21:02:46 <markmcclain> #startmeeting Networking
21:02:47 <openstack> Meeting started Mon Dec  9 21:02:46 2013 UTC and is due to finish in 60 minutes.  The chair is markmcclain. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:02:48 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:02:51 <openstack> The meeting name has been set to 'networking'
21:03:04 <markmcclain> #link https://wiki.openstack.org/wiki/Network/Meetings
21:03:09 <markmcclain> #topic Annoncements
21:03:18 <markmcclain> #info Icehouse-1 was tagged and released last week
21:03:46 <markmcclain> #link http://tarballs.openstack.org/neutron/neutron-2014.1.b1.tar.gz
21:04:01 <markmcclain> It's now time to turn our focus onto I-2
21:04:31 <markmcclain> #info Icehouse-2 release date is Jan 23rd
21:04:40 <markmcclain> #info code submission deadline is Jan 21st
21:05:22 <markmcclain> Don't let the January date sneak up on you
21:06:00 <markmcclain> with the holidays approaching for some the project velocity will slow down a bit
21:06:24 <markmcclain> so getting an early start on I-2 work is a must
21:06:50 <markmcclain> #info The first phase of 3rd Party Testing is to be completed by Jan 23rd
21:07:08 <markmcclain> We've already got several in our community working on setting this up
21:07:25 <markmcclain> if you have questions feel free to ask in our IRC channel or on the ML
21:07:34 <markmcclain> Any questions before we dive into the agenda?
21:08:14 <markmcclain> #topic Bugs
21:08:37 <markmcclain> anteaya: has gone through and updated the gate blocking bug list in the agenda
21:08:54 <markmcclain> anteaya: any we need to highlight?
21:09:26 <anteaya> mlavalle: has an update for https://bugs.launchpad.net/neutron/+bug/1251448
21:09:31 <uvirtbot> Launchpad bug 1251448 in neutron "BadRequest: Multiple possible networks found, use a Network ID to be more specific. " [High,New]
21:09:57 <mlavalle> yes…. This one is like a mutating virus….. it has several manifestations
21:10:29 <markmcclain> mlavalle: should we split out the known variants of this bug?
21:10:37 <mlavalle> it is triggered by a tempest tests that fails in it setup or teardown method
21:11:11 <mlavalle> once that happens, a network is not deleted and all the other tests after that that try to create an instance connected to a network fail
21:11:20 <markmcclain> ok
21:11:35 <markmcclain> do we know why the tearDown fails()
21:11:37 <markmcclain> ?
21:11:42 <mlavalle> right now, the test triggering the problem is a vpnaas test that has a bug
21:11:48 <markmcclain> ah
21:11:58 <markmcclain> did you loop nati_ueno in to help?
21:12:07 <mlavalle> earlier today I filed this bug https://bugs.launchpad.net/tempest/+bug/1259282
21:12:09 <uvirtbot> Launchpad bug 1259282 in tempest "Neutron VPN API test failing during VPN service creation" [Undecided,New]
21:12:27 <mlavalle> it's an easy fix…. I will submit the patch tonight or tomorrow morning
21:12:42 <markmcclain> I like easy fixes :)
21:12:42 <nati_ueno> mlavalle: let me jump in this issue
21:13:02 <mlavalle> nati_ueno: sure, go ahead
21:13:20 <markmcclain> mlavalle thanks for triaging it
21:13:26 <markmcclain> nati_ueno: thanks for working on it
21:13:30 <mlavalle> once we fix this one, we are likely to encounter another manifestation of the problem…..
21:13:46 <gongysh> https://review.openstack.org/#/c/60756/
21:14:13 <mlavalle> the specific one will depend on the fixes to race conditions that salv-oralando, armax and marun are working on
21:14:20 <mlavalle> i'll keep on eye on it
21:14:27 <mlavalle> and debug it
21:14:37 <mlavalle> i'll keep the team informed
21:14:50 <markmcclain> mlavalle: thanks for updating
21:14:51 <marun> mlavalle: thank you!
21:14:54 <rkukura> mlavalle: How regularly do you see DB lock timeouts in the logs for these failures, such as when trying to delete the port that eventually prevents the network from being deleted?
21:15:25 <mlavalle> rkukura: a week ago that was happening a lot…. not in the current mutation
21:15:40 <anteaya> if we can move to the next bug?
21:15:44 <emagana> emagana is late... sorry but I am here!
21:15:47 <mlavalle> but let's see what happens after nati_ueno submits his fix
21:15:51 <anteaya> https://bugs.launchpad.net/nova/+bug/1210483 needs a champion
21:15:53 <uvirtbot> Launchpad bug 1210483 in neutron "ServerAddressesTestXML.test_list_server_addresses FAIL" [Critical,Confirmed]
21:16:02 <anteaya> it is also a critical neutron bug
21:16:30 <anteaya> anyone willing to dig into this bug a bit?
21:17:13 <marun> anteaya: it looks like a bug we are already working on from different angles (ip assignment race)
21:17:18 <nati_ueno> mlavalle: you mean ML2 fix for port deletion?
21:17:34 <markmcclain> marun: agreeds
21:17:36 <anteaya> marun: okay, should the status of the bug reflect that?
21:17:45 <mlavalle> nati_ueno: yes
21:17:49 <markmcclain> anteaya: I'll update it
21:17:57 <markmcclain> and link to the correct other bug
21:18:00 <anteaya> awesome thanks markmcclain
21:18:10 <anteaya> so that is it from me for gate blocking bugs
21:18:14 <markmcclain> #action markmcclain to update 1210483
21:18:18 <nati_ueno> mlavalle: Thank you for your let me recall it
21:18:20 <anteaya> I didn't look at the critical bugs
21:18:23 <markmcclain> anteaya: thanks for the update
21:18:33 <markmcclain> Any other bugs the team should be tracking?
21:18:36 <nati_ueno> mlavalle: https://review.openstack.org/#/c/53478/ this one ?
21:18:40 <anteaya> I do want to say awesome job to everyone for the last few weeks adressing these
21:18:56 <anteaya> markmcclain: did you want to address the critical bugs?
21:19:03 <anteaya> I didn't look at them myself
21:19:17 <anteaya> I can for next week though, if you would like
21:19:30 <markmcclain> yeah that will work
21:19:35 <anteaya> k
21:19:42 <markmcclain> I know the current state of the work on them for now
21:19:48 <anteaya> k
21:19:54 <anteaya> should we?
21:20:14 <markmcclain> the LP states should be current
21:20:33 <markmcclain> we've still got a bit to cover
21:20:55 <markmcclain> so if anyone has questions feel free to ping the IRC channel after the meeting
21:21:03 <markmcclain> #topic Docs
21:21:07 <markmcclain> emagana: hi
21:21:22 <emagana> markmcclain: Hi There!
21:21:39 <emagana> I just updated the wiki with the status of the High priority tickets
21:21:58 <emagana> so far I am the only lazy one!
21:22:13 * markmcclain refreshes
21:22:39 <emagana> not sure if you want me go over each one?
21:22:55 <markmcclain> looks like there's an assignee for each
21:23:06 <markmcclain> so not sure we need to go into depth this week
21:23:10 <emagana> but I want to remind everyone about the doc bug day on Dec 20th!!!  Please, attend it!!  :-)
21:23:20 <annegentle> woo for doc bug day!
21:23:23 <markmcclain> #info Doc Bug Day Decemeber 20th
21:24:05 <nati_ueno> doc bug day!
21:24:09 <emagana> ok, I will be sure we make progress this week!
21:24:31 <emagana> s/we/I
21:24:48 <markmcclain> great.. thanks for updating
21:24:52 <markmcclain> #topic Nova Parity
21:24:55 <markmcclain> beagles: hi
21:25:00 <beagles> hi!
21:25:11 <beagles> so...
21:25:56 <beagles> there has been some interest in documenting mapping nova network deployments to neutron...
21:26:12 <beagles> I've created a wiki as noted in the agenda as a starting point
21:26:22 <markmcclain> awesome
21:27:44 <beagles> because so much has been going on it has been hard to track everything that is relevant to parity... but I'm pulling together everything in another wiki... also indicated on the agenda
21:28:24 <markmcclain> perfect.. I'll take a look at it in the morning
21:28:25 <beagles> while I'd love to be all rainbows... I/we have to figure out how to map where we are to where we need to get ...
21:28:36 <beagles> it's a  little chaotic at the moment
21:29:00 <markmcclain> understandable… want to sync up later this week?
21:29:07 <anteaya> beagles: will you be around in channel this week for questions?
21:29:22 <beagles> and that is pretty much what I'll be focusing on as much as I can until I'm satisfied :)
21:29:28 <beagles> anteaya, yup
21:29:33 <anteaya> great
21:29:37 <anteaya> I'll find you
21:29:53 <markmcclain> beagles: ok… feel to ask for help in the channel if you need to parallelize some tasks
21:29:59 <beagles> actually you can ping me any time... if I can't answer right away, I'll say so but I'm really great at getting distracted so you have a good chance :)
21:30:10 <beagles> markmcclain, ack
21:30:12 <markmcclain> hehe
21:30:20 <beagles> what the
21:30:24 <beagles> sorry
21:30:25 <markmcclain> Anything else?
21:30:40 <mlavalle> beagles: a few months ago I put together this ether pad, regarding gate-tempest-devstack-vm-neutron-full gate job: https://etherpad.openstack.org/p/gate-tempest-devstack-vm-quantum-full
21:31:13 <mlavalle> I think rosella_s you and I should talk to coordinate efforts
21:31:24 <beagles> mlavalle, yeah... I have the tab open to plagari... borrow from  in a bit ;)
21:31:37 <rossella_s> mlavalle: agreed
21:31:39 <beagles> mlavalle, agreed
21:31:51 <markmcclain> great
21:31:59 <mlavalle> much of what is failing in that job are tests that were designed forr nova-network that don't translate to neutron so easily
21:32:05 <markmcclain> this is good segue
21:32:08 <markmcclain> #topic Tempest
21:32:13 <beagles> heh
21:33:06 <markmcclain> mlavalle: looks like the grenade patch is moving along nicely
21:33:15 <mlavalle> so, as I said last week, I put together a wiki page for new contributors who want to help in the API tests effort
21:33:24 <markmcclain> mlavalle: awesome
21:33:25 <anteaya> the grenade patch: https://review.openstack.org/#/c/59039/
21:33:26 <mlavalle> please look at https://wiki.openstack.org/wiki/Neutron/TempestAPITests
21:33:44 <EmilienM> anteaya: almost merged
21:33:53 <mlavalle> I am asking for feedback from this team and the Tempest team as well
21:33:56 <EmilienM> got last recommendations from infra-core
21:34:13 <anteaya> EmilienM: \o/
21:34:50 <anteaya> mlavalle: great job on this page
21:35:02 <markmcclain> mlavalle: thanks for putting this together
21:35:04 <mlavalle> please feel free to make any changes you guys see fit
21:35:06 <anteaya> a good set of instructions for getting started
21:35:19 <anteaya> s/guys/folks
21:35:42 <mlavalle> I am going to let a few days pass…. after that I will send an email to the ML to start getting recruits
21:35:59 <anteaya> mlavalle: what are you waiting for?
21:36:12 <anteaya> no harm in emailing now, unless you have a specific concern
21:36:22 <markmcclain> +1 no need to wait
21:36:23 <mlavalle> just letting people give feedback
21:36:23 <anteaya> in which case voice it so we can address it
21:36:36 <mlavalle> ok, i'll send the email later today
21:36:38 <anteaya> they can do so during the ml thread discussion
21:36:44 <anteaya> mlavalle: awesome, thanks
21:37:10 <mlavalle> I also had a conversation with rosella_s earlier today
21:37:25 <mlavalle> regarding the gate job I mentioned to beagles
21:37:31 <markmcclain> ok
21:37:38 <mlavalle> trying to get everybody on the same page
21:37:53 <anteaya> mlavalle: can you link us to the conversation?
21:38:04 <markmcclain> thanks for coordinating
21:38:04 <anteaya> so we can be on the same page too
21:38:28 <markmcclain> we can chat about this after meeting because I'm concerned that we're running short on tie
21:38:30 <markmcclain> *time
21:38:36 <anteaya> k
21:38:45 <markmcclain> Anything other items tempest items?
21:39:02 <mlavalle> i'm done
21:39:25 <markmcclain> mlavalle: thanks for coordinating and updating
21:39:28 <markmcclain> #topic IPv6
21:39:30 <markmcclain> sc68cal: hi
21:39:33 <sc68cal> hello
21:40:01 <sc68cal> So, we're making some headway on the dnsmasq/dhcp side of things -
21:40:06 <markmcclain> great
21:40:09 <sc68cal> Big show stopping bug is this one
21:40:10 <sc68cal> https://bugs.launchpad.net/neutron/+bug/1257446
21:40:11 <uvirtbot> Launchpad bug 1257446 in neutron "Creating a dual-stacked network causes dhcp for both stacks to fail" [Medium,Triaged]
21:40:23 <markmcclain> error in the config file?
21:40:37 <sc68cal> Basically if you create two subnets, one v4 and one v6, dnsmasq is invoked with bad command line opts
21:41:01 <markmcclain> interesting
21:41:12 <sc68cal> For our network deployment - we are looking to do a slaac mode, so that dnsmasq is not even provisioned on the v6 side
21:41:17 <sc68cal> since we have northbound router doing RAs
21:41:33 <sc68cal> https://blueprints.launchpad.net/neutron/+spec/ipv6-provider-nets-slaac
21:41:36 <sc68cal> But
21:41:36 <markmcclain> yeah I do know that some want DHCPv6
21:41:48 <markmcclain> so we'll have to support that use case at some point in time
21:42:03 <sc68cal> That same code that triggers disabling dnsmasq, we want to make it so you can do dhcpv6 and whatever else dnsmasq supports
21:42:10 <sc68cal> make it configurable, etc.
21:42:12 <markmcclain> ok
21:42:18 <sc68cal> Next bug
21:42:26 <sc68cal> https://bugs.launchpad.net/nova/+bug/1251235
21:42:29 <uvirtbot> Launchpad bug 1251235 in nova "IPv6 DAD failure due to hairpinning" [Undecided,In progress]
21:42:30 <sc68cal> sort of related to nova-network parity
21:42:43 <beagles> yeah.. I saw that
21:42:51 <sc68cal> So, Nova turns on hairpinning on the bridge interface which breaks IPV6 dad
21:43:11 <sc68cal> but we're having some back and forth about how to fix it - I did it one way, then another, got -1's from different peolpe
21:43:23 <markmcclain> ok
21:43:24 <sc68cal> My thought is to possibly toss it onto the ML for discussion.
21:43:33 <anteaya> sc68cal: +1
21:43:38 <markmcclain> makes the most sense
21:43:48 <markmcclain> Anything other items we need to discuss this week?
21:44:02 <sc68cal> Not at this point, but feedback from -core is always appreciated
21:44:17 <sc68cal> I'm doing a lot of work that makes it "work" for me, with eye towards wider community
21:44:31 <sc68cal> but it's always nice to have some other peolpe checking
21:44:50 <markmcclain> agreed.. I'll add it to my queue for the morning
21:45:01 <markmcclain> sc68cal: thanks for updating
21:45:04 <markmcclain> #topic ML2
21:45:09 <markmcclain> rkukura/mestery: hi
21:45:17 <mestery> hi
21:45:17 <rkukura> hi
21:45:29 <rkukura> mestery: go ahead
21:45:31 <mestery> OK
21:45:58 <mestery> We're continuing to discuss ML2 in the weekly ML2 meeting, which was moved to Wednesdays at 1600UTC on #openstack-meeting-alt
21:46:12 <mestery> This week, we're going to discuss the blueprint ovs-firewall-driver in more detail.
21:46:34 <mestery> There are some folks interested in that area, and it makes sense to see how that fits into ML2.
21:46:59 <mestery> Other than that, we're also looking at this review still: https://review.openstack.org/37893
21:47:06 <mestery> This is the "Extra Port Info" for TypeDrivers.
21:47:24 <mestery> rkukura: Did I miss anything else urgent from ML2?
21:47:50 <rkukura> On the last topic, looking for feedback on letting TypeDrivers define arbitrary attributes
21:48:02 <rkukura> beyond just physical_network and segmentation_id
21:48:25 <rkukura> this forces moving from provder extension to multiprovider
21:48:32 <rkukura> so please chime in if you care
21:48:49 <markmcclain> arbitrary attributes are always tricky
21:48:49 <mestery> Thanks for bringing that one up rkukura.
21:49:07 <markmcclain> salv-orlando: have you taken a look from the API perspective?
21:49:17 <rkukura> with multiprovider, they are in list of dicts, so not an issue for API
21:49:56 <markmcclain> ok
21:50:15 <rkukura> so the question is really whether to drop the old provider extension (eventually) or keep it going for single-segment networks that don't have extra attributes
21:50:35 <rkukura> there's an email thread at http://lists.openstack.org/pipermail/openstack-dev/2013-December/021117.html
21:50:58 <rkukura> that's all for ml2 I think
21:51:25 * mestery nods in agreement.
21:51:30 <markmcclain> Thanks for updating and make sure to chime in on the thread for those that have throughts on the issue
21:51:36 <markmcclain> #topic VPN
21:51:49 <hemanthravi> last weeks IRC there was mention of a webex to discuss tempest reqmts for ML2: Is there a recording of this?
21:51:50 <markmcclain> I saw that the WIP for SSL API has been proposed
21:51:59 <nati_ueno> SSL-VPN API patch is under review.
21:52:00 <nati_ueno> yes
21:52:11 <nati_ueno> https://review.openstack.org/#/c/58897/
21:52:14 <markmcclain> hemanthravi: to my knowledge no
21:52:34 <mestery> hemanthravi: I have not set that up yet, I need to do that yet this week, FYI.
21:52:49 <hemanthravi> thx, will look for it on the ML
21:53:09 <anteaya> hemanthravi mestery we have asterisk set up for meetings, it is an opensource tool
21:53:24 <mestery> anteaya: thanks, will ping you on that.
21:53:29 * beagles snickers
21:53:32 <markmcclain> nati_ueno: besides asking for folks to review anything else to update?
21:53:49 <nati_ueno> Rajesh Mohan is working on agent implementation for that API
21:53:57 <markmcclain> great!
21:53:58 <anteaya> https://wiki.openstack.org/wiki/Infrastructure/Conferencing
21:54:08 <beagles> anteaya, used to work for Digium
21:54:18 <nati_ueno> We can reuse most of code in ipsec, so code should be small.
21:54:21 <nati_ueno> That's all from me
21:54:26 * mestery now knows to ping beagles for help with asterisk. :)
21:54:34 <beagles> heheh
21:54:35 <Swami> nati_ueno: Did we get any closure on where to store the certificates
21:54:35 <anteaya> :D
21:54:38 <markmcclain> nati_ueno: thanks for the update
21:54:41 <amotoki> nati_ueno: it looks a separate extension from IPsec VPN. Do you have a plan to merge them with service provider?
21:55:01 <nati_ueno> Swami: yes we will have application independent resoruces
21:55:21 <nati_ueno> amotoki: API should be separated because same service impl didn't support every vpns
21:55:27 <Swami> nati_ueno:thanks
21:55:34 <nati_ueno> amotoki: I'm thinking to separate IPsec and VPNService
21:55:58 <amotoki> nati_ueno: thanks.
21:56:10 <markmcclain> nati_ueno: feel free to share your idea on ML for discussion
21:56:17 <markmcclain> ok running short of time
21:56:26 <nati_ueno> markmcclain: sure
21:56:30 <markmcclain> #topic FWaaS
21:56:31 <markmcclain> https://review.openstack.org/#/c/60699
21:56:41 <markmcclain> one item up for review and SumitNaiksatam and I need to sync up
21:56:48 <markmcclain> Anything else to add this week?
21:56:53 <SumitNaiksatam> hi, yeah garyduan posted the patch
21:57:13 <SumitNaiksatam> and we markmcclain we need to get your input on the proposed blueprints
21:57:34 <SumitNaiksatam> thats it, since we are running short on time
21:57:58 <markmcclain> ok.. whens the next subteam meeting?
21:58:08 <SumitNaiksatam> wednesday
21:58:16 <SumitNaiksatam> 10 AM PST
21:58:42 <markmcclain> ok.. I'll be on a plane… I'll try to sync up with up tomorrow
21:58:42 <anteaya> SumitNaiksatam: if you can, try to use utc for time
21:58:48 <anteaya> it is easier for the group
21:58:48 <gongysh> SumitNaiksatam: what is the utc time?
21:58:55 <salv-orlando> Hi guys I am really late - anything you needed from me
21:59:02 <markmcclain> 1800UTC
21:59:12 <SumitNaiksatam> markmcclain: yeah its mentioned on the meetings page
21:59:52 <markmcclain> salv-orlando: nothing new added to your queue
22:00:03 <gongysh> 1800UTC is too early for Chinese guys.
22:00:08 <markmcclain> #topic Open Discussion
22:00:14 <salv-orlando> markmcclain: yeee christmas came early
22:00:33 <anteaya> gongysh: what time range is better for your location?
22:00:49 <gongysh> we are at UTC+8
22:00:55 <marun> gongysh: you'll have to fight with europe, they get screwed too
22:01:01 <anteaya> gongysh: so what time range is better?
22:01:07 <marun> 5am for cst right?
22:01:35 <anteaya> when wanting another option, it is helpful to offer an alternative
22:01:40 * salv-orlando love the assumption that weird hours are not suitable for us
22:01:43 <salv-orlando> us = US
22:01:46 <marun> anteaya: I don't think there is a good solution, honestly.  When I was in UTC+2, midnight meetings
22:01:56 <anteaya> there is no good solution
22:01:57 <marun> anteaya: When I'm in JST, 6am
22:02:07 <gongysh> I think UTC21 +
22:02:08 <nati_ueno> I have question about 3rd party CI.
22:02:12 <marun> anteaya: the only people that have it easy are in north america
22:02:21 <nati_ueno> To support Router and Security group extension is mandatory?
22:02:27 <anteaya> but it is helpful to have information to consider
22:02:44 <anteaya> and I agree, having flexibility for meeting times helps all contributors
22:02:47 <gongysh> better 22:00,  this is the current time.
22:02:49 <salv-orlando> We should move to the moon. As it's tidally locked, time zones would count nothing there.
22:02:51 <markmcclain> nati_ueno: no those are not required to be supported by a plugin
22:02:56 <marun> salv-orlando: well, you're an outlier.  people in israel aren't fans of midnight
22:03:02 <nati_ueno> markmcclain: OK Thanks
22:03:21 <markmcclain> marun: speaking of which.. it is midnight
22:03:23 <emagana> nati_ueno: Thanks for asking, I had that concern as well.
22:03:27 <anteaya> ha ha ha
22:03:28 <markmcclain> and we're out of time this week
22:03:29 <salv-orlando> marun: I'm not saying I like it; it's right in the middle of the only productive hours I have!
22:03:36 * salv-orlando s/only/most/
22:03:41 <nati_ueno> emagana: your welcome :)
22:03:42 <markmcclain> Thanks everyone for stopping by and talk to everyone on the ML and IRC
22:03:47 <markmcclain> #endmeeting