17:01:09 <pballand> #startmeeting CongressTeamMeeting 17:01:10 <openstack> Meeting started Tue Aug 19 17:01:09 2014 UTC and is due to finish in 60 minutes. The chair is pballand. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:01:11 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:01:11 <sarob_> Morning 17:01:13 <openstack> The meeting name has been set to 'congressteammeeting' 17:01:19 <thinrichs> Hi all 17:01:53 <rajdeep> hi 17:02:06 <pballand> Items to discuss today: Alpha release progress, policy summit, open discussion 17:02:38 <pballand> thinrichs: would you mind starting with an update on the alpha release? 17:02:39 <skn_> Hi Guys, sorry couldn't make it last week 17:02:53 <pballand> hi skn_, no worries, glad you could make it today 17:03:00 <thinrichs> Sure. 17:03:13 <thinrichs> We're targeting our first release Friday. 17:03:32 <thinrichs> The goal is to have basic monitoring in place and Nova/Neutron datasource drivers. 17:03:39 <skn_> Yeah, got pulled into something for a couple of weeks, will resume Congress work from this week 17:03:43 <thinrichs> API support as well. 17:03:55 <thinrichs> That is, curl API support. 17:04:19 <thinrichs> Last Friday we were supposed to be code complete, which we mainly succeeded at. 17:04:30 <LouisF> is this congress meeting? 17:04:38 <thinrichs> This week we're testing, fixing minor things, improving documentation. 17:04:42 <thinrichs> LouisF: yep 17:04:48 <kudva> hi 17:04:51 <LouisF> hi 17:05:03 <thinrichs> Yesterday I had a completely successful API test. 17:05:17 <thinrichs> So everything seems to be working as we'd expect. 17:05:34 <pballand> awesome 17:05:47 <thinrichs> The hope is that we'll have all the small code bits merged into master by Thurs morning. 17:06:02 <kudva> thinrichs: does that mean the tests will run now? 17:06:12 <skn_> Great stuff, thinrichs! 17:06:24 <thinrichs> On Thurs, if everyone could help by testing out the install instructions and the API, that would be great! 17:06:35 <thinrichs> kudva: tests should run with tox -epy27 17:06:41 <skn_> I'd love to 17:07:03 <kudva> thinrichs: great! can't wait to get the tests working again... 17:07:04 <thinrichs> Or ./run_tests.sh but without -N so that it creates a virtual env. 17:07:14 <thinrichs> congress/README.rst should be up to date. 17:07:19 <pballand> I’ve been able to run the tests with ./run_tests.sh -N once all of the test dependencies are in 17:07:49 <kudva> pballand: okay, will try it asap 17:07:50 <thinrichs> pballand: I had an issue where once I installed with devstack then installed test dependencies, my devstack install broke. 17:08:13 <thinrichs> But I haven't tried it for a while—I understand there was a bug in infra that I kept hitting. 17:08:17 <pballand> we have a couple of ways people are running congress, so the more people we have trying it out and reporting issues/solutions the better 17:08:52 <kudva> pballand: I am running it without devstack, just clone and install congress from git. 17:09:08 <rajdeep> tests seems to be working fine 17:09:09 <thinrichs> There are just a few outstanding changes in review that need to be merged. 17:09:22 <rajdeep> i checked our code coverage also it is around 57% 17:09:40 <thinrichs> rajdeep: what do you mean by 'code coverage', and where did 57% come from? 17:09:53 <sarob_> 57.3% 17:09:54 <pballand> rajdeep: thanks for doing that - code coverage is somethign we haven’t been tracking so far, but it’s a reasonable time to start doing so 17:10:19 <pballand> rajdeep: any notable code paths missing coverage? 17:10:52 <rajdeep> plexxi code 17:10:59 <rajdeep> and the runtime 17:11:09 <rajdeep> are two areas where we could add more tests 17:11:40 <thinrichs> rajdeep: we're only exposing a fraction of the functionality of runtime in the API. 17:12:03 <pballand> rajdeep: ok, no surprises there - we’ve been working on those slowly but surely 17:12:13 <sarob_> Rajdeep: so where does the number come from? 17:12:14 <rajdeep> yeah so we need to add unit test cases there 17:12:29 <rajdeep> you can run coverage command for tox 17:12:34 <rajdeep> tox -e cover 17:12:45 <sarob_> Rajdeep: thx 17:13:03 <pballand> thinrichs: I think your doc patch went in as well, right? 17:13:22 <thinrichs> pballand: yes—but ayip is making another editing pass. 17:13:37 <rajdeep> i added a few new tests for webservice 17:13:41 <thinrichs> Apparently there are several large holes I can't see. 17:13:59 <pballand> great! it would also be great to have more eyes on the docs to help newbies as much as possible 17:14:21 <Radu_> The new doc's readme for a fresh install seems to be going well. Installing devstack on a new vm now. 17:14:46 <thinrichs> I'd love to have people try out the install on a bunch of different distros/versions. 17:15:03 <thinrichs> Maybe it's worth putting a list of the distros we've had success with somewhere in the README. 17:15:23 <sarob_> Thinrichs: good idea 17:15:28 <pballand> thinrichs: I like that idea 17:15:47 <pballand> I have a short list of instructions I follow for ubunu 12.04, so I can add those 17:16:06 <rajdeep> i checked it on ubuntu 13.10 17:16:13 <rajdeep> seesm to work fine 17:16:18 <pballand> rajdeep: thanks for adding the tests - I think they need to be reabased though 17:16:54 <rajdeep> pballand -- will do that 17:17:32 <thinrichs> I'll try to send out a reminder on Thur morning to the ML asking people to test the install/API/docs. 17:17:33 <pballand> rajdeep: actually, nevermind - I refactored the code I was thinking of in test_webservice out :-P 17:17:52 <bauzas> hi Congress team, I'm just reviewing your code but I'm having some concerns about the design you made with WSGI natively, do you plan to work in the future with Pecan or Flask ? 17:18:58 <pballand> hi bauzas - I can speak to that in a minute 17:19:02 <bauzas> pballand: sure 17:19:13 <pballand> thinrichs, rajdeep: any other updates? 17:19:24 <thinrichs> That's it for me. 17:19:58 <rajdeep> i made some changes in README but ran into rebase problems -- need to look into it 17:20:22 <rajdeep> https://review.openstack.org/#/c/114896/1/README.rst 17:20:43 <pballand> kudva: any updates on the builtins? 17:21:16 <pballand> in paralle, skn_ - are you still working on your use case? 17:21:48 <skn_> Hi pballand 17:22:07 <skn_> Yeah, I have something written up as well as some progress in the code 17:22:23 <skn_> Where do you guys put the specs? 17:22:39 <skn_> I'd like to put that in so that you can take a look 17:22:55 <kudva> pballand: yes, I added code to do syntax checking. It works. need to run the new tests and I'll check in 17:23:18 <kudva> pballand: I have finished the coding, was having problems running the tests 17:23:23 <pballand> skn_: https://wiki.openstack.org/wiki/Congress#How_To_Create_a_Blueprint 17:23:45 <pballand> skn_: we started putting the specs in the code repo, but are trying to get a spec repo done 17:23:56 <pballand> sarob: can you update us on the specs repo? 17:24:11 <sarob_> Pballand: sure 17:24:33 <pballand> kudva: great - are there any particular blockers to discuss here, or is it something to work out over email? 17:24:38 <sarob_> Pballand: the infra code is finally working 17:25:24 <sarob_> Pballand: I'm working through the req of the specs doc test 17:25:44 <sarob_> Pballand: should be done tomorrow 17:26:04 <pballand> sarob_: is it fair for people to follow the instructions on the wiki, and you will update it when the new repo is ready? 17:26:06 <sarob_> Pballand: I'm leaving for linuxcon 17:26:08 <skn_> pballand: so after we have the spec repo, that's where I should upload, right? 17:26:22 <kudva> pballand: It's going well. I will check in the syntax checking asap, and then move forward 17:26:33 <sarob_> Pballand: sure. I'll move when repo is ready 17:26:43 <pballand> kudva: great, thanks! 17:27:25 <pballand> I want to talk a bit about the policy summit 17:27:33 <pballand> bauzas: I haven’t forgot about your question 17:27:35 <sarob_> Shirley 17:27:47 <pballand> #link https://etherpad.openstack.org/p/juno-midcycle-policy-summit 17:27:53 <bauzas> pballand: no worries, it's more likely an open question about the level of integration you have 17:28:24 <pballand> we’ve been getting some good feedback and signups for the policy summit in Sept 17:28:50 <pballand> currently, the agenda is fairly open 17:29:21 <pballand> we have some ideas (there is more than enough to discuss), but want to give the community time to participate in the planning 17:29:44 <pballand> if there is something in particular people want to discuss, please addi it to the etherpad, so we can discuss 17:30:03 <sarob_> Mmcclain, mikal should have something on the agenda 17:30:15 <pballand> we’ll be calling on community leaders for each section 17:30:24 <sarob_> I'll work with them over this week to add 17:30:53 <pballand> does that make sense / are there any questions or comments about the summit? 17:31:50 <pballand> #topic Open Discussion 17:32:10 <rajdeep> what is the agenda for day 2 post lunch 17:32:14 <skn_> sarob/pballand: day 2 is for use cases? 17:33:06 <sarob_> Skn_ that's correct 17:33:11 <pballand> skn_: the high level idea was to spend day one getting familiar with various projects policy strategy and goals, and day two would be spent applying what we leared to use cases 17:33:48 <skn_> Got it! I am hoping I could do a talk on the security use cases 17:35:04 <pballand> skn_: that would be great - can you pencil that in under “proposed overviews” on the etherpad? Then we will schedule it 17:35:42 <skn_> pballand: Oh ok, great, let me add it now 17:35:58 <pballand> LouisF: I didn’t get a chane to welcome you to the meeting - are you new to congress? 17:36:37 <LouisF> pballand: yes, currently in gbp 17:37:15 <LouisF> pballand: looking forward to sept summit 17:37:39 <sarob_> Thinrichs, arosen: how's the demo? 17:37:53 <skn_> pballand: I added this to the proposed overviews, it would make sense for day 2, I guess 17:38:08 <sarob_> We have slot for next weeks ops summit 17:38:37 <thinrichs> sarob_: there are many details about the alpha release earlier in the meeting notes. But in short, it's looking good for a Friday release. 17:38:39 <pballand> sarob_: cool - I’ll be ready to give a demo - how long is the slot? 17:39:03 <sarob_> 40 min 17:39:14 <pballand> I’ve booked my travel to SA - sarob_ and I will be at the Ops Summit to talk about policy 17:39:23 <pballand> sarob_: perfect! 17:39:29 <sarob_> Pballand: good 17:39:34 <sarob_> Ness 17:40:01 <sarob_> I'm on the way to linuxcon 17:40:10 <sarob_> Before ops summit 17:40:18 <pballand> bauzas: The short answer on the API framework is that the design came from work I’ve done previously, was was done before we started pusing openstack compatibility 17:40:23 <sarob_> Mestery and other odl people there 17:40:44 <bauzas> pballand: thanks for your answer 17:41:02 <pballand> bauzas: we’re currently focused on getting the code where others can pick up congress and integrate policy across openstack projects 17:41:05 <sarob_> Anything you guys are wanting to ask these guys? 17:41:10 <bauzas> pballand: by reading your bps, I can't see where you plan to integrate with common Openstack guidelines 17:41:35 <bauzas> pballand: on the other hand, I can see you're pretty interested in being incubated one day or later 17:41:45 <pballand> bauzas: code compatability and reuse is important, but the mechanics of the API are not a focus area right now 17:42:28 <pballand> from a look-and-feel perspective, we definitely aim to follow REST best practices 17:42:39 <bauzas> pballand: sounds reasonable, but I'm just wondering when you guys plan to work on that 17:42:54 <bauzas> pballand: I can see you have specs ? 17:42:59 <pballand> the implementation can certainly change, but I don’t see it as urgent unless there are problems 17:43:13 <thinrichs> bauzas: have a pointer to the OS guidelines handy? 17:43:22 <bauzas> thinrichs: wish I would have :) 17:43:41 <bauzas> thinrichs: that's mostly coming from other projects readability 17:43:48 <pballand> bauzas: we’ve started following the specs process, but don’t yet have specs for everything 17:43:54 <pballand> (still getting the specs repo up :) ) 17:44:07 <bauzas> pballand: ok, so, meaning that we can propose without that ? 17:44:34 <thinrichs> So there's no doc describing OS guidelines/best-practices? 17:44:37 <pballand> bauzas: we have a process for proposing specs, if that’s what you’re asking 17:44:54 <pballand> bauzas: https://wiki.openstack.org/wiki/Congress#How_To_Create_a_Blueprint 17:44:58 <bauzas> thinrichs: well, each project has their own rules 17:45:06 <sarob_> Bauzas I wasn't aware pecan had become a standard 17:45:33 <bauzas> sarob_: Pecan is not a standard but a good opportunity for starting projects 17:45:41 <sarob_> Bauzas it's an interesting option for reuse though 17:46:00 <sarob_> Bauzas okay got it 17:46:03 <bauzas> sarob_: it was discussed during Icehouse and Juno summits, and I know that's something that can be considered 17:46:23 <bauzas> do you have any design docs in the wiki or elsewhere? 17:46:42 <bauzas> because I can't see if you plan to integrate with a messaging queue or so 17:47:00 <bauzas> I identified you made up the Keystone middleware auth 17:47:18 <bauzas> but IIRC, the engine is called thru a python lib ? 17:47:44 <thinrichs> bauzas: design doc (linked from wiki): https://docs.google.com/a/vmware.com/document/d/1f2xokl9Tc47aV67KEua0PBRz4jsdSDLXth7dYe-jz6Q/edit 17:47:54 <pballand> sorry, got dropped from the channel for a second.. 17:47:59 <bauzas> thinrichs: awesome, thanks 17:48:19 <sarob_> Bauzas its important to get critical feedback 17:48:36 <pballand> bauzas: from the API perspecive, we’ve focused on the consuption aspects 17:48:40 <thinrichs> bauzas: That design doc is concerned with functionality, not implementation. 17:48:43 <bauzas> pballand: ok 17:49:00 <bauzas> pballand: ok, I need to look further and see your API model 17:49:08 <pballand> the framework also is designed to keep developers from falling in to the trap of tightly-coupling the API to the application 17:49:40 <bauzas> pballand: what do you mean ? you still need input validation on the API ? 17:49:53 <pballand> bauzas: yes, there is a bp for that 17:50:02 <bauzas> pballand: k 17:50:21 <bauzas> pballand: got it 17:50:43 <bauzas> pballand: ok, thanks for your feedback, greatly appreciated 17:50:46 <pballand> bauzas: the goal is that those writing to congress don’t have to touch the REST API (they just write to a data-source interface, and their info is exposed by the framework) 17:51:13 <pballand> those writing APIs follow a simple model class, and the framework offloads validation, etc for them 17:51:20 <bauzas> pballand: mmm ok I see what you mean 17:51:37 <bauzas> pballand: ok, will see what I can do then 17:51:38 <pballand> and those working in the API framework, have to understand the details - which is where using existing frameworks has advantanges 17:52:13 <bauzas> ok, will go further 17:52:23 <pballand> the api framework is the 3rd tier of interaction (by developer time), however, so that is why it is not the top priority 17:52:54 <bauzas> k thanks 17:53:02 <pballand> with that said, we’d be (very) happy to consider blueprints and contributions that bring us closer :) 17:53:34 <bauzas> Juno-3 is closing :) 17:54:20 <thinrichs> Time check: 5 min remaining 17:54:59 <pballand> anything else people want to bring up before we close? 17:55:17 <sarob_> Opendaylight 17:55:36 <sarob_> Anything you guys want me to discuss with 17:55:41 <skn_> What about ODL? 17:55:43 <sarob_> At linuxcon 17:56:57 <pballand> sarob: details? 17:57:29 <sarob_> odl track at linuxcon 17:57:52 <sarob_> I'm going to listen mostly 17:58:06 <sarob_> Anything to discuss from this group? 17:58:37 <sarob_> Around policy 17:58:39 <skn_> ODL group policy project working with Congress, you mean 17:59:00 <sarob_> Along those lines, sure 17:59:14 <skn_> That's one of the relevant things of interest for Congress, for sure 17:59:38 <pballand> sarob_: I don’t have anythin in particular to discuss, but it would be good to listen for areas of collaboration 17:59:41 <sarob_> Ping me on the ML if you think of something 17:59:52 <sarob_> Yup will do 18:00:55 <pballand> ok, that’s it on time 18:00:57 <pballand> thanks everyone! 18:01:00 <sarob_> Thx 18:01:05 <skn_> Bye everyone 18:01:07 <pballand> #endmeeting