21:10:12 <devkulkarni> #startmeeting Solum Team Meeting
21:10:13 <openstack> Meeting started Tue Aug 25 21:10:12 2015 UTC and is due to finish in 60 minutes.  The chair is devkulkarni. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:10:14 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:10:18 <openstack> The meeting name has been set to 'solum_team_meeting'
21:10:22 <devkulkarni> #topic Roll Call
21:10:27 <devkulkarni> Devdatta Kulkarni
21:10:30 <dimtruck> o/
21:10:31 <muralia> murali allada
21:10:40 <devkulkarni> hi dimtruck, muralia
21:10:43 <dimtruck> hi devkulkarni
21:10:45 <muralia> hey
21:10:54 <devkulkarni> hi james_li_
21:11:00 <devkulkarni> hi datsun180b
21:11:06 <james_li_> james li
21:11:15 <devkulkarni> #link https://wiki.openstack.org/wiki/Meetings/Solum#Agenda_for_2015-08-25_2100_UTC
21:11:21 <devkulkarni> today's agenda ^^
21:11:47 <devkulkarni> nice to see you all
21:12:11 <devkulkarni> the agenda is mostly about reviews
21:12:36 <devkulkarni> #topic Announcements
21:12:49 <devkulkarni> I don't have any prepared announcements
21:13:02 <devkulkarni> anybody want to share something with the team?
21:13:29 <muralia> i see lots of patches from you
21:13:40 <devkulkarni> muralia: yes
21:13:41 <datsun180b> oh i'm here, just distracted
21:13:47 <devkulkarni> I am about to get to that
21:13:53 <devkulkarni> okay, since no announcements from anyone, moving on to next topic
21:14:02 <devkulkarni> #topic Review Action Items
21:14:29 <devkulkarni> I there were no action items from the last meeting
21:14:42 <devkulkarni> #topic Blueprint/Task Review and Discussion
21:14:56 <devkulkarni> ok, so this is the major topic for todays meeting
21:15:04 <devkulkarni> I have three items:
21:15:13 <devkulkarni> (devkulkarni/randallburt) Deployer plugins (Notes: Need one more +2)
21:15:20 <devkulkarni> I just saw james_li_ approve this
21:15:25 <james_li_> yes
21:15:32 <devkulkarni> thanks for that james_li_
21:15:44 <james_li_> it has been there for a while
21:15:54 <devkulkarni> james_li_: yes, that is correct
21:16:11 <devkulkarni> now that this is merged, we can start implementing the plugin model for deployers
21:16:22 <devkulkarni> I am really looking forward to that
21:16:30 <devkulkarni> because, once that is in place
21:16:32 <dimtruck> +1
21:16:46 <devkulkarni> it will be straightforward (hopefully) to integrate with magnum
21:16:53 <dimtruck> +10000 :D
21:16:56 <devkulkarni> :)
21:17:14 <devkulkarni> I will be checking with randallburt on what are the next steps
21:17:24 <devkulkarni> and what needs to be done to start the implementation
21:17:45 <james_li_> should operator have an option of not running solum with magnum?
21:17:54 <devkulkarni> dimtruck, muralia: your knowledge with magnum should be useful when we get to implementing the integration
21:18:00 <muralia> yup
21:18:04 <devkulkarni> james_li_: good question
21:18:23 <devkulkarni> I would assume that we would provide operator-level flags/overrides to choose what configuration solum uses
21:18:31 <muralia> james_li: yes, operators can choose any backed to deploy apps
21:18:40 <dimtruck> yup!
21:18:41 <devkulkarni> whether it is heat+magnum or just heat
21:19:42 <devkulkarni> more will emerge about pros/cons of this on that once we start implementing and experimenting I suppose
21:20:09 <devkulkarni> So next one in the review section is:
21:20:11 <devkulkarni> (devkulkarni) Request to review API patches. (Notes: These patches add the workflow resource and connect it to the Solum worker (and deployer))
21:20:19 <devkulkarni> let me give some background on this..
21:20:43 <devkulkarni> so as you all know we have accepted a spec to add 'app' and 'workflow' resource to solum
21:21:05 <devkulkarni> the four patches that I have listed in meeting agenda are implementing the workflow resource
21:21:20 <devkulkarni> before you look into the code
21:21:28 <devkulkarni> it might help to check out this spec:
21:21:50 <devkulkarni> #link https://review.openstack.org/#/c/215832/
21:22:13 <devkulkarni> this outlines the approach that I have taken to connect the workflow resource to the main solum engine (specifically, worker and deployer)
21:22:30 <devkulkarni> the four code patches follow the approach outlined in the spec
21:22:40 <devkulkarni> I have not yet added tests
21:22:50 <devkulkarni> I will be doing that starting tomorrow
21:23:14 <devkulkarni> please take sometime, whenever you get a chance, to review these patches
21:23:34 <devkulkarni> imo, once these patches merge, we will have basic workflow resource implemented (end-to-end)
21:23:35 <james_li_> devkulkarni: i guess you did some tests manually
21:23:42 <devkulkarni> james_li_: yes —
21:23:49 <devkulkarni> I have tested manually
21:23:55 <james_li_> with CLI
21:23:57 <james_li_> ?
21:23:57 <devkulkarni> can show you if interested
21:24:03 <devkulkarni> yes, with CLI
21:24:07 <james_li_> ok
21:24:10 <james_li_> cool
21:24:26 <devkulkarni> I am hoping that we can get these patches merged soon
21:24:47 <devkulkarni> so that before the long weekend (next week), we could demo the end-to-end behavior with the new code
21:25:08 <devkulkarni> so looking for reviews.
21:25:15 <devkulkarni> datsun180b: as an fyi
21:25:29 <devkulkarni> you might see that I have added 'status' to the app model
21:25:51 <devkulkarni> it is a singular status rather than a dictionary as you have outlined in your spec
21:25:52 <datsun180b> yeah, that part's tricky to convey
21:25:59 <devkulkarni> I have done that on purpose
21:26:11 <datsun180b> older ideas, older models. i won't fuss about changes like that
21:26:12 <devkulkarni> my main goal for now is to get to feature parity with what we currently have
21:26:19 <devkulkarni> cool
21:26:30 <devkulkarni> just wanted to inform you about the diversion from the spec
21:26:39 <datsun180b> rules are made to be broken
21:26:56 <devkulkarni> once this first iteration is complete, I will go back and add finer states etc.
21:27:06 <datsun180b> that sounds entirely reasonable
21:27:11 <devkulkarni> awesome!
21:27:54 <devkulkarni> so the next item is the CLI patches:
21:28:01 <devkulkarni> (devkulkarni) Request to review CLI patches. (Notes: These patches add interactive input functionality to the new app commands.)
21:28:16 <devkulkarni> let me give some background on these set of patches
21:29:10 <devkulkarni> basically we want to achieve two things: 1) support the new app resource 2) add all the error checks, interactive inputs, etc. that we currently support to the new app paths
21:29:51 <devkulkarni> code that invokes the app and workflow rest APIs was already submitted as WIP by datsun180b
21:30:08 <datsun180b> as long as feature parity is the goal i guess we can keep the interactive prompts
21:30:08 <devkulkarni> that has been the starting point for item 2
21:30:16 <datsun180b> but ftr i don't like 'em
21:30:26 <devkulkarni> datsun180b: I would like to get to feature parity first
21:30:46 <devkulkarni> that way, if anyone wants to use it, the experience will be what we currently support
21:31:02 <datsun180b> right, so i'm not going to fight any of those changes
21:31:17 <devkulkarni> datsun180b: thanks :)
21:31:33 <devkulkarni> so starting from your WIP patch
21:31:52 <devkulkarni> I am adding piece-by-piece support for other things
21:31:59 <devkulkarni> the patches should be easy to review
21:32:17 <devkulkarni> the code is almost similar to existing code
21:32:33 <devkulkarni> with simplifactions/changes to support the app-file model instead of the plan-file
21:32:58 <devkulkarni> btw
21:33:28 <devkulkarni> in both, API and CLI, I am planning to support git-triggers after the first iteration
21:33:45 <datsun180b> aiming high
21:33:47 <devkulkarni> i.e., once the current workflow patches on the API side and the CLI patches are merged
21:34:11 <devkulkarni> datsun180b: actually, not aiming for git-trigger stuff at all for now
21:35:07 <devkulkarni> will start looking at porting the git-trigger related things (trusts in API, tokens in CLI) after the basic end-to-end workflow patches are completed
21:35:37 <devkulkarni> datsun180b: does sound like a good plan to you?
21:35:47 <devkulkarni> s/does/does that/
21:36:10 <datsun180b> seems like the most straightforward
21:36:15 <devkulkarni> datsun180b: cool
21:36:17 <muralia> ya
21:36:33 <devkulkarni> so overall the next steps/plan in my mind is as follows:
21:36:48 <devkulkarni> 1) complete basic end-to-end workflow (in progress)
21:37:01 <devkulkarni> 2) complete basic CLI changes (in progress)
21:37:15 <devkulkarni> 3) git-trigger on API
21:37:23 <devkulkarni> 4) tokens/github support on CLI
21:37:31 <devkulkarni> 5) pluggable deployers
21:37:39 <devkulkarni> 6) magnum integration
21:37:52 <devkulkarni> 6) non-destructive app update functionality
21:38:05 <devkulkarni> 7) support for multi-repo applications
21:38:33 <devkulkarni> the last item may need overall changes
21:39:01 <devkulkarni> thoughts/comments?
21:39:29 <devkulkarni> oh btw, one more thing
21:39:40 <devkulkarni> sometime in between these
21:39:50 <devkulkarni> I would like to get james_li_'s bash-to-python patches merged
21:40:04 <devkulkarni> so james_li_, I will take a look at the config generator issue
21:40:13 <devkulkarni> but not right now
21:40:15 <james_li> thanks
21:40:35 <devkulkarni> it will have to wait till I get at least till item 4 completed
21:40:57 <devkulkarni> cool
21:41:17 <devkulkarni> so moving on to the next topic
21:41:24 <devkulkarni> #topic Open Discussion
21:41:38 <datsun180b> i've got a short one
21:41:43 <devkulkarni> datsun180b: floor is yours
21:41:51 <datsun180b> devkulkarni in your workflow table you're adding a deleted_at field to workflow
21:41:59 <datsun180b> who will delete workflows?
21:42:35 <devkulkarni> datsun180b: in our current model, app delete deletes all the assemblies
21:42:50 <devkulkarni> so it could be similar..
21:42:55 <devkulkarni> however,
21:43:29 <devkulkarni> for preserving the history of workflow executions
21:43:55 <devkulkarni> we are not going to delete the record from the workflow table. hence you had the 'deleted' boolean flag
21:44:09 <devkulkarni> so then, the 'deleted_at' field is to record the time when this flag was set
21:44:22 <datsun180b> i don't remember a deleted boolean
21:44:31 <datsun180b> but in that case, carry on
21:44:32 <devkulkarni> it is there in the spec as well
21:44:52 <datsun180b> perhaps it was soft deleted in my memory
21:44:56 <devkulkarni> datsun180b: I remember we discussing about soft delete
21:45:00 <devkulkarni> right
21:45:32 <datsun180b> then i'm happy with your current procedure
21:45:40 <devkulkarni> datsun180b: cool
21:46:01 <datsun180b> now there's a point where your code in these patches outweighs what i set out with in my WIPs
21:46:36 <datsun180b> so how many grains until we have a pile here?
21:46:44 <devkulkarni> yes — there are some additions which build upon the WIP
21:47:14 <devkulkarni> datsun180b: I don't anticipate more patches (on the API side)
21:47:26 <devkulkarni> the four patches cover the basic end-to-end functionality
21:47:41 <devkulkarni> I will be adding tests as part of these patches themselves
21:47:48 <datsun180b> still looking at the latest revisions but they look like they're all there
21:48:50 <devkulkarni> datsun180b: saw your comments on CLI patches
21:48:58 <devkulkarni> thanks for checking those out
21:49:36 <datsun180b> i don't have any substantial quarrel with the CLI changes
21:49:45 <devkulkarni> thanks
21:50:10 <devkulkarni> I can move filter_lps back to solum.py
21:50:26 <devkulkarni> the reason I moved it to cli_utils was because it is required from both AppCommands and OldAppCommands
21:50:37 <devkulkarni> I understand that this is temporary
21:50:48 <devkulkarni> so here is what I can do
21:51:03 <devkulkarni> move it back to solum.py, but not make it part of any class
21:51:29 <datsun180b> those sort of functions are getting numerous as is
21:52:02 <devkulkarni> true
21:52:45 <devkulkarni> we can revisit this once we are at feature parity with existing functionality
21:52:50 <datsun180b> that's fine then
21:52:56 <datsun180b> let me go double that +1 up
21:53:18 <devkulkarni> datsun180b: oh, cool
21:53:25 <devkulkarni> that is nice of you datsun180b :)
21:54:12 <devkulkarni> we are almost at the end of our meeting time
21:54:30 <devkulkarni> anything else for us to discuss today?
21:54:37 <james_li> none from me
21:54:54 <datsun180b> nothing further
21:55:05 <devkulkarni> want to convey sincere thanks to all for keeping looped in on solum reviews
21:55:24 <devkulkarni> james_li, datsun180b, dimtruck, muralia thanks for joining in today
21:55:30 <devkulkarni> see you all next week
21:55:35 <muralia> thanks
21:55:41 <devkulkarni> #endmeeting