15:01:30 <mlavalle> #startmeeting neutron_l3
15:01:33 <openstack> Meeting started Thu Oct 15 15:01:30 2015 UTC and is due to finish in 60 minutes.  The chair is mlavalle. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:01:34 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:01:37 <openstack> The meeting name has been set to 'neutron_l3'
15:01:44 <mlavalle> Good morning everyone
15:01:50 <regXboi> I guess I was wrong :)
15:02:20 <mlavalle> agenda for today is: https://wiki.openstack.org/wiki/Meetings/Neutron-L3-Subteam#Agenda
15:02:59 <mlavalle> ooops! I just realized I forgot to update the date.... but the agenda itself is up to dat
15:03:01 <john-davidge> hi
15:03:23 <mlavalle> #topic Announcements
15:03:37 <adduarte> hello
15:03:50 <mlavalle> well, the obvious anoucements is that we are 10 days away from the Tokyo summit!
15:04:02 <carl_baldwin> #excited
15:04:17 <mlavalle> looking forward to see you as many as possible of you there!
15:04:41 * regXboi will be missing out ... #sadpanda
15:04:42 * john-davidge is rewatching Lost In Translation multiple times
15:05:05 <mlavalle> the other announcement I think is relevant is to start paying attention to mitaka-1. I just pinged armax and he informs it is december 1 o 2
15:05:46 <mlavalle> any other announcements from the team?
15:06:14 <carl_baldwin> It’ll be here before we know it.
15:06:38 <mlavalle> ok, moving on......
15:06:49 <mlavalle> #topic Bugs
15:07:06 <mlavalle> so first up is https://bugs.launchpad.net/neutron/+bug/1365473
15:07:06 <openstack> Launchpad bug 1365473 in neutron "Unable to create a router that's both HA and distributed" [High,In progress] - Assigned to John Schwarz (jschwarz)
15:07:50 <mlavalle> adduarte: merged one partial fix a couple of weeks ago
15:08:00 <mlavalle> thank you very much!
15:08:11 <adduarte> the merge still outstanding is the server side
15:08:31 <adduarte> the agent can create the dvr ha router
15:08:42 <mlavalle> and jschwarz is working on https://review.openstack.org/#/c/143169/
15:08:44 <adduarte> but the server side won't allow it
15:08:48 * carl_baldwin will review today.
15:08:54 <mlavalle> I think this is what adduarte is referring to
15:08:59 <carl_baldwin> Needs rebase due to merge conflict.
15:09:06 <regXboi> ack
15:09:12 <adduarte> yes I think John has done quite a lot of work on it.
15:09:14 <neiljerram> hi all, sorry to be late (again)
15:10:19 <mlavalle> yeah, we need to rebase it and take from there
15:10:30 <mlavalle> any more comments?
15:10:48 <mlavalle> ok....
15:11:00 <mlavalle> next up is  https://bugs.launchpad.net/neutron/+bug/149435
15:11:00 <openstack> Launchpad bug 149435 in update-manager (Ubuntu) "Update Manager freezes when clicking "install updates" after having cancelled root password insertion" [Undecided,Invalid]
15:11:08 <regXboi> typo?
15:11:13 <mlavalle> nope wrong bug
15:11:44 <regXboi> https://bugs.launchpad.net/neutron/+bug/1494351
15:11:44 <openstack> Launchpad bug 1494351 in neutron "Observed StaleDataError in gate-neutron-dsvm-api tests if reference IPAM driver is used" [High,In progress] - Assigned to Carl Baldwin (carl-baldwin)
15:11:55 <regXboi> how about that one?
15:11:58 * carl_baldwin getting deja-vu
15:12:07 <mlavalle> https://bugs.launchpad.net/neutron/+bug/1494351
15:12:20 <mlavalle> this is the correct one
15:12:50 <carl_baldwin> pavel_bondar should be the assignee.
15:13:18 <mlavalle> he pointed out that he is going to be on vacation until the 20th
15:13:35 <mlavalle> and wanted someone to take over for that period
15:13:47 <carl_baldwin> mlavalle: True.  I’ll revisit my review and see how it is looking.
15:14:11 <carl_baldwin> regXboi: Let’s chat out of band on this one.
15:14:16 <regXboi> carl_baldwin: ack
15:14:25 <mlavalle> I think there was a rebasing problem carl_baldwin was helping with
15:14:46 <carl_baldwin> mlavalle: That was an infra issue and is resolved.
15:15:15 <mlavalle> cool....
15:15:29 <mlavalle> next up is  https://bugs.launchpad.net/neutron/+bug/1486795
15:15:29 <openstack> Launchpad bug 1486795 in neutron "DVR: create or update port by using notify specific host rather than fanout" [High,In progress] - Assigned to Oleg Bondarev (obondarev)
15:16:09 <mlavalle> obondarev has been working on the fix:  https://review.openstack.org/#/c/231555
15:16:17 <mlavalle> he seems to be making progress
15:16:31 <mlavalle> any comments from the team on this one?
15:16:36 <regXboi> there are actually a couple of different PSs that are alternatives
15:16:49 <regXboi> so we should also look at the alternatives that are in the commit message
15:16:50 <mlavalle> do you have the url's?
15:17:02 <regXboi> https://review.openstack.org/221209 and https://review.openstack.org/215457
15:17:20 <regXboi> they are in the commit message for obondarev's patch
15:17:25 <mlavalle> will add them to the tracking list. thanks!
15:17:51 <regXboi> yw
15:18:11 <mlavalle> next up is https://bugs.launchpad.net/neutron/+bug/1486828
15:18:11 <openstack> Launchpad bug 1486828 in neutron "DVR: Notify specific agent when dealing with floating ips" [High,In progress] - Assigned to changzhi (changzhi)
15:18:39 <Swami> mlavalle: I thought that a similar patch was merged yesterday
15:18:53 <haleyb> The first of that series for create FIP was merged yesterday
15:19:10 <mlavalle> Swami: for 1486828?
15:19:16 <regXboi> mlavalle: yes
15:19:20 <Swami> haleyb: yes it was for associated fips.
15:19:21 <haleyb> https://review.openstack.org/231031
15:19:55 <mlavalle> haleyb: but there are more after this one, right?
15:20:01 <Swami> mlavalle: Yes it addresses the same issue, but just the association case, so it is partial fix.
15:20:01 <haleyb> there is still update and delete patches to merge
15:20:33 <carl_baldwin> haleyb: Those others seem to have broken off of the chain.  Do you have a link handy?
15:20:34 <Swami> mlavalle: yes there is delete and update. Probably oleg needs to rebase his delete patch since the first one landed.
15:20:44 <mlavalle> I have  https://review.openstack.org/#/c/215136 and https://review.openstack.org/#/c/231455
15:20:46 <haleyb> They somehow got lost from the depends section once it merged
15:20:47 <regXboi> https://review.openstack.org/#/c/231455/ for delete
15:20:53 <regXboi> https://review.openstack.org/#/c/215136/ for update
15:21:05 <mlavalle> regXboi: yeah, we are in synch
15:21:08 <carl_baldwin> regXboi: Thanks
15:21:13 <regXboi> no worries
15:22:03 <mlavalle> ok, next up is a new one:  https://bugs.launchpad.net/neutron/+bug/1505575
15:22:03 <openstack> Launchpad bug 1505575 in neutron "Fatal memory consumption by neutron-server with DVR at scale" [High,In progress] - Assigned to Oleg Bondarev (obondarev)
15:22:27 <regXboi> mlavalle: did you mean to skip one?
15:22:38 <mlavalle> regXboi: i'll come back
15:22:43 <regXboi> mlavalle: ack
15:23:02 <mlavalle> obondarev has a fix proposed for this one  https://review.openstack.org/#/c/23406
15:23:20 <mlavalle> https://review.openstack.org/#/c/234067 ^^^^
15:23:40 <mlavalle> it makes sense to me.....
15:24:14 <mlavalle> any comments?
15:24:14 <regXboi> yes, but jenkins needs to be mollified
15:24:43 <Swami> mlavalle: is this the right link
15:25:16 <Swami> mlavalle: got it, I used the first one.
15:26:18 <mlavalle> finally from me:  https://bugs.launchpad.net/neutron/+bug/1476097
15:26:18 <openstack> Launchpad bug 1476097 in neutron "[fwaas]Support fwaas to control east-west traffic in dvr router" [High,Triaged] - Assigned to lee jian (leejian0612)
15:26:31 <mlavalle> this is an approved rfe
15:27:13 <mlavalle> problem is that I only see an abandoned patchset associated with it. abandoned back in august and no more activity ssince then
15:27:15 <Swami> mlavalle: Yes I am working with mickey and sridhar to come up with a proposal. There are some implications to existing DVR design, but will discuss with the community for approval.
15:27:29 <mlavalle> Swami: ah perfect.....
15:27:50 <mlavalle> I will add a note the bug so we remember
15:27:57 <Swami> mlavalle: sounds good.
15:28:17 <mlavalle> and I believer regXboi added one more bug to the agenda.....
15:28:27 <regXboi> yes, this is an FYI
15:28:38 <regXboi> it came in this week and it's entangled code
15:28:55 <regXboi> so it's more so for folks to be aware of
15:29:38 <mlavalle> any other bugs we might be missing?
15:29:51 <carl_baldwin> regXboi: link?
15:30:04 <regXboi> https://bugs.launchpad.net/neutron/+bug/1505571
15:30:04 <openstack> Launchpad bug 1505571 in neutron "VM delete operation fails with 'Connection to neutron failed - Read timeout' error" [Undecided,Incomplete] - Assigned to Sonu (sonu-sudhakaran)
15:30:16 <Swami> mlavalle: there are a bunch of DVR bugs that are being worked out, may be we should also include those bugs.
15:30:18 <carl_baldwin> regXboi: Thanks!
15:30:53 <regXboi> carl_baldwin: yw
15:31:13 <mlavalle> Swami: let's discuss those in the dvr section
15:31:19 <Swami> mlavalle: thanks
15:31:32 <mlavalle> ok, moving on.....
15:31:52 <mlavalle> #topi Routed network segments
15:32:05 <mlavalle> #topic Routed network segments
15:32:44 <mlavalle> carl_baldwin: any updates here?
15:33:23 <carl_baldwin> I’ll just bring up the discussion going on in the spec.  The question brought forward is whether to continue with RoutedNetworkGroup as a DB object for grouping networks, or to solve the VM scheduling problem with tags (opaque metadata).
15:33:33 <carl_baldwin> So, the spec review is quite active.
15:34:22 <carl_baldwin> Please go read and weigh in if you feel strongly.  But, let’s keep the discussion technical and respectful.  :)
15:34:36 <mlavalle> will do :-)
15:34:46 <Swami> carl_baldwin: +1
15:34:51 <vikram> will go through this week
15:35:04 <mlavalle> thanks for the update
15:35:12 <mlavalle> next one is.....
15:35:18 <mlavalle> #topic DVR
15:35:24 * carl_baldwin ’s way of saying the discussion is heating up.  But, its all good.
15:35:30 <mlavalle> ping haleyb, regXboi Swami
15:35:31 <Swami> mlavalle: hi
15:35:31 <regXboi> I'll give Swami the floor
15:35:52 <Swami> mlavalle: The patches are up for review
15:36:25 <Swami> mlavalle: I think it is in a good shape, I have addressed some last review comment. I think it should see the light today.
15:36:53 <Swami> #link https://review.openstack.org/#/c/228582/
15:37:12 <Swami> #link https://review.openstack.org/#/c/229561/
15:37:29 <Swami> #link https://review.openstack.org/#/c/230079/
15:37:47 <Swami> #link https://review.openstack.org/#/c/225319/
15:38:02 <Swami> Additionally there was bug that was reported yesterday on DVR and LBaaS.
15:38:23 <Swami> I am triaging it right now, and once confirmed and I will update the bug report.
15:38:57 <regXboi> Swami: link would be good
15:39:09 <Swami> #link https://bugs.launchpad.net/neutron/+bug/1504726
15:39:10 <openstack> Launchpad bug 1504726 in neutron "The vm can not access the vip of load balancer under DVR enviroment" [Undecided,New] - Assigned to Swaminathan Vasudevan (swaminathan-vasudevan)
15:39:39 <regXboi> thx!
15:39:44 <mlavalle> thanks!
15:39:49 <Swami> Also I had some findings with the IPV6 issue, where the VM was not able to get an IP address
15:39:59 <Swami> In SLAAC and dhcp.
15:40:04 <haleyb> https://bugs.launchpad.net/neutron/+bug/1501969
15:40:04 <openstack> Launchpad bug 1501969 in neutron "No dhcp IPv6 assigned (slaac/slaac) with interface-add after VM boot" [Medium,Confirmed] - Assigned to Swaminathan Vasudevan (swaminathan-vasudevan)
15:40:10 <Swami> It seems that the RA from the router is not reaching the VM.
15:40:48 <Swami> I am looking at the flow rules right now to make sure if there is any flow rule that is preventing the packet from the reaching the VM instance.
15:40:59 <carl_baldwin> Swami: As a summary, what are you top issues for getting the gate job under control?
15:41:32 <Swami> carl_baldwin: The ARP issue and the FIP Namespace issue are the two concerns.
15:41:46 <haleyb> Swami: could you post the tcpdump log and/or ip6tables-save output as well for the slaac bug?
15:42:21 <Swami> haleyb: sure. I could not see any thing for ip6tables-save when I ran it. I will try it again.
15:43:03 <Swami> carl_baldwin: I assume that those two bugs should make the gate job stable.
15:43:14 <carl_baldwin> Swami: Thanks.
15:43:27 <Swami> haleyb: regXboi: Do you guys think any else could stabilize the DVR job.
15:43:27 <neiljerram> Swami: if it is iptables, possibly "ip6tables -L -v | grep DROP" will show the problem more readily than ip6tables-save.
15:43:44 <regXboi> Swami: I think we should get those patches in and then reassess
15:44:10 <Swami> regXboi: exactly that is my thought. Unless we get something in, we cannot come to a conclusion.
15:44:22 * regXboi isn't a big fan of making multiple changes at once
15:44:23 <Swami> neiljerram: thanks will check it out.
15:44:30 * haleyb 's brain is just programmed to iptables-save from working on the iptables_manager code
15:44:47 <mlavalle> haleyb: Lol
15:44:59 <Swami> haleyb: neiljerram: I will update further on my findings for the iptables issue.
15:45:04 <Swami> That's all I had.
15:45:18 <mlavalle> any thing else from haleyb regXboi ?
15:45:41 <neiljerram> Swami: please feel free to subscribe me to that bug, as I've had a bit of experience with RA and DHCPv6 trouble
15:45:41 <regXboi> I'm hoping to have some performance profiling for next week
15:45:59 <Swami> haleyb: I have addressed your review comments, can you take a look at it and also the dependent patches.
15:46:13 <regXboi> I've got the setup that I was using to find the network node issues changed over to DVR and just need to find time to run things through
15:46:19 <Swami> neiljerram: sure I will add you to the subscription list.
15:46:43 <haleyb> Swami: i'll look.  nothing else from me
15:46:51 <mlavalle> ok, moving on....
15:47:05 <mlavalle> #topic BGP dynamic routing
15:47:31 <vikram> hi
15:47:56 <tidwellr> hi
15:48:42 <mlavalle> any updates for today?
15:48:50 <vikram> tidwellr: I hope rebasing is in progress
15:48:56 <tidwellr> it is
15:49:00 <vikram> tidwellr: By when it will be done?
15:49:26 <vikram> tidwellr: Awaiting to post my changes ;)
15:49:44 <tidwellr> I had hoped to just decompose and have it rebased for you then, but that's not happening as quickly as I had hoped
15:49:50 <tidwellr> I'll rebase today
15:49:57 <vikram> tidwellr: no worries
15:50:17 <tidwellr> mlavalle: I don't have any updates
15:50:37 <mlavalle> vikram: anything else?
15:50:43 <mlavalle> tidwellr: thanks!
15:50:54 <vikram> mlavalle: nope
15:51:02 <mlavalle> ok..... moving on
15:51:08 <mlavalle> #topic DNS
15:51:26 <mlavalle> https://review.openstack.org/#/c/212213/ is passing the Jenkins tests now
15:51:26 <neiljerram> saw your most useful summary email on the ML - thanks
15:51:59 <carl_baldwin> mlavalle: ++
15:52:05 <mlavalle> It has to be said that the floating ips functionality for dns is complete
15:52:28 <mlavalle> I need to add the same to ports with external addresses
15:52:47 <mlavalle> but I wanted to have something completely functional for Tokyo
15:52:59 <mlavalle> so that is why I've focused on fip's so far
15:53:34 <mlavalle> I have also already the changes to nova to start sending the hostname to the port
15:53:48 <mlavalle> I will push the first patchset at the end of this week
15:53:58 <carl_baldwin> mlavalle: Do you have a review up for that?
15:54:22 <mlavalle> carl_baldwin: no, I will push the first patchset tomorrow
15:54:29 <carl_baldwin> mlavalle: ack
15:54:58 <mlavalle> but I have an example flow that I executed a couple of nights ago showing it working http://paste.openstack.org/show/476210/
15:55:12 <mlavalle> I used it to reply to a ML question
15:55:40 <mlavalle> there you can see that the vm's hostname is being pushed to designate along with the fip
15:55:53 <mlavalle> A and PTR records are being created
15:56:17 <carl_baldwin> mlavalle: cool
15:56:19 <carl_baldwin> !
15:56:45 <mlavalle> finally, I created a slide show in google for Tokyo: https://docs.google.com/presentation/d/1pdlMQjmIRfUc34MKXd0g-FWVkVeT7Scnf9EX46JXC-Q/edit#slide=id.p
15:57:16 <mlavalle> I believe I gave acccess to carl_baldwin and Kiall. Please let me know if indeed you can access it
15:57:46 <mlavalle> and that's it from me
15:58:01 <mlavalle> #topic Open discussion
15:58:24 <mlavalle> any additional topics to be shared with the team?
15:58:38 <regXboi> I've got a new DVR bug to bring up in channel when we are done
15:58:51 <neiljerram> I have undertaken to be a mentor for an outreachy applicant, for some Neutron project...
15:58:51 <regXboi> :)
15:59:05 <mlavalle> neiljerram: +++
15:59:09 <neiljerram> but I'm not sure what would be a good project!
15:59:32 <neiljerram> Would be happy for any advice - perhaps after this in openstack-neutron
15:59:47 <vikram> neiljerram: how about networking-onos?
15:59:49 <Swami> neiljerram: I would recommend that you choose the project that you are most aware of.
16:00:11 <mlavalle> ok, time is up
16:00:13 <carl_baldwin> neiljerram: could depend on the applicant, too.
16:00:17 <mlavalle> #endmeeting