22:00:59 <danwent> #startmeeting
22:00:59 <openstack> Meeting started Tue Nov  1 22:00:59 2011 UTC.  The chair is danwent. Information about MeetBot at http://wiki.debian.org/MeetBot.
22:01:00 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic.
22:01:06 <somik> salv-orlando: living the good life :)
22:01:13 <danwent> #info Agenda: http://wiki.openstack.org/Network/Meetings
22:01:26 <danwent> troytoman around?
22:01:54 <troytoman> o/
22:02:01 <danwent> #topic melange status
22:02:17 <danwent> congrats to melange… going to be its own project
22:02:30 <troytoman> main update is that I conferred with vishy and a number of others and we decided melange should stand alone
22:02:40 <troytoman> so I am going to submit it for incubator status.
22:02:57 <carlp> awesome, congrats!
22:03:05 <troytoman> so, we are setting that up this week and hope to have it in front of PPB next week
22:03:11 <danwent> cool
22:03:23 <troytoman> this will simplify packaging, etc. and help us build a community around the project
22:03:23 <danwent> troytoman: what does this mean for packaging.  going to do it on your own?
22:03:26 <danwent> :)
22:03:34 <troytoman> it was getting lost in the noise with nova
22:03:55 <edgarmagana> hi everybody!
22:04:05 <danwent> troytoman:  one question.  melange is already working with keystone, is that correct?
22:04:10 <troytoman> we have also start implementing MAC address assignment in Melange
22:04:17 <troytoman> danwent: yes
22:04:20 <danwent> great
22:04:48 <troytoman> this will allow us to assure unique MAC addresses across Nova installations that use a single melange instance
22:04:53 <danwent> troy: saw that review in nova.  we probably have some suggestions on how the QuantumManager code can be structured to be more inline with the existing QM code.
22:05:00 <danwent> brad will be reviewing
22:05:11 <bhall> danwent: already provided feedback to them (and got a response)
22:05:19 <troytoman> it will also help carlp sort out how to tie in DCHP/address discovery
22:05:21 <bhall> they're going to talk it over and figure out another way to do it
22:05:22 <danwent> but overall, we should easily be able to integrate melange mac generation into QM
22:05:27 <danwent> bhall: great
22:05:51 <danwent> ok, anything else on melange?  questions?
22:06:22 <troytoman> danwent: good
22:06:22 <troytoman> I think that is it on melange
22:06:33 <danwent> #topic quantum status
22:06:47 <danwent> essex-1 is about a week away: https://launchpad.net/quantum/+milestone/essex-1
22:07:01 <danwent> want to go through any bp/bugs that aren't already in code review or done
22:07:11 <danwent> salv: https://blueprints.launchpad.net/quantum/+spec/authentication-for-quantum
22:07:20 <danwent> still comfortable with putting the keystone code back in?
22:07:35 <salv-orlando> I'm facing an issue with that at the moment
22:07:48 <salv-orlando> Basically the code does not work anymore :)
22:07:55 <danwent> an issue indeed :)
22:08:01 <danwent> due to keystone changes?
22:08:06 <danwent> or quantum changes?
22:08:09 <salv-orlando> I need to spend a few hours on it; something probably changed server-side.
22:08:29 <danwent> ok, good to know.
22:08:40 <salv-orlando> Auth is quite independent from quantum. What I'm observing is basically that I always get a 401 on token validation.
22:08:57 <danwent> please do update the status of the BP once you spend some time on it.  we can probably pull in keystone folks if needed.
22:09:06 <salv-orlando> Anyway, I hope to propose code for merge on thurs
22:09:14 <danwent> great, thanks.
22:09:19 <salv-orlando> tomorrow instead I will propose for merge api-framework-essex
22:09:31 <danwent> anticipating my next question… smart :P
22:09:44 <salv-orlando> Finally, I have a couple of bugs on my plate. They should be merge-prop by friday.
22:09:53 <danwent> great.
22:10:03 <danwent> bhall, two bugs on your plate
22:10:05 <vishy> salv-orlando: make sure you are testing based on keystone diablo
22:10:09 <danwent> that are "in progress"
22:10:16 <bhall> they're in code review
22:10:19 <vishy> salv-orlando: i wouldn't try to track keystone trunk at this point
22:10:28 <bhall> there is no launchpad bug state for "in code review" that I can find
22:10:35 <danwent> bhall: ok, please update status
22:10:39 <danwent> ah
22:10:54 <vishy> bhall: if you reference the bug or blueprint in the commit message it will create a link and a comment for you
22:11:00 <danwent> vishy:  will that remain true for a while?
22:11:13 <danwent> vishy:  my question was about keystone
22:11:26 <salv-orlando> vishy: thanks for the advie. will do that.
22:11:29 <vishy> danwent: yes, I don't think any of the projects will move to keystone trunk until the api changes have stabalized
22:11:42 <vishy> *stabilized
22:11:46 <danwent> vishy: ok, good to know
22:11:53 <vishy> stablized?
22:12:05 <danwent> :)
22:12:09 <danwent> i hate spelling
22:12:24 <bhall> vishy: noted.. I'm going to go through and amend the quantum commits that haven't merged yet with bug/bp references
22:12:40 <danwent> Ok, any other oustanding issues we have mentioned for essex-1?  we'll talk reviews next.
22:12:47 <danwent> have -> have not
22:13:17 <salv-orlando> not that I know of
22:13:47 <danwent> #info big review in quantum for essex-1 is quantum packaging: https://review.openstack.org/1094
22:14:13 <danwent> #info other big review is actually in nova, dhcp changes to quantum manager: https://review.openstack.org/#change,916
22:14:34 <bhall> I have a few more comments on the packaging review but I don't think they're major.. I'll submit the comments today
22:14:37 <danwent> we definitely need a big review push for the packaging.  I think Brad and I have committed to reviewing this week.
22:14:41 <bhall> would be nice to get some more eyes on it
22:14:45 <salv-orlando> To be honest, I won't have time for doing that review before the weekend.
22:14:55 <edgarmagana> I am also reviewing packaging and testing it
22:14:58 <danwent> there will be more smaller reviews coming down the pipeline later in the week, so I'd like to get started on the packaging one ASAP
22:15:08 <danwent> edgar: great
22:15:22 <danwent> salv: yeah, sounds like you have dev on your plate this week.  maybe you can do a final review early next week.
22:15:28 <edgarmagana> dan: I already provided some comments and more to come
22:15:29 <salv-orlando> this merge is likely to cause conflicts with all other branches.
22:15:37 <salv-orlando> which are currently under development.
22:15:54 <danwent> salv: agreed.
22:16:06 <salv-orlando> It woul dbe great if we could merge it by the end of the week. This will leave us time to fix the conflicts in branches under dev.
22:16:20 <danwent> salv: I think that's a good idea.
22:16:37 <danwent> let's target reviews tomorrow, so hopefully we can get a few rounds of feedback in by end of week.
22:16:47 <bhall> wouldn't the other way around be easier?
22:17:00 <bhall> get the other revies in first and then fix conflicts in one fell swoop?
22:17:01 <salv-orlando> bhall: push back packaging?
22:17:06 <danwent> bhall: I don't think salv's work is ready yet.
22:17:21 <bhall> I meant early next week.. but I guess we don't have much time next week
22:17:28 <bhall> ok.. either way is ok with me
22:17:54 <danwent> #help please review https://review.openstack.org/#change,1094 ASAP
22:18:26 <danwent> #info tyler + team should be ready for reviews + quick feedback this week
22:18:41 <danwent> both tyler and mark couldn't attend today, but I will sync with them via email.
22:18:54 <danwent> Ok, any other concerns around essex-1 reviews?
22:19:36 <salv-orlando> that changeset is the only one currently waiting for review for quantum, is that correct?
22:19:52 <bhall> there is more
22:19:55 <bhall> lemme find a link
22:19:56 <danwent> https://review.openstack.org/#q,status:open+project:openstack/quantum,n,z
22:20:03 <danwent> #info current reviews: https://review.openstack.org/#q,status:open+project:openstack/quantum,n,z
22:20:07 <bhall> ok, dan got it
22:20:33 <danwent> we're going to try and branch a bit before the actual milestone drop, like other openstack projects.
22:20:57 <danwent> #info we're going to try and branch a bit before the actual milestone drop, like other openstack projects.  If you're committed right before a milestone, please be aware.
22:21:11 <danwent> Ok, anything else essex-1 related, before we move on?
22:22:06 <danwent> Ok, carlp, want to give an update on jenkins + quantum?
22:22:27 <carlp> sure!
22:23:06 <carlp> I worked with Monty last week and we got the jenkins slave setup.  I've been really busy with real-life obligations but as of this morning it appears to be working.  Details are in the blueprints.
22:23:29 <danwent> carlp: can you send out links with #info tags?
22:23:46 <carlp> We have two worker nodes, both VT capable machines.  Jenkins can login to both of them.  They have 100 VLANs for use in testing
22:24:06 <carlp> #info https://jenkins.openstack.org/job/netstack-dreamhost/
22:24:19 <carlp> #info https://blueprints.launchpad.net/quantum/+spec/quantum-functional-tests
22:24:52 <danwent> ok so what is our goal in terms of getting the first functional test up and running?
22:25:06 <carlp> The biggest issue at the moment is that we need to define the actual testing environment.  We can now spin up libvirt VMs on the workers to do the testing, but we need to define that OS and package set
22:25:17 <carlp> Hardware and networking wise, we should be good to go
22:25:31 <danwent> carlp:  cool.  brad is doing some work with devscript that may be of use
22:25:44 <danwent> may save you some work.
22:25:46 <bhall> devstack
22:25:48 <bhall> he means
22:25:54 <danwent> indeed :)
22:26:08 <carlp> bhall: Here is some info monty sent me on how they do bare metal installs now :http://ci.openstack.org/jenkins.html#integration-testing
22:26:13 <carlp> we can steal from that or not
22:26:18 <bhall> ok, I'll take a look
22:26:26 <carlp> We can also just run code without a VM.  That is ready to rock now.
22:26:43 <danwent> carlp: I know several other people are looking to draft behind you in terms of setting up jenkins + quantum, so please do send out docuemtnation when you have the chance.  great work so far.
22:27:08 <carlp> It's actually pretty simple, but I will document the environment and the slave better in those blueprints
22:27:42 <danwent> cool.  please send an email to the list with any info on how we can help
22:27:57 <danwent> anything else on jenkins/functional test?
22:28:09 <carlp> I think that's it for now
22:28:15 <zns> danwent, vishy: you there?
22:28:24 <danwent> i'm here
22:29:03 <danwent> ok, and final update for quantum is that we should be targeting blueprints at essex-2 now: https://launchpad.net/quantum/+milestone/essex-2
22:29:23 <danwent> Any other questions or comments on quantum?
22:29:40 <zns> I just read you and vishy were talking about moving to Keystone and API stabalization (sp!). The API implementation in 2011.3 has been stable since Sept 30. So wondering if I understood that right?
22:30:19 <danwent> I think vishy was just telling salv-orlando that he should be using the keystone api from diablo, not tracking trunk.
22:30:22 <salv-orlando> zns: I think we were talking about Quantum's auth middleware not working at the moment.
22:30:28 <danwent> salv-orlando was seeing 401 issues with keystone
22:30:31 <salv-orlando> and danwent told the rest :)
22:30:41 <zns> ah. OK.
22:30:56 <zns> Anything you need us to do, salv-orlando?
22:31:21 <salv-orlando> zns: I spend only a few hours on it. That might be something very trivial. I'll get back at you If I'm not able to solve the issue in the next couple of days.
22:31:34 <salv-orlando> I'm doing probably too many things in parallel :)
22:31:37 <zns> OK.
22:31:45 <danwent> #topic open discussion
22:31:51 <danwent> anything else folks?
22:32:13 <danwent> #endmeeting