22:00:19 <danwent> #startmeeting
22:00:20 <openstack> Meeting started Tue Apr 10 22:00:19 2012 UTC.  The chair is danwent. Information about MeetBot at http://wiki.debian.org/MeetBot.
22:00:21 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic.
22:00:47 <carlp> o/
22:00:50 <danwent> ok, who's here?  I know a couple people are out today
22:01:08 <mestery> o/
22:01:11 <rkukura> o/
22:01:11 <danwent> #info: agenda: http://wiki.openstack.org/Network/Meetings
22:01:39 <danwent> #topic Quantum Essex Release
22:01:42 <cdub> o/
22:01:49 <danwent> #info quantum essex release is out: https://launchpad.net/quantum/essex/2012.1
22:02:12 <danwent> congrats to everyone who contributed.  Now on to Folsom as a core project :)
22:02:28 <danwent> To my knowledge, there are no major outstanding issues with the release.
22:02:52 <SumitNaiksatam> Hi All!
22:02:58 <danwent> Thanks to sumit for reviewing the Quantum Admin Guide
22:03:02 <danwent> those changes are in.
22:03:18 <danwent> I'm still in progress of creating a 1.1 version of the API guide.
22:03:48 <SumitNaiksatam> tough job :-)
22:03:49 <danwent> And we still need to consider adding distro specific content to the Quantum Admin Guide (current install instructions are just based on installing using tarball or setup.py)
22:04:02 <danwent> SumitNaiksatam: not so tough, but definitely boring :)
22:04:18 <SumitNaiksatam> :-)
22:04:21 <danwent> #topic Distro Release Testing
22:04:31 <mestery> Do we want distro specific instructions in the admin guide, or is that better to have in each distro's quantum install guide?
22:05:10 <danwent> mestery: I think nova and others include distro speciific install instructions.
22:05:11 <GheRivero> hi all
22:05:12 <annegentle> distro specific is best for later fitting into the Compute Admin guide
22:05:28 <mestery> danwent: Fair enough
22:05:40 <danwent> in fact, i'd eventually like to get away from documenting the source install, once we have reliable and well tested distro packaging
22:06:11 <mestery> danwent: That makes sense
22:06:23 <danwent> Ok, as far as fedora testing, my experience has been a very good one.  thanks to rkukura for his work on that
22:06:33 <GheRivero> real stuff/deployment should be done using distro packages
22:06:57 <mestery> GheRivero: Agree 100%
22:06:58 <danwent> I still haven't done a full deployment including nova, keystone, glance, etc. but I was easily able to get quantum-server and agents up and running.
22:07:13 <danwent> Anyone else seen any issues with fedora packagin?
22:07:16 <danwent> packaging :)
22:07:28 <mestery> danwent: I am running on fedora with devstack, lookign to do a full install on fedora later this week
22:07:37 <mestery> will report back with any issues, but so far so good
22:07:38 <danwent> mestery: thanks.
22:07:47 <rkukura> We are working on selinux policy now, so you may need to run permissive for now
22:08:20 <danwent> damn, was just going to ask him to explain.  Is this b/c we use type=ethernet?
22:08:21 <cdub> hmm, selinux police took out rkukura
22:08:25 <danwent> :)
22:08:33 <danwent> they are known to do that.
22:08:40 <cdub> danwent: no, the type=ethernet hits a different security issue (cgroup device whitelist)
22:08:58 <danwent> cdub: ok, i was familiar with the cgroup issue.
22:09:23 <danwent> #todo: rkukrua send email to list about issues with fedora packaging and selinux
22:09:31 <cdub> will be nice to get past that one, but it's low on the list
22:09:39 <danwent> Ok, on the topic of Ubuntu, we had some issues in the last version I tested.
22:10:04 <danwent> it wasn't until right before the meeting that I realized that the ubuntu folks apparently revved their packaging and seem to have fixed some of the issues with agents.
22:10:20 <danwent> So its probably worth another test of ubuntu packaging
22:11:02 <danwent> anyone who does that testing, please report to the ML.  I'll try to get to it soon.
22:11:29 <danwent> anything else on distro testing?  GheRivero, have you tried out quantum on stackops?
22:12:12 <GheRivero> this week a new intership started, he will be working on quantum on debian
22:12:33 <danwent> GheRivero: cool!
22:12:39 <GheRivero> and if everything goes properly, we (Debian, not stackops) could have a gsoc student working on quantum too
22:12:53 <danwent> double cool
22:13:03 <danwent> ok, anything else on distro testing?
22:13:22 <danwent> #topic Folsom Summit Sessions
22:13:39 <danwent> #info summit sessions posted here: http://summit.openstack.org/sessions/track/15
22:13:45 <danwent> (i hope you all can see that link)
22:14:01 <mestery> danwent: Clicking that links gets me a "Forbidden"
22:14:01 <cdub> Forbidden
22:14:02 <mnewby> nope
22:14:12 <somik> nope
22:14:17 <somik> HTTP Forbidden
22:14:17 <danwent> damn, I was afraid it was only visible to PTLs
22:14:34 <danwent> ok, check out: http://paste.openstack.org/show/12648/
22:14:36 <mnewby> http://summit.openstack.org/ works
22:14:45 <danwent> this is a first cut at the schedule for the Quantum track
22:14:46 <mnewby> got it
22:15:06 <mestery> looks like quantum is on days 2 and 3, nothing on day 1?
22:15:08 <zaitcev> yeah, we're logging in via openid from launchpad at summit site, but still 403
22:15:32 <danwent> there's currently an issue that prevents me from pushing this to sched.org, but it will be there soon.
22:16:02 <danwent> mestery: yes, I believe that is what ttx said.
22:16:12 <mestery> cool, looks good
22:16:31 <ttx> yes track review panel is tracklead-only
22:16:32 <danwent> might be nice to have some longer breaks in between sessions, but having two days solid makes the scheduling much easier, as it helps the bin-packing problem
22:16:36 <ttx> should have done a read-only view
22:16:43 * ttx <-- lazy
22:16:57 <danwent> ttx: no worries, just do it for october :)
22:17:29 <ttx> (the code is actually up on Launchpad :)
22:17:46 <danwent> ttx: ah, that's the correct response
22:18:07 <ttx> though I should really clean it up, looks quite bad :)
22:18:28 <danwent> On the topic of the summit for Quantum, I suspect our biggest challenge as a team will be converging around APIs and back-end implementation model for IPAM + L3-forwarding + NAT + floating IPs.
22:18:37 <danwent> Sumit has done a good job of sending out proposals to the list
22:18:56 <danwent> so the rest of us need to do a better job of starting to give feedback prior to the summit.  I
22:19:17 <mnewby> noted
22:19:33 <danwent> We'll have two 55 minute session devoted to this, but even that won't be enough I believe unless we start getting toward a similar model prior to the summit by using the ML.
22:19:51 <danwent> so now that Essex is out and the docs are almost done, we should really swing our attention to this.
22:20:40 <danwent> anything to add SumitNaiksatam, or other co-drivers for this topic?
22:20:58 <SumitNaiksatam> danwent: sounds good
22:21:19 <SumitNaiksatam> just for the benefit of everyone, the L3 forwarding proposal has been revised, made simpler
22:21:30 <danwent> ok.  any other questions on the summit?  We'll also be having a session on quantum + horizon, but it will be on the Horizon track.
22:21:35 <SumitNaiksatam> posted here: http://wiki.openstack.org/quantum-l3
22:21:52 <danwent> if you are interested in being a driver on any of the posted issues, please ping me, or one of the existing drivers.
22:22:28 <danwent> I will also be adding one more session on vif-plugging topics.  Thanks to rkukura for volunteering to squeeze his session into a small spot.
22:23:05 <danwent> #topic open discussion
22:23:09 <danwent> anything?
22:23:34 <danwent> looking forward to seeing you all at the summit next week.  Obviously, there will be no netstack IRC meeting next week due to the summit.
22:23:58 <mestery> Hoping to enjoy a few pints with everyone next week!
22:24:00 <edgarmagana> see you all soon!
22:24:08 <danwent> mestery: indeed :)
22:24:11 <danwent> #endmeeting