14:00:15 #startmeeting networking 14:00:15 Meeting started Tue Sep 8 14:00:15 2015 UTC and is due to finish in 60 minutes. The chair is mestery. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:16 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:18 The meeting name has been set to 'networking' 14:00:22 #link https://wiki.openstack.org/wiki/Network/Meetings Agenda 14:00:23 hi 14:00:24 o/ 14:00:33 o/ 14:00:36 hi 14:00:37 hi 14:00:38 Basically, today is 1) Announcements 2) Bugs 3) RC1 (with some yelling about only merging stuff approved to go in) 14:00:44 #topic Announcements 14:00:46 hi 14:00:49 hi 14:00:49 Liberty-3 is out! 14:00:52 #link https://launchpad.net/neutron/+milestone/liberty-3 14:01:03 Hello 14:01:07 #info Thanks to armax for going above and beyond last week 14:01:12 Buenos Dias! (Good Morning) 14:01:16 We as a project are lucky to have armax around, and I wanted to thank him! 14:01:22 o/ :) 14:01:27 armax++ 14:01:33 mestery: sorry I am late.. 14:01:37 armax^2 14:01:37 emagana: No worries 14:01:38 :-) 14:01:39 * armax blushes 14:01:50 * mestery thinks most people wish he was on vacation all the time after how awesome armax was last week 14:01:58 Seriously, nice work armax 14:01:59 hi 14:02:04 And nice work to the entire TEAM! 14:02:05 mestery: ok, now you’re pushing it 14:02:07 :) 14:02:08 :) 14:02:09 armax: absolutely! 14:02:16 Liberty-3 was a great team effort by all. 14:02:19 We should all be proud of that. 14:02:29 #link https://wiki.openstack.org/wiki/ReleaseNotes/Liberty#OpenStack_Liberty_Release_Notes 14:02:36 #info Please keep updating Liberty release notes 14:02:50 That's all the announcements I had 14:02:57 Anyone have anything else to share with the broader team this week? 14:03:22 OK, lets slide along to bugs! 14:03:24 #topic Bugs 14:03:38 Starting at the top ... 14:03:49 #link https://bugs.launchpad.net/neutron/+bug/1484148 14:03:50 Launchpad bug 1484148 in neutron "neutronclient gate broken following VPNaaS infra changes" [Critical,Confirmed] - Assigned to Paul Michali (pcm) 14:04:05 This one belongs to pc_m, who doesn't appear to be here obviously :) 14:04:34 I've posted some comments in the bug now, we'll move along unless someone else has an update here. 14:04:52 #link https://bugs.launchpad.net/neutron/+bug/1488282 14:04:53 Launchpad bug 1488282 in neutron "Gate failures with 'the resource could not be found'" [High,Confirmed] - Assigned to Salvatore Orlando (salvatore-orlando) 14:05:02 This one is assigned to salv-orlando :) 14:05:10 I believe 2 weeks ago he volunteered to triage it a bit more 14:05:25 Reported by armax 14:06:07 There was a tempest fix for that I believe? 14:06:10 * mestery looks at the logstash query 14:06:18 mestery but I declared it as "non neutron" 14:06:20 * pc_m sorry late 14:06:20 mestery: comments by salvatore say it seems like a nova bug 14:06:22 HenryG: that was a different NotFound 14:06:34 HenryG: actually we had a few that hit at the same time 14:06:35 hi salv-orlando :) 14:06:35 I'm just waiting for a confirmation from the nova people to remove neutron from the affected list 14:06:43 one fix went into tempest 14:06:45 salv-orlando: Awesome 14:06:47 the other in neutron 14:06:53 Also, the logstash query shows no hits since 9-3 14:06:58 So .... 14:07:02 this one, as salv-orlando points out 14:07:06 might be noa 14:07:07 nova 14:07:10 salv-orlando: may be we should move neutron out from "confirmed" to something else? 14:07:21 ajo: not so hasty 14:07:28 O:) 14:07:36 lol 14:07:39 ajo: sure. will do. sir. 14:07:40 typically when you do so, the bug reassert itself as neutron’s fault ;) 14:07:41 mestery: I'm here 14:07:45 I proved with this one 14:07:49 once armax says it's ok to do 14:07:51 lol 14:07:52 pc_m: Once second 14:07:53 yeah, we like to close unfixed bugs 14:07:55 https://bugs.launchpad.net/neutron/+bug/1490051 14:07:56 Launchpad bug 1490051 in neutron "test_assert_pings_during_br_int_setup_not_lost fails with oslo_rootwrap.wrapper.NoFilterMatched" [High,In progress] - Assigned to Ihar Hrachyshka (ihar-hrachyshka) 14:08:03 I closed it and it showed up 14:08:08 armax: Silly you ;) 14:08:20 mestery: well that was part of the strategy to lure the bug out 14:08:24 bugs have souls you know? 14:08:30 armax, we need another phantom state (not closed) ... not to disturb the bugs... 14:08:31 Oh for sure they do 14:08:32 he assumed no logstash hits during weekend and labour day == no bug 14:08:37 lol 14:08:47 ihrachys: Doesn't the entire world get Labor Day off? ;) 14:08:51 ihrachys: no hits during the weekend no 14:08:56 OK 14:08:58 Lets leave it oipen 14:09:03 salv-orlando looks to ahve this under hand 14:09:06 mestery: well, if entire world == NA, then yes 14:09:10 pc_m: This one was your's https://bugs.launchpad.net/neutron/+bug/1484148 14:09:11 Launchpad bug 1484148 in neutron "neutronclient gate broken following VPNaaS infra changes" [Critical,Confirmed] - Assigned to Paul Michali (pcm) 14:09:14 mestery: yes but on may 1st 14:09:14 ihrachys: lol 14:09:17 pc_m: Any updates there? 14:09:22 I changed the project-config job and the neutron-client code. Infra indicated they want Neutron cores to review the project config change, before they will review. 14:09:22 ihrachys: we had an occurence today by the looks of it 14:09:34 #link https://review.openstack.org/#/c/209887/ 14:09:39 armax: I debug it now, lots of weird stuff seen in logs 14:09:48 The other change, needs this upstreamed, so it can be tested. 14:09:57 pc_m: Thanks! 14:10:07 armax: Can you review 209887 today yet? 14:10:13 Currently, VPN tests are skipped, so it works. 14:10:28 dougwig is out hunting this week with no access to cell or email (so we shoudl all fill up his inbox and voicemail) :) 14:10:32 Works as in, doesn't fail, but is not really tested. 14:11:09 OK, lets move to the next one I'd like to discuss 14:11:14 #link https://bugs.launchpad.net/neutron/+bug/1491131 14:11:17 Launchpad bug 1491131 in neutron kilo "Ipset race condition" [High,New] 14:11:17 mestery, lol 14:11:34 This one appears to affect Kilo as well as master 14:11:53 ah, yes, the racy thing, I totally forgot 14:12:29 shihanzhang has a patch out which may be related (https://review.openstack.org/#/c/122368/) 14:12:34 I will follow that up during this week , seems assigned to cedric 14:12:40 It was abandoned 14:12:46 ajo: Awesome, thanks! 14:12:47 related in the sense that it acvoid triggering the bug 14:12:58 *avoids triggering 14:13:00 mestery, there's some previous patch which already got in that fixes some of the issues they see in kilo 14:13:14 salv-orlando: Well, that's related isn't it? :) 14:13:15 but, there's something that triggers ipset delete when it's already gone (or may be never created...) 14:13:18 ajo: The patch got into master? 14:13:37 ajo,: do you know if someone beyond the reporter managed to reproduce it? 14:13:54 mestery, I will review the whole thread and link the right patches. 14:13:56 salv-orlando: not that I know 14:14:21 ajo: Thanks! 14:14:26 salv-orlando: I guess it could make sense to make ipset.delete("aaa") not fail, if the ipset didn't exist 14:14:49 it could hide some source of races, but at least, we wouldn't block the agent. A warning may work.. 14:15:16 That's the last bug I wanted to discuss before we move along. Does anyone have any other bugs they want to highlight here? 14:15:16 salv-orlando, i am trying to reproduce it but at the moment i fail to reproduce it 14:15:56 ajo: we could put in place measures to avoid triggering the bug, but we'd also be masking the underlying race condition. so we might not necessarily do something good 14:16:06 as at some point the race will explode in our face 14:16:20 salv-orlando, yeah, but the failure mode is horrendous: broken agent until agent restart 14:16:22 (well in the face of the ref impl lieutenant) 14:16:35 *cough cough* kevinbenton *cough cough* 14:16:36 salv-orlando: at least, we could reduce the impact 14:16:57 mestery: kevinbenton is surely still sleeping 14:16:58 and may be link the bug to a logstash query to look for ocurrences 14:17:07 said that, salv-orlando , I couldn't find those ocurrences in logstash upstream 14:17:13 armax: He's driving home from disneyland actually and told me he'd miss today's meeting 14:17:17 (for ipset trying to delete something not existing..) 14:17:44 but he did so last night 14:17:51 mestery: last time I checked LA is not that far 14:17:55 from SF 14:17:57 :) 14:18:03 armax: Nope, he's on the road right now in fact :) 14:18:08 He must have not gotten enough Mickey Mouse 14:18:15 :-) 14:18:20 OK, lets move on before we lost the meeting completely :) 14:18:22 NExt up ... 14:18:24 mestery: um…he’s probably eating corn somewhere on the road 14:18:27 lol 14:18:29 #topic Docs 14:18:31 * carl_baldwin sorry to be late 14:18:32 emagana: What have you got for us today? 14:18:36 Hello! 14:18:44 * mestery sends carl_baldwin to the back of the room 14:18:50 mestery: not much really 14:18:54 emagana: That's ok :) 14:18:58 I have not seen new review for us.. 14:19:13 * markmcclain gets in line behind carl_baldwin 14:19:25 however, after release-3 is when we should invest even more in Documenting the new features 14:19:35 lol 14:19:48 emagana: That sounds good! 14:20:00 emagana: Is there a plan, and does it involve Sam-I-Am as well? :) 14:20:01 emagana, me owes QoS stuff 14:20:05 mestery: I should have more details next week, after our networking docs meeting on Friday 14:20:10 emagana: Thanks! 14:20:23 #info Plan for documentation changes for new Liberty-3 items will be unveiled next week after networking docs meeting on Friday 14:20:24 ajo: Yeah.. I will help there 14:20:50 mestery: Always involves Sam-I-Am 14:20:51 thanks emagana , may be a simple template or initial "empty" review on where to put things would help, then I can follow up adding the details 14:21:02 :) 14:21:13 ajo: sounds good.. You will have it 14:21:17 emagana++ 14:21:20 I will add you as reviewer 14:21:43 mestery: all from me! 14:21:52 emagana: Thanks! 14:22:14 #topic Liberty-RC1 14:22:20 #link https://launchpad.net/neutron/+milestone/liberty-rc1 14:22:33 Lets walk through what we have targeted and have a hard look at what realistically will make it 14:22:39 Because I still think what's there isn't realistic at this point 14:22:47 First up 14:22:49 #link https://blueprints.launchpad.net/neutron/+spec/get-me-a-network 14:22:59 haleyb: This seems like it's Mitaka at this point, does that sound right? 14:23:00 mestery: yes, I simply let the stuff spill over RC1 14:23:08 armax: I would have done the same :) 14:23:27 mestery: yes 14:23:33 haleyb: That was easy :) 14:23:34 mestery: what's the deadline for the RC1 cut? 14:23:41 haleyb: not even foundational work? 14:23:53 ajo: https://wiki.openstack.org/wiki/Liberty_Release_Schedule 14:24:00 haleyb: that prepares it for an early land in M? 14:24:02 We'll start cutting RC candidates the week of Sept. 21 14:24:30 ah, ok, so it's not fixed for RC, just between 21sept to 9oct 14:24:30 armax: yes, possibly foundation work 14:24:53 haleyb: well...by now we’d need a little more than possibly 14:24:54 haleyb: If you land a patch or two for this, we'll let them in pending review then. 14:25:04 :) 14:25:10 OK 14:25:14 Next up 14:25:16 #link https://blueprints.launchpad.net/neutron/+spec/lbaas-ref-octavia 14:25:21 dougwig is gone, so perhaps blogan or xgerman can give us an update here 14:25:31 I believe this one is still targeted as being landable, though no code shows up there 14:25:36 brogan worked all weekend so we should be good for the RC 14:25:39 mestery: yeah i believe it is still landable 14:25:47 +! 14:25:48 mestery: that must require landing code in more than one repo, though right? 14:25:49 +1 14:25:55 Yay! 14:25:56 mestery: so stuff might be out of our control 14:25:59 armax: Right! 14:26:02 armax: correct 14:26:06 Where did hte code land xgerman blogan_ ? 14:26:17 armax: Well, it's in our control as it's in the Stadium 14:26:19 octavia and lbaas repo 14:26:22 Awesome 14:26:33 #info Octavia as the default LBaaS implementation will make Liberty 14:26:44 blogan_: Feel free to mark Implemented once you're done 14:26:48 nice! 14:26:55 mestery: what about project-config changes, devstack changes etc 14:27:10 we got them covered 14:27:19 armax: See, xgerman has it covered? 14:27:41 mestery: tight! 14:27:51 yep, very tight 14:28:00 Lets move on, we've got plenty of yak shaving to complete while painting the bike shed yet 14:28:06 #link https://blueprints.launchpad.net/neutron/+spec/wsgi-pecan-switch 14:28:14 I spoke to kevinbenton today 14:28:17 And he says we still have a shat at this one 14:28:25 Though blogan_ said service plugins don't load with the current branch, which is sad 14:28:37 kevinbenton remains vigilant he can get this landed this week yet 14:28:45 * mestery waits for armax 14:28:55 mestery: that’s doable 14:28:57 mestery: well i would say most probably do work, but its not the same as the current wgi server 14:28:59 mestery: we have targeted code 14:29:07 mestery: I've got some time to help kevinbenton with any reviews/work needed to land it 14:29:07 mestery: so we have an idea of what to expect 14:29:24 markmcclain: Awesome! Please have a look here: https://review.openstack.org/#/q/project:openstack/neutron+branch:feature/pecan+status:open,n,z 14:29:26 #link https://review.openstack.org/#/q/project:openstack/neutron+branch:feature/pecan+status:open,n,z 14:29:30 armax: We do 14:29:33 blogan_: OK 14:29:37 So, this one looks like it's covered 14:29:39 mestery: blogan_ did a hack that forces the pecan app to load all plugins at startup 14:29:46 salv-orlando: Nice 14:29:48 let's say that from an end user perspective it works 14:29:53 rofl 14:29:53 mestery: this is just discussing whether it should roll back into master correct? it's not going to replace the current wgi is it? 14:30:05 blogan_: God no, not at this stage! :) 14:30:09 We're just rolling back to master 14:30:12 mestery: but tell armax not to look at the code. He'll faint. 14:30:14 So we can release it as experimental 14:30:52 salv-orlando: thanks for the warning 14:31:00 sorry got dc'ed 14:31:02 armax: Please, refrain from looking for a bit :) 14:31:08 I’ll make sure I have a cpr kit next to me when I do 14:31:15 Awesome 14:31:17 OK, lets move along 14:31:18 #link https://blueprints.launchpad.net/neutron/+spec/vm-without-l3-address 14:31:29 armax reverted this one, luckily he had a CPR kit close by 14:31:37 The author has reached out 14:31:46 But the new direction this one goes in requires an API change (yuck) 14:31:49 I'm inclined to boot this one out 14:31:51 Thoughts? 14:31:55 mestery: unfortunately I hadn't 14:32:10 Port security is a work around IMO 14:32:12 * mestery takes a note to get armax portable CPR kits to keep near his computer 14:32:21 kevinbenton: Sounds fair 14:32:30 kevinbenton: is alive! 14:32:32 Shall we move this one out to Mitaka then? 14:32:37 kevinbenton armax: ^^^^ 14:32:54 I think so. We need a more fine grained port security API 14:33:03 Which will take some time to develop 14:33:07 AGreed 14:33:15 #info vm-without-l3-address is a Mitaka feature now! 14:33:19 kevinbenton: well, it depends on how fast we can get consensus on the approach 14:33:32 but yeah, probably it’s safer than shipping a hole in the ship 14:33:47 no-one wants neutron to sink 14:33:51 mestery: do you? 14:33:54 The only difference is that they won't have Mac spoofing protection 14:34:05 armax: Of course not my friend 14:34:08 They can still do the non L3 addressed VM 14:34:11 armax: but almost everyone rocks the boat ;) 14:34:14 It's out of Liberty 14:34:14 Lets move along 14:34:16 #link https://blueprints.launchpad.net/neutron/+spec/external-dns-resolution 14:34:19 agree. we may need to combine unaddress port/port security with rbac 14:34:27 This one looks close, mlavelle is working on it now 14:34:39 I expect it to land this week once he re-spins to address current issues 14:34:57 mestery: +1 This one is not intrusive. 14:35:06 carl_baldwin: Cool 14:35:13 Speaking of carl_baldwin .... 14:35:14 #link https://blueprints.launchpad.net/neutron/+spec/address-scopes 14:35:15 and some operators need this one 14:35:21 * mestery calls carl_baldwin up to the front of the classroom 14:35:29 I meant the dns resolution one 14:35:30 carl_baldwin: Thoughts on this one making Liberty at this point? 14:35:35 emagana: ++ 14:35:35 * carl_baldwin walks up… 14:35:46 carl_baldwin: I know you've been pushing hard on this one for the past few weeks yet 14:35:57 The optimistic in me wants to say yes. It is functional. But... 14:36:28 It needs tests to be completed. I suspect some interaction with DVR and I need to look at what will happen with static routes. 14:36:53 carl_baldwin: OK, I can leave it in there for now and we can re-evaluate next week? 14:36:55 Sound fair? 14:37:02 * salv-orlando thinks carl_baldwin deserves a wildcard as he was already mothballed for liberty with pluggable-ipam 14:37:02 I can join the test team. 14:37:05 Though I really want it in I think it would be better to merge early in Mitaka. 14:37:34 carl_baldwin: Well, if that's your thinking, lets just move it out now. Sound ok? 14:37:41 mestery: ok 14:37:50 * carl_baldwin stashes away the wildcard for later. 14:37:53 lol 14:38:15 :D 14:38:22 Next up 14:38:23 #link https://blueprints.launchpad.net/neutron/+spec/better-quotas 14:38:31 salv-orlando: I think we just need to review and merge 3 more patches right? 14:38:37 Or maybe it's 4 14:38:55 I think it's 3 14:39:08 Awesome 14:39:10 I can add some more testing coverage during RC phase, I reckon that would be ok, wouldn't it? 14:39:17 salv-orlando: ++ 14:39:21 Thanks salv-orlando 14:39:33 Next up 14:39:35 #link https://blueprints.launchpad.net/neutron/+spec/restructure-l2-agent 14:39:44 rossella_s has been working on this one for quite a while now 14:39:45 carl_baldwin: take out that wildcard for salv-orlando :P :) 14:39:52 9 outstanding patches still rossella_s? 14:39:58 * carl_baldwin hands it to salv-orlando 14:40:01 mestery, 5 14:40:06 rossella_s: Not bad! 14:40:14 rossella_s: Are you getting enough reviews for these? 14:40:20 rossella_s: Or do you need more people to help focus on these? 14:40:36 in the last days I got more review. Reviewers are always very welcome anyway 14:40:43 * carl_baldwin can jump in 14:40:51 carl_baldwin: Please do! 14:40:57 carl_baldwin thanks! 14:41:08 #info 5 patches left for agent refactoring, please jump in and help rossella_s land these this week! 14:41:13 Thanks rossella_s! 14:41:23 mestery, thank you 14:41:27 Next up 14:41:28 #link https://blueprints.launchpad.net/neutron/+spec/add-availability-zone 14:41:47 If I'm not mistaken, I think russellb was reviewing these and testing htem recently 14:42:00 armax also had a pass on these ones 14:42:07 mestery: yes, I meant to continue 14:42:08 yeah i got the first patch to work 14:42:11 wanted to do another pass 14:42:14 amotoki said API patch is OK 14:42:16 rossella_s: I will re-review them all, sorry for being a bit slow on that. 14:42:16 russellb: \o/ 14:42:20 I'm waiting his review to Scheduler patch. 14:42:32 So, looks like armax russellb and amotoki have this one covered 14:42:43 perhaps we need to review them combined with the scheduler patch. 14:42:48 amotoki: ++ 14:42:48 ok if it's targeted, i'll be sure to do another pass ... i'm traveling this week again though 14:42:55 hopefully can find some time in the evening 14:42:59 ajo thanks a lot! 14:43:07 amotoki++ russellb++ armax++ 14:43:08 :) 14:43:17 russellb: Thanks! 14:43:20 lets check it next week again 14:43:25 And thanks to amotoki and armax and ajo too :) 14:43:28 amotoki: ++ 14:43:30 Thanks! 14:43:38 Next up 14:43:39 #link https://blueprints.launchpad.net/neutron/+spec/lbaas-l7-rules 14:43:44 blogan xgerman: Any update here? 14:43:53 dougwig made it sound like this one may be out of Liberty now 14:43:58 I think we deferred that to M 14:44:05 xgerman: Cool, thanks! 14:44:05 not to me, I did nothing related :D 14:44:44 Next up 14:44:45 #link https://blueprints.launchpad.net/neutron/+spec/vlan-aware-vms 14:44:57 I haven't reviewed this patches, has anyone been looking at these ones? 14:45:02 I know they landed only a few weeks ago 14:45:14 mestery: looks like they have not had any beefy review 14:45:26 armax: I figured 14:45:36 mestery: they were submitted too late in the cycle to get enough attention 14:45:38 I'm inclined to give it one more week, but I don't remain optimistic these land in Liberty 14:45:44 armax: ++ 14:46:20 I think they can be delayed to M 14:46:24 at this point 14:46:33 too bad 14:46:44 Yup 14:46:46 Last one 14:46:47 #link https://blueprints.launchpad.net/neutron/+spec/neutron-flavor-framework-templates 14:46:56 Since dougwig isn't here, I'm inclined to just move this to Mitaka 14:47:11 mestery: he’s not doing the work though 14:47:19 well, we have the CLI and are working on a flavor -> service type thingy 14:47:23 mestery: someone from xgerman is doing it 14:47:32 I just moved it out, should I move it back in? 14:47:35 xgerman: Are you saying htis has a shot now? 14:47:48 there is still a chance 14:47:48 xgerman: Please move the assignee to the actual person doing the work :) 14:47:53 xgerman: Ok ,letaving it in 14:47:55 mestery: so long as the server side is fairly complete yes 14:47:56 ok, will do 14:48:11 but CLI is clearly blocked on the server side getting in shape 14:48:19 armax: WEll 14:48:20 Also 14:48:26 We're releasing the Liberty client today 14:48:37 Which means we won't release another one until after the Liberty release I guess 14:48:39 * ihrachys dances 14:48:46 * mestery dances with ihrachys 14:48:48 mmh, maybe then we are out 14:48:52 * ajo dances too 14:49:06 xgerman: Please add the reviews to the BP as well: https://blueprints.launchpad.net/neutron/+spec/neutron-flavor-framework-templates 14:49:20 xgerman: the cli was ready to merge 14:49:21 it is a race condition between CLI dependency freeze and RC features :-( 14:49:27 Yes 14:49:28 well, my powers don’t let me change much on the BP 14:49:32 xgerman: I only stopped it because I looked at hte server side code 14:49:36 …and fainted 14:49:36 xgerman: In the work area 14:49:46 * mestery rushes the CPR kit to armax 14:49:56