia for review of https://etherpad.openstack.org/p/MinimalCLI (for tasks specific to blueprints)
16:39:24 <adrian_otto> muralia: Should we carry this forward, or are we happy with this now?
16:39:24 <muralia> I've looked through the etherpad. https://etherpad.openstack.org/p/MinimalCLI
16:39:34 <muralia> it's good for now
16:39:46 <noorul> I have questions
16:39:48 <muralia> i had to make a few changes to the app create command
16:39:56 <muralia> sure
16:40:00 <adrian_otto> noorul: please proceed
16:40:08 <noorul> 1. Are we considering plan file for m1?
16:40:31 <devkulkarni> noorul: I can answer that
16:40:40 <muralia> I'm workign with Angus to decide this. most probably not, or atleast, just a basic one with just a repo url
16:41:02 <devkulkarni> asalkeld took first stab at implementing it. he found some issues which I think he is working out with alex.
16:41:31 <adrian_otto> noorul: Once concern we are working on is how best to express the difference between an application configuration, the related workflow(s), and the deployment architecture.
16:41:35 <devkulkarni> In the mean time I think asalkeld is proceeding with a very simple plan file (which just contains git repo url), or an option of no plan file.
16:41:54 <adrian_otto> piling that all into one file made sense conceptually, but in practice turned out to be a bit messy
16:42:15 <alexheneveld> + the project config file and the application deployment plan should be separate
16:42:19 <alexheneveld> +1 ^^^
16:42:25 <adrian_otto> so we are interested in ideas for how to address this
16:42:52 <alexheneveld> in particular you may have multiple deployment plans, for different environments etc
16:42:57 <adrian_otto> one possibility is what alexheneveld is suggesting which is to use separate files for the config and deployment
16:43:26 <adrian_otto> we could still have a single plan file that makes references to other files as needed
16:43:57 <alexheneveld> and project config (is there a better name for this?) focuses on per-project stuff, like how to build it, test it, what to promote when where etc
16:44:07 <paulczar> with m1 functionality being very limited,  do we really need a plan file ?   or I guess more,  can we just have a default plan file and have the CLI/API auto select it ?
16:44:09 <alexheneveld> +1 adrian_otto project config would refer to individual deployment plans
16:44:13 <devkulkarni> alexheneveld: is there a etherpad that you and asalkeld are discussing this in?
16:44:15 <adrian_otto> noorul: the short answer to your question is yes, there will be a plan file. The uncertainty at this point is exactly what belongs in there, and what should be broken out into supporting files
16:44:17 <noorul> Is this decision factored in this etherpad ?
16:44:33 <noorul> I think we should focus on m1
16:44:58 <alexheneveld> devkulkarni: so far i don't think so ... i caught asalkeld quite late his time and it was a quick chat but i expect he will write it up
16:45:01 <noorul> For m1 do we need plan file?
16:45:09 <muralia> For M1, we might just want to keep it simple. one option is to just have one build/deployment workflow in solum for M1 and just pass in a plan file with a git repo url in it. or even ignore the plan file and just pass in the git repo url to the assembly create CLI call.
16:45:09 <adrian_otto> so one way to continue making forward progress on the m1 target is to defer some of the design effort, and focus on the simple case first.
16:45:21 <muralia> I think the best would be to not use a plan file and keep it simple.
16:45:41 <noorul> When I spoke to Angus y'day, I requested him to send an email to ML with planless idea
16:45:54 <alexheneveld> +1 quite possibly we don't need deployment plan file for m1, there is a default way an app always gets deployed
16:46:03 <noorul> I think he will be doing it may be by tomorrow
16:46:12 <adrian_otto> that would essentially amount to setting the repo URL as an environment variable, or an argument to the CLI
16:46:39 <adrian_otto> as that's the minimum piece if information needed for the simple workflow
16:46:46 <adrian_otto> s/if/of/
16:47:02 <muralia> adrian_otto: it could be sent in as an argument to the 'assembly create' command
16:47:13 <devkulkarni> +1 to argument to CLI
16:47:19 <adrian_otto> ok, we are getting a bit crunched on time for today
16:47:20 <alexheneveld> (the plan file would really be for setting different deployment configurations, as i see it)
16:47:36 <adrian_otto> I do see this decision as very important
16:47:46 <adrian_otto> would we like to continue now, or revisit this?
16:47:50 <muralia> we could continue discussion on this etherpad. https://etherpad.openstack.org/p/MinimalCLI
16:48:14 <adrian_otto> muralia: okay, thanks. We can also schedule a breakout meeting for this if there is sufficient interest
16:48:15 <devkulkarni> I would not like to clutter the cli etherpad with discussion about plan
16:48:28 <muralia> sure
16:48:29 <adrian_otto> then link to another etherpad in that one
16:48:38 <noorul> devkulkarni: I think this is very important
16:48:53 <noorul> devkulkarni: We need to narrow down the scope for m1.
16:49:11 <devkulkarni> noorul: yes, I am not disagreeing. there was another etherpad (something called solum-demystified)
16:49:35 <muralia> https://etherpad.openstack.org/p/solum-demystified
16:49:39 <devkulkarni> I think plan was being discussed there. so just suggesting that we do plan brainstorming there. but no hard preference. anything that works
16:49:55 <adrian_otto> so let's ask it… who feels strongly that we should have a comprehensive plan file as part of m1, versus a streamlined workflow for the general case, and a simplified approach to accompany it?
16:50:28 <adrian_otto> I am partial to the simple approach, which will allow us freedom to iterate on it
16:50:31 <muralia> +1 for simple approach
16:50:36 <devkulkarni> +1 to streamlined workflow for general case
16:50:40 <stannie> simple and then iterate
16:50:41 <stannie> +1
16:50:53 <tomblank> +1 simple approach and then iterate on it
16:50:54 <datsun180b> +1 for simple
16:50:57 <rajdeep_> i think we should focus on getting app deployed without too many complications
16:51:13 <adrian_otto> ok, good
16:51:17 <alexheneveld> +1
16:51:19 <adrian_otto> alternate viewpoints to consider?
16:51:52 <adrian_otto> ok , I will move to a #agreed
16:52:40 <adrian_otto> #agreed to use a simplified plan file approach for m1, so that it is suitable for a general use case, and allows additional features to be added in subsequent iterations.
16:52:46 <adrian_otto> does that sound about right?
16:53:05 <devkulkarni> so we are still saying we will have a plan file — just making sure..
16:53:14 <devkulkarni> but a simplified one
16:53:20 <adrian_otto> devkulkarni: it may be implicit
16:53:26 <adrian_otto> there will be a plan resource regardless
16:53:33 <muralia> no, we shouldnt. I'm guess that will be hashed out in the email that angus will send out soon
16:53:37 <adrian_otto> but we might actually bypass the file itself by using the CLI
16:54:04 <devkulkarni> sure, okay.
16:54:37 <adrian_otto> ok, since we have limited time remaining, I'd like owners of blueprints to place any updates in the whiteboards on those respective BP's.
16:54:48 <adrian_otto> if you would like us all to hear your update, you are welcome to make it now.
16:54:53 <devkulkarni> adrian_otto: okay, will do.
16:55:12 <devkulkarni> updates on specify-lang-pack. aratim's work got merged!!
16:55:23 <adrian_otto> whoot
16:55:33 * paulmo_ cheers Arati
16:55:37 <adrian_otto> #topic Open Discussion
16:55:40 <devkulkarni> stannie is planning to work on the rest of the actions (PUT/POST, etc.) from that BP
16:56:14 <adrian_otto> excellent
16:57:00 <devkulkarni> adrian_otto: gokrovke, asalkeld, and I were chatting about similarities and differences between murano.
16:57:08 <adrian_otto> I will be proposing additional core reviewer candidate(s) over the next week
16:57:16 <adrian_otto> if you have input, I'm open to hearing from you
16:57:18 <devkulkarni> have you had a chance to sync up with gokrovke about this?
16:57:38 <adrian_otto> devkulkarni: I had a chat with gokrokve, yes.
16:57:41 <stannie> devkulkarni: I am currently working on functional tests for components so Julien Vey started working on PUT/POST for LP
16:57:48 <devkulkarni> how does murano's workflow DSL line up with what we want to express?
16:58:03 <devkulkarni> stannie: sounds good!! you guys rock :)
16:58:13 <stannie> since he needs PUT/POST etc for LP in the functionnal test he started on LP :)
16:58:21 <noorul> Yes!
16:58:34 <adrian_otto> devkulkarni: they are in the same ballbpark. I think there is a good opportunity for us to work together
16:59:16 <adrian_otto> I'm convinced that neither project wants unnecessary overlap, so it makes sense to coordinate efforts to the extent convenient
16:59:20 <noorul> we were discussing thins the othe rday
16:59:29 <gokrokve> adrian_otto: Agree. There is a lot of synergy between the projects.
17:00:03 <devkulkarni> gokrove, adrian_otto: sounds good
17:00:11 <adrian_otto> gokrokve: yes. It might be wise for us to co-author a Wiki page that discusses the similarities
17:00:20 <adrian_otto> thanks everyone
17:00:23 <gokrokve> adrian_otto: Sure.
17:00:25 <adrian_otto> #endmeeting