00:01:25 <ekcs> #startmeeting congressteammeeting
00:01:26 <openstack> Meeting started Thu Aug 17 00:01:25 2017 UTC and is due to finish in 60 minutes.  The chair is ekcs. Information about MeetBot at http://wiki.debian.org/MeetBot.
00:01:27 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
00:01:29 <openstack> The meeting name has been set to 'congressteammeeting'
00:02:22 <ekcs> hi all! as usual, the collaborative topics list for the meeting is here: https://etherpad.openstack.org/p/congress-meeting-topics
00:02:39 <ekcs> feel free to add/comment
00:09:50 <thinrichs> Hi all.
00:10:15 <thinrichs> (Sorry I'm late.  Got to leave early too.)
00:11:23 <ekcs> oh hi thinrichs !
00:11:39 <ekcs> ramineni on vacation this week and masahito may be traveling.
00:12:25 <thinrichs> Ok
00:12:33 <ekcs> so we can spend some time talking or just catch up next week.
00:15:34 <thinrichs> Either way.  I only have 15min or so.
00:17:07 <ekcs> ok. well. I’ve been dropping down thoughts in the PTG etherpad.
00:17:13 <ekcs> https://etherpad.openstack.org/p/congress-ptg-queens
00:17:47 <ekcs> I think in the past few cycles we had clear deficiencies and problems to work on.
00:18:36 <ekcs> at this point though, I feel we may need to consider higher level directions to inform our lower level priorities.
00:18:47 <ekcs> your input would be super helpful.
00:19:44 <ekcs> that’s one thing I wanted to bring to your attention.
00:20:31 <thinrichs> I think it'd be valuable to start targeting specific use-cases that folks have and solve them end-to-end.
00:20:45 <thinrichs> We're doing that to some extent with the policy library and remediating problems.
00:21:42 <thinrichs> Talking to Bryan from ATT and to NFV operators as you suggest could point to real pain points
00:22:43 <thinrichs> There's also the suggested use-case around configuration invariants that came up a couple weeks back.
00:23:42 <ekcs> right.
00:24:25 <thinrichs> It may be a cycle where we spend a lot of time talking to people to understand what the major pain points in OpenStack are that we can help with.
00:25:15 <thinrichs> One thing that was on my list for a long time is working to help people manage all the policy.json files.
00:25:55 <thinrichs> Security groups is another popular one
00:25:56 <ekcs> hmm good point about the possibility of focusing the cycle on that instead of features.
00:26:59 <ekcs> For me I feel like there is a bit of a bifurcation between the traditional enterprise use cases and the NFV use cases.
00:27:47 <ekcs> And I’d like to understand much better whether/where congress provides value in the NFV use cases.
00:27:54 <thinrichs> That could be.  It'd be fine to pick 1 class of use-cases and focus on that for a while.
00:28:19 <thinrichs> Probably even smart to pick 1 class of use-cases
00:28:25 <thinrichs> Or even better 1 use case
00:29:19 <ekcs> hmm makes sense.
00:32:50 <ekcs> one way to break down the classes of use cases:
00:33:06 <ekcs> 1. “human” policy kind of cases.
00:33:21 <ekcs> 2. “fault-recovery” kind of cases
00:33:42 <ekcs> I think congress was originally designed to address 1.
00:33:53 <ekcs> but there have emerged many people interested in using congress for 2.
00:34:32 <ekcs> and what use-case(s) we pick to focus on may depend on a better understanding of whether/where congress plays in 2.
00:36:29 <ekcs> so I’m studying up on things and hopefully connecting with some people to have more context for PTG.
00:37:09 <thinrichs> Talking to people who have ideas definitely sounds like a good plan.
00:37:39 <thinrichs> (I'm not sure what 1 is — someone looking at violations in a dashboard?)
00:40:00 <ekcs> that’d be one case. I’m still formulating what it means really. but seems qualitatively diferent from the idea of
00:41:11 <ekcs> writing policy as a way to react to contingencies in a self-managing system.
00:47:03 <ekcs> ending meeting now. I’m available in #congress if anyone wants to talk!
00:47:05 <ekcs> #endmeeting