16:03:45 <rkukura> #startmeeting networking_ml2
16:03:45 <openstack> Meeting started Wed Mar  5 16:03:45 2014 UTC and is due to finish in 60 minutes.  The chair is rkukura. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:03:46 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:03:49 <openstack> The meeting name has been set to 'networking_ml2'
16:04:09 <rkukura> #link https://wiki.openstack.org/wiki/Meetings/ML2 Agenda
16:05:00 <rkukura> icehouse feature freeze was yesterday, but there are a few straggling merges and some FFEs for ML2
16:05:15 <rkukura> plus some time to fix bugs
16:05:29 <rkukura> and work on docs
16:06:38 <rkukura> #topic Action Item Review
16:06:53 <rkukura> there were no official action items recorded last week
16:07:00 <rkukura> are there any that didn't get recorded?
16:07:05 <trinaths> Hi rkukura
16:07:23 <mestery> rkukura: I don't think so, no.
16:07:49 <trinaths> Hi mestery
16:08:03 <rkukura> So for today, lets focus on merges that potentially get into icehouse
16:08:10 <rkukura> #topic Mechanism driver status for icehouse
16:08:58 <HenryG> Hi
16:09:02 <shivharis> rkukura: brocade md is in 'about to be' merged state. thanks for the last minute help, appreciate it
16:09:08 <rkukura> see https://wiki.openstack.org/wiki/Meetings/ML2#Meeting_March_5.2C_2014 for the ones I'm aware of
16:09:19 <trinaths> I have completed code and submitted for review and 3rd party setup is ready today..
16:09:27 <shivharis> mestery: thanks for the reviews.
16:09:29 <trinaths> but its defered by mark
16:09:45 <rkukura> here, approved means the patch has been reviewed and approved, but is either waiting to get merged, or maybe needs rebasing
16:10:08 <mestery> trinaths: You will have to ask markmcclain for a FFE, it's out of our hands now.
16:10:27 <shivharis> rkukura: mark initiated the merge
16:10:30 <trinaths> FFE ??
16:10:44 <rkukura> FFE is feature freeze extension
16:10:56 <rkukura> so far, ODL has an official FFE
16:11:18 <rkukura> mestery: Is that on track to finish review and have 3rd party testing this week?
16:11:41 <mestery> rkukura: I hope so, dealing with some issues in the third party testing.
16:11:51 <trinaths> oh.. can you guide me on how can I achieve this .. is it via email.. ?? can I mail mark with a request to FFE
16:12:08 <amotoki> trinaths: when did you propose the code?
16:12:32 <trinaths> Its a week back
16:12:41 <trinaths> but mark hase some review comments
16:12:55 <trinaths> and I have updated then and posted back review
16:12:57 <rkukura> trinaths: Can you paste the link to the review?
16:13:03 <amotoki> i think it is one of important factors to consider FFE. it seems to be late for feature proposal review deadline....
16:14:03 <trinaths> rkukura: here it is https://review.openstack.org/#/c/78092/
16:14:44 <rkukura> trinaths: Were earlier versions reviewed separately?
16:14:56 <trinaths> yes
16:15:32 <trinaths> old versions
16:15:35 <trinaths> https://review.openstack.org/#/c/71868/
16:15:46 <trinaths> https://review.openstack.org/#/c/73985/
16:15:58 <trinaths> https://review.openstack.org/#/c/70089/
16:16:13 <trinaths> I was struck with Ci setup
16:16:24 <trinaths> now I have it running in my testsetup
16:16:29 <rkukura> trinaths: I'll add these links to the agenda page. I think you'll need to negotiate with markmcclain to see if an FFE can be granted, and recruit reviewers
16:17:06 <HenryG> The cisco_apic mech driver missed 3rd party testing deadline yesterday, but we will get it working today and ask for FFE.
16:17:14 <rkukura> It usually helps to keep the same Change-Id for updates so the history is obvious
16:17:25 <trinaths> rkukura: thankyou for the help.. Do I need to mail mark or post in openstackdev list regarding FFE
16:17:26 <amotoki> why do you post a new review instead of restoring the review? It allows us to track the review history.
16:17:46 <rkukura> trinaths: I'd try both if needed
16:18:23 <trinaths> okay.. since, I dont have an email of mark.. can you share..!
16:18:27 <rkukura> HenryG: I put deferred? for the apic status, knowing you were negotiating an FFE
16:18:32 <trinaths> will post it now
16:18:49 <HenryG> rkukura: thanks. The cisco CI has to test many products and it takes a lot of manual labor to keep it running smoothly.
16:19:17 <rkukura> trinaths: Mark's email should be in any of the recent gerrit reviews he's done
16:20:11 <rkukura> Other MDs listed as deferred are cisco_ucs and huawei. Are we looking for FFEs for these, or are they going to target juno?
16:20:22 <trinaths> yes.. got it.. will post a mail for FFE.. now
16:20:41 <sadasu> for cisco_ucs I am working on 3rd party testing setup
16:20:53 * pcm_ mmcclain@yahoo-inc.com
16:21:02 <sadasu> working through some issues...
16:21:05 <rkukura> Its important that any of these FFEs get reviews from ML2 team members for them to have any chance to make icehouse
16:21:26 <sadasu> once I have that ready, looking for FFE too
16:21:43 <asomya> https://review.openstack.org/#/c/73355/
16:21:53 <asomya> https://review.openstack.org/#/c/73372/
16:22:00 <asomya> These are the two patches for the APIC MD
16:22:27 <rkukura> Ok, we'll update the agenda with links and "FFE?" for anything hoping to get an FFE.
16:22:49 <rkukura> Anything else on mechanism drivers for icehouse? Did I miss any?
16:23:20 <irenab> rkukura: when the code should be ready to get FFE?
16:23:42 <rkukura> to get an FFE for a blueprint, it really should already be in review
16:23:49 <irenab> there is the following MD for SRIOV https://review.openstack.org/#/c/74464/, but still in WIP
16:24:15 <yamamoto_> s/ryu/ofagent/ in the agenda please
16:24:31 <rkukura> irenab: You'll need to discuss an FFE with markmclain
16:24:39 <rkukura> I'll add the link to the agenda
16:24:54 <irenab> rkukura: Thanks
16:24:56 <amotoki> yamamoto_: done
16:25:09 <rkukura> yamamoto_: ok
16:25:11 <yamamoto_> thanks
16:25:21 <rkukura> anything else on MDs?
16:25:42 <rkukura> #topic Other BPs for icehouse
16:26:07 <rkukura> marun: Any update on the migration BP?
16:26:23 <rkukura> This one has an FFE and is in review
16:28:12 <rkukura> I guess marun isn't really here - lets give https://review.openstack.org/#/c/76533/ some review attention - its critical to finish deprecating the old monolithic plugins
16:29:30 <rkukura> Is Cedric (zzelle) here?
16:30:35 <rkukura> Guess not. His partial specs BP implementation looks ready to merge to me, but seems to have missed the deadline.
16:31:05 <rkukura> Put the wrong link in the agenda for the 1st part
16:31:24 <rkukura> #link https://review.openstack.org/#/c/71904/
16:32:16 <rkukura> If others think this is important, please review it. This is a possible FFE in my opinion.
16:32:58 <rkukura> Are there any other ML2 BPs that I missed in the agenda that might make icehouse?
16:34:10 <rkukura> ok, moving on...
16:34:28 <rkukura> #topic Bugs for icehouse
16:35:21 <rkukura> Of the two I'm working on that effect the MD API, the 1st was approved but the verify failed
16:35:37 <rkukura> I'll fix it if needed, so I think that should get merged soon
16:36:09 <rkukura> The 2nd, which moves port binding outside the transactions is in progress and should be in review by the end of the week.
16:37:33 <rkukura> What other ML2 bug fixes are being worked on for icehouse?
16:37:54 <rcurran> i've got a few for cisco_nexus md
16:38:10 <banix> I have this: Not critical but in case you want to have a look: https://review.openstack.org/#/c/69792/
16:39:25 <banix> Mark did review it a few days back; updated to address his suggestions.
16:39:30 <trinaths> rkukura: can you guide me on recruit reviewers..
16:39:36 <rkukura> I'd suggest adding those links to the agenda to encourage reviews and so we can track them next week
16:39:56 <shivharis> rkukura, banix: do we expect a precommit-rollbacks?
16:39:58 <trinaths> i feel yourself and mestery can be my code reviewers .. any suggestions
16:40:16 <shivharis> in plugins?
16:40:32 <banix> shivharis, : I do not think so
16:40:32 <rkukura> trinaths: I'm certainly willing to review it.
16:40:56 <trinaths> rkukura: thanks a lot sir.. :)
16:41:03 <shivharis> banix: i;ll look at your patch, have not seen it yet
16:41:18 <rkukura> shivharis: Doesn't throwing an exception in *-precommit already result in rollback?
16:41:18 <trinaths> Hi Mark
16:41:25 <banix> shivharis: Thanks.
16:42:36 <shivharis> rkukura: if the postcommit fails, there needs to be explicit rollback, i look at the patch and comment, i dont know how it is implemented. later.
16:43:18 <rkukura> shivharis: Is that addressed by banix's patch (which I haven't reviewed yet)?
16:43:49 <shivharis> rkukura: i have not seen the implementation, will take a look today
16:44:00 <rkukura> shivharis: Same here.
16:44:09 <rkukura> Any other bugs to discuss?
16:44:47 <rkukura> I see DVR has been added to the agenda
16:45:00 <carl_baldwin> Hi
16:45:02 <rkukura> #topic Distributed Virtual Router
16:45:13 <carl_baldwin> Should be quick.
16:45:44 <carl_baldwin> I just wanted to point out that DVR design includes some work in L2.  A document describing the work is available for review.
16:46:03 <carl_baldwin> #link https://docs.google.com/document/d/1depasJSnGZPOnRLxEC_PYsVLcGVFXZLqP52RFTe21BE/edit#heading=h.5w7clq272tji
16:46:06 <rkukura> carl_baldwin: Is this likely to make icehouse, or is it for juno?
16:46:34 <carl_baldwin> The target is currently Juno-1.  Not Icehouse.
16:46:58 <carl_baldwin> DVR is a use case under the L3 sub team.
16:47:01 <carl_baldwin> #link https://wiki.openstack.org/wiki/Meetings/Neutron-L3-Subteam
16:47:16 <carl_baldwin> Is there anyone interested/willing to review?
16:48:10 <amotoki> i am interested in it.
16:48:11 <rkukura> carl_baldwin: Is the idea to package this as an MD in ML2, or does the plugin itself need to change?
16:48:41 <rkukura> Seems like the L2 population MD might be a good model
16:48:42 <carl_baldwin> rkukura: A good question.  I have only begun to review the document myself.
16:49:16 <carl_baldwin> That would be a great question to add as a comment to the document if it is not clear.
16:49:25 <rkukura> carl_baldwin: Thanks for bringing this up. I encourage everyone to review and comment on the doc.
16:49:26 <Sukhdev> carl_baldwin: I want to look into this
16:49:43 <carl_baldwin> Great, thank you.  That is all I had for this meeting.
16:49:48 <rkukura> Anything else on DVR today?
16:49:53 <carl_baldwin> Sukhdev: Thank you for your interest.
16:49:55 <carl_baldwin> Nothing else.
16:50:07 <rkukura> #topic Open Discussion
16:50:15 <Sukhdev> carl_baldwin: Does it provide hooks for physical devices?
16:50:50 <carl_baldwin> Sukhdev: I don't think that has been addressed yet.  Another great question to ask in the document.
16:51:57 <rkukura> Anything else before we wrap up?
16:52:12 <Sukhdev> rkukura: QQ
16:52:47 <rkukura> ?
16:52:48 <Sukhdev> rkukura: ML2 agent work - that replaces ovs agent is all in for I3?
16:54:14 <rkukura> Sukhdev: I'm not aware of anything on that for icehouse
16:54:35 <amotoki> /FYI/ I updated the links in the agenda based on the discussion in the meeting.
16:54:49 <Sukhdev> rkukura: OK - somebody brought up yesterday - was curious
16:54:49 <rkukura> amotoki: Thanks - really appreciate it!
16:54:59 <shivharis> Sukhdev: your comment was not understood, can u please rephrase
16:55:16 <rkukura> Sukhdev: Is this the "modular agent" that was discussed a bit at the last summit?
16:55:27 <trinaths> I have sent a mail to openstack-dev and mark regarding FFe.. hope for the best
16:55:39 <Sukhdev> shivharis: Oh sorry - rkukura, yes that is what I meant
16:56:04 <rkukura> Sukhdev: Lets see if there is interest in this at the juno summit
16:56:12 <rkukura> Anything else?
16:56:28 <rkukura> We've got 4 minutes, but I don't mind finishing early!
16:56:31 <Sukhdev> rkukura: sounds good
16:56:44 <Sukhdev> early finish is always good :-)
16:56:53 <banix> bye :)
16:57:04 <irenab> thanks
16:57:04 <yamamoto_> good night
16:57:05 <rkukura> Thanks everyone!
16:57:10 <sadasu> thanks
16:57:12 <Sukhdev> bye
16:57:12 <trinaths> thanks rkukura
16:57:19 <shivharis> bye all
16:57:22 <rkukura> Keep up the reviews and finish up the coding!
16:57:29 <rkukura> #endmeeting