14:02:10 <carl_baldwin> #startmeeting neutron_routed_networks
14:02:10 <openstack> Meeting started Tue Aug 23 14:02:10 2016 UTC and is due to finish in 60 minutes.  The chair is carl_baldwin. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:02:11 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:02:14 <openstack> The meeting name has been set to 'neutron_routed_networks'
14:02:29 <carl_baldwin> #topic Announcements
14:02:50 <carl_baldwin> We had the mid-cycle last week. I think it was productive for me.
14:03:24 <john-davidge> carl_baldwin: Agreed!
14:03:58 <carl_baldwin> I appreciated the opportunity to meet some new folks and see others of you again.
14:04:05 <carl_baldwin> Newton-3 is coming at the end of the month.
14:04:30 <mlavalle> #link https://releases.openstack.org/newton/schedule.html
14:04:39 <carl_baldwin> With the gate having so many issues, it may be difficult to make it. But, I think we can still get a few things in.
14:04:42 <carl_baldwin> mlavalle: Thanks!
14:05:51 <carl_baldwin> Any other announcements?
14:06:17 <carl_baldwin> Moving on...
14:06:20 <carl_baldwin> #topic docs
14:06:41 <carl_baldwin> I totally let this slip by without any discussion at the mid-cycle.
14:07:26 <carl_baldwin> But, I did post a review with the little that I had. If you could review it, that would be great. At this point, it is probably missing a lot of content.
14:07:29 <carl_baldwin> #link https://review.openstack.org/#/c/356013/
14:08:00 <carl_baldwin> It is really just my first quick stab at it.
14:08:10 <carl_baldwin> (Ignore the +2, I don't know what that reviewer was thinking)
14:08:28 <carl_baldwin> #topic testing
14:09:01 <mlavalle> yb and janzian have been making good progress
14:09:04 <yb> this week, I have write all the 3 test scenarios for the scenario test
14:09:21 <yb> The second one has problem running in multi-node environment
14:09:53 <yb> And I run all the test cases on Miguel vagrant environment
14:09:59 <mlavalle> #link https://review.openstack.org/#/c/347188/
14:10:25 <mlavalle> yb: I will take a look at that failure later this week
14:10:41 <yb> I think the environment issue Miguel addressed maybe a problem now
14:10:54 <carl_baldwin> yb: what issue is that?
14:10:55 <yb> mlavalle: thanks
14:11:20 <yb> you can see that in Miguel's last commends
14:11:31 <carl_baldwin> yb: okay, I'll take a look today.
14:12:01 <yb> thanks. Because I assume that the test running on two network-node that has two different segments
14:12:54 <mlavalle> and janzian's api tests for segments are here: https://review.openstack.org/#/c/353115/
14:12:55 <yb> I think That is all for my updates
14:13:12 <carl_baldwin> mlavalle: Is janzian 's ready for review?
14:13:18 <carl_baldwin> I could review that today also.
14:13:23 <mlavalle> carl_baldwin: nope
14:13:32 <carl_baldwin> mlavalle: Okay. Let us know.
14:13:43 <janzian> I should have a new patchset coming by the end of the day as well
14:13:43 <mlavalle> he will address some issues and later will be ready for review
14:14:07 <carl_baldwin> mlavalle: sounds good
14:14:30 <carl_baldwin> Good work guys. Anything else to discuss?
14:14:58 <yb> carl_baldwin:That is all for me
14:15:17 <carl_baldwin> yb: thanks
14:15:47 <carl_baldwin> #topic Deferred IPs for ports
14:15:58 <carl_baldwin> #link https://review.openstack.org/#/c/346278/
14:16:30 <carl_baldwin> Looks like we're close. I think all of the depends-on patches have merged. But, I got some feedback while traveling that I need to stop ignoring.
14:16:37 <carl_baldwin> I'll take a look at that today.
14:18:10 <carl_baldwin> I guess maybe we should just discuss some of the patches that are still in progress.
14:18:17 <carl_baldwin> #topic patches
14:18:25 <carl_baldwin> #link https://review.openstack.org/#/q/status:open+(topic:bp/routed-networks+OR+topic:bp/neutron-routed-networks+OR+topic:bp/service-subnets)
14:18:52 <carl_baldwin> The top one on my list is still https://review.openstack.org/317358
14:19:14 <xiaohhui> hi
14:19:16 <carl_baldwin> I'd like kevinbenton to look at it another time. I know he's been very busy lately.
14:19:21 <carl_baldwin> xiaohhui: hi
14:19:22 <xiaohhui> I made a update today
14:20:36 <xiaohhui> It fails the jekins now, but I don't think it is the code issue
14:20:59 <carl_baldwin> xiaohhui: The gate has been difficult to say the least.
14:21:08 <john-davidge> xiaohhui: there are still some instability issues in the gate, i wouldn't worry for now
14:21:08 <carl_baldwin> xiaohhui: Thanks for updating.
14:21:42 <carl_baldwin> xiaohhui: I'll have another look myself today.
14:22:18 <xiaohhui> Thanks, Carl
14:22:37 <carl_baldwin> #link https://review.openstack.org/#/c/350613
14:22:58 <carl_baldwin> john-davidge: Did you see my change here? I hope you don't mind that I updated it.
14:23:20 <john-davidge> carl_baldwin: You're update to the join/filter is much better, thank you! I'm not sure how we didn't think of that last week
14:23:37 <john-davidge> carl_baldwin: I'm currently reviewing your comment here https://review.openstack.org/#/c/350613/8/neutron/db/ipam_non_pluggable_backend.py
14:24:15 <john-davidge> if a subnet_id doesn't appear as a key in ip_pools, then we'll get an error on L73
14:24:16 <carl_baldwin> john-davidge: I wouldn't worry about that one too much. It is dead code afterall.
14:24:25 <john-davidge> I'm currently trying to decide if that can actually happen
14:24:26 <haleyb> The DVR job in that patch is failing pretty bad - http://logs.openstack.org/13/350613/8/check/gate-tempest-dsvm-neutron-dvr/df5dcec/logs/screen-q-svc.txt.gz?level=ERROR
14:24:34 <haleyb> don't know if it's related to our changes
14:25:00 <carl_baldwin> john-davidge: It won't because of L64
14:25:29 <john-davidge> carl_baldwin: I must be misunderstanding what the line does then
14:25:35 <john-davidge> *that line
14:26:15 <carl_baldwin> john-davidge: A defaultdict is a magic dict that generates a default value for a key when accessed in that way.
14:26:19 <john-davidge> either way, removing it doesn't result in any test failures with non-pluggable ipam re-enabled, so im happy to remove it
14:26:43 <john-davidge> carl_baldwin: in that case it's fine :) I'll push a patch soon to get rid of it
14:26:54 <carl_baldwin> john-davidge: Either way, since the tests failed, we need to run them again anyway at some point.
14:27:22 <haleyb> carl_baldwin: yes, did you see my link ^^ ?  strange failure
14:27:23 <carl_baldwin> john-davidge: that'll help other reviewers so that they aren't distracted by that changed.
14:27:26 <carl_baldwin> *change
14:27:39 <john-davidge> carl_baldwin: Thanks for all your reviews on this subject, I'm sure we can get this through before FF
14:28:01 <carl_baldwin> haleyb: thanks for point that out.
14:28:26 <carl_baldwin> haleyb: looking now...
14:29:06 <haleyb> carl_baldwin: that seems related to the dhcp port allocation issues i've seen before, but not exactly the same
14:30:08 <john-davidge> haleyb: That final traceback and InvalidInput error looks concerning, but I don't think it's caused by this change
14:30:16 <haleyb> https://bugs.launchpad.net/neutron/+bug/1583759 is our original bug, but that describes the symptom, not the problem
14:30:16 <openstack> Launchpad bug 1583759 in neutron "Invalid input for operation: IP allocation requires subnets for network" [High,New] - Assigned to Bharat Kumar (bharatkumar)
14:30:16 <carl_baldwin> haleyb: Yeah, I don't have any ideas yet.
14:30:21 <john-davidge> haleyb: Will keep an eye on it, thatnks for raising it
14:30:43 <haleyb> john-davidge: yeah, it's some kind of deadlock, but IP allocation fails because of it
14:32:24 <haleyb> carl_baldwin: the OSC change for subnet types is waiting for osc-lib i believe, i'll check on it today
14:32:40 <carl_baldwin> haleyb: Thanks.
14:32:43 <haleyb> https://review.openstack.org/#/c/342976/
14:33:29 <carl_baldwin> haleyb: I wonder if the switch to pluggable IPAM has exacerbated that issue somehow.
14:34:03 <haleyb> carl_baldwin: it could have, i definitely see that failure more in logstash now
14:35:47 <carl_baldwin> Okay, we should probably look in to this some more as a priority.
14:37:43 <carl_baldwin> haleyb: Is it always associated with a failure?
14:38:50 <haleyb> carl_baldwin: no, sometimes i see a pass, but the log has that error, at least that's what i remember from looking at the multinode-dvr jobs yesterday
14:40:45 <carl_baldwin> haleyb: Do you have time to look more into the failure on john-davidge 's patch or do you need some help?
14:41:32 <haleyb> carl_baldwin: i can look, but wonder if they're related to https://bugs.launchpad.net/neutron/+bug/1612798 which kevinbenton has a patch for
14:41:32 <openstack> Launchpad bug 1612798 in neutron "Move db retry logic away from API layer" [Critical,In progress] - Assigned to Kevin Benton (kevinbenton)
14:42:09 <mlavalle> several bugs need that fix
14:42:44 <haleyb> sql is not my forte
14:43:00 <carl_baldwin> gerrit just stopped responding for me
14:43:29 <haleyb> it needs more coffee :)
14:44:09 <carl_baldwin> There it goes.
14:45:09 <carl_baldwin> Sounds like we need to continue focusing some attention on gate issues.
14:45:28 <carl_baldwin> Just to wrap up this meeting, I'd like to bring up two more patches.
14:45:34 <carl_baldwin> Just FYI
14:45:46 <haleyb> that seemed to be armando's point last night
14:45:49 <carl_baldwin> #link https://review.openstack.org/#/c/293305/
14:46:07 <carl_baldwin> This is adding name and description to segments. Hopefully, we can get this worked out soon.
14:46:27 <carl_baldwin> #link https://review.openstack.org/#/c/358658/
14:46:56 <carl_baldwin> This is mlavalle 's patch for generic resource pools inventory. Good to see this coming, mlavalle !
14:47:09 <mlavalle> not ready to review yet
14:47:15 <carl_baldwin> mlavalle: ack
14:47:22 <mlavalle> Made good progress is the mid-cycle
14:47:27 <carl_baldwin> I'm curious though. So, I'm going to peek anyway. :)
14:47:36 <mlavalle> you are welcome
14:47:50 <carl_baldwin> Any other patches to bring up?
14:48:43 <carl_baldwin> If not, let's all jump over to the neutron room and see how the battle with the gate is going. Hopefully, we can help out.
14:49:09 <carl_baldwin> Great work, everyone!
14:49:13 <carl_baldwin> #endmeeting