| *** vivek-ebay has quit IRC | 00:06 | |
| *** minwang2 has joined #openstack-lbaas | 00:16 | |
| *** minwang2 has quit IRC | 00:20 | |
| *** vivek-ebay has joined #openstack-lbaas | 00:20 | |
| *** minwang2 has joined #openstack-lbaas | 00:29 | |
| *** minwang2 has left #openstack-lbaas | 00:30 | |
| *** minwang2 has joined #openstack-lbaas | 00:30 | |
| *** Varun_Lodaya has joined #openstack-lbaas | 00:37 | |
| *** chlong has joined #openstack-lbaas | 00:42 | |
| *** vivek-ebay has quit IRC | 00:42 | |
| *** macjack has quit IRC | 00:58 | |
| *** Varun_Lodaya has quit IRC | 01:21 | |
| *** Brian has joined #openstack-lbaas | 01:35 | |
| *** Brian is now known as Guest68286 | 01:35 | |
| *** Guest68286 has quit IRC | 01:37 | |
| *** vivek-ebay has joined #openstack-lbaas | 01:43 | |
| *** manishg has quit IRC | 01:44 | |
| *** vivek-ebay has quit IRC | 01:47 | |
| *** KunalGandhi has joined #openstack-lbaas | 01:53 | |
| *** denghui has joined #openstack-lbaas | 01:59 | |
| *** Brian has joined #openstack-lbaas | 02:04 | |
| *** Brian is now known as Guest83407 | 02:05 | |
| *** KunalGan_ has joined #openstack-lbaas | 02:28 | |
| *** KunalGandhi has quit IRC | 02:30 | |
| *** vivek-ebay has joined #openstack-lbaas | 02:52 | |
| *** Varun_Lodaya has joined #openstack-lbaas | 03:02 | |
| *** KunalGandhi has joined #openstack-lbaas | 03:08 | |
| *** bharathm has joined #openstack-lbaas | 03:08 | |
| *** vivek-ebay has quit IRC | 03:10 | |
| *** KunalGan_ has quit IRC | 03:12 | |
| *** KunalGandhi has quit IRC | 03:14 | |
| *** KunalGandhi has joined #openstack-lbaas | 03:15 | |
| *** KunalGandhi has quit IRC | 03:20 | |
| *** vivek-ebay has joined #openstack-lbaas | 03:28 | |
| *** vivek-eb_ has joined #openstack-lbaas | 03:33 | |
| *** vivek-ebay has quit IRC | 03:34 | |
| openstackgerrit | Brandon Logan proposed openstack/neutron-lbaas: Use admin context in octavia driver, improve logs https://review.openstack.org/225542 | 03:46 |
|---|---|---|
| *** manishg has joined #openstack-lbaas | 03:54 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia: Update controller logging https://review.openstack.org/225543 | 03:55 |
| *** manishg has quit IRC | 03:59 | |
| *** vjay15 has joined #openstack-lbaas | 04:03 | |
| *** Guest83407 has quit IRC | 04:12 | |
| *** KunalGandhi has joined #openstack-lbaas | 04:14 | |
| *** vjay15 has quit IRC | 04:18 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia: Fix a typo in the .coveragerc file https://review.openstack.org/225548 | 04:26 |
| rm_work | ALMOST got this gate change merged johnsom | 04:27 |
| rm_work | working on the infra peeps now :) | 04:27 |
| *** vivek-eb_ has quit IRC | 04:32 | |
| *** vivek-ebay has joined #openstack-lbaas | 04:32 | |
| *** minwang2 has quit IRC | 04:39 | |
| *** vjay15 has joined #openstack-lbaas | 04:39 | |
| *** minwang2 has joined #openstack-lbaas | 04:40 | |
| sbalukoff | Nice! | 04:41 |
| rm_work | merged | 04:43 |
| rm_work | rechecking the other patch | 04:43 |
| rm_work | https://review.openstack.org/#/c/223265/ just showed up in zuul checks | 04:43 |
| rm_work | hmm though it looks like it's using the old job config :/ | 04:44 |
| *** vjay15 has quit IRC | 04:44 | |
| rm_work | might have to wait a few minutes and try again once this fails | 04:44 |
| *** vjay10 has joined #openstack-lbaas | 04:46 | |
| *** manishg has joined #openstack-lbaas | 04:55 | |
| *** manishg has quit IRC | 04:59 | |
| *** KunalGan_ has joined #openstack-lbaas | 05:03 | |
| *** KunalGandhi has quit IRC | 05:04 | |
| *** vivek-eb_ has joined #openstack-lbaas | 05:11 | |
| *** vivek-ebay has quit IRC | 05:11 | |
| *** Brian has joined #openstack-lbaas | 05:13 | |
| *** Brian is now known as Guest68836 | 05:13 | |
| *** Shijia has joined #openstack-lbaas | 05:15 | |
| Shijia | Hi! I hava a problem about how to configure lbaasv2 in devstack. My devstack enabled senlin and magnum now. When set enable_service q-lbaasv2 in local.conf, it will say do not enable lbaasv1 and lbaasv2 both. Just disable q-lbaasv1 doesn't work. Any suggestions? | 05:19 |
| rm_work | Shijia: I am not sure what your setup looks like, but compare to this (what I use to spin up my devstack): https://gist.github.com/rm-you/f7585ca4932b3ee1eed9 | 05:29 |
| rm_work | maybe you can compare and find what is different? | 05:30 |
| johnsom | Shijia Make sure you don't have just lbaas enabled either. | 05:30 |
| rm_work | johnsom: when did you start working on my schedule? :P | 05:31 |
| johnsom | Hahaha, well, I'm not really on your schedule. I'm just about to turn in. | 05:31 |
| johnsom | Thought I would check IRC before doing so as I wasn't watching IRC while working on my laptop. | 05:32 |
| johnsom | FYI, the new coverage package is breaking the sonar test. I thought it was just a config typo, but the tox -e cover now fails with two issues. Strangely the normal py27 and py34 pass those. | 05:33 |
| *** Shijia_ has joined #openstack-lbaas | 05:33 | |
| rm_work | T_T | 05:33 |
| rm_work | weird | 05:34 |
| johnsom | So, it may just be a bad coverage release. It's not hurting anything (non-voting) so I'm going to pick it up in the morning | 05:34 |
| rm_work | k | 05:34 |
| rm_work | i'm going to recheck this one more time and it should be good | 05:34 |
| rm_work | now that the tests have registered | 05:34 |
| rm_work | if i'd been more patient it would be close to passing by now <_< | 05:34 |
| johnsom | Ok, cool. I hope everything passes. | 05:34 |
| johnsom | Catch you about lunch time, pacific time, tomorrow... grin | 05:35 |
| johnsom | Or whenever you log back on. | 05:36 |
| *** Shijia has quit IRC | 05:36 | |
| *** vivek-eb_ has quit IRC | 05:37 | |
| *** KunalGan_ has quit IRC | 05:42 | |
| *** Shijia__ has joined #openstack-lbaas | 05:43 | |
| *** Shijia_ has quit IRC | 05:46 | |
| *** minwang2 has quit IRC | 05:49 | |
| rm_work | lol indeed | 05:53 |
| *** manishg has joined #openstack-lbaas | 05:55 | |
| *** nmagnezi has joined #openstack-lbaas | 05:56 | |
| *** numans has joined #openstack-lbaas | 05:59 | |
| *** Shijia has joined #openstack-lbaas | 05:59 | |
| *** Shijia__ has quit IRC | 06:00 | |
| *** manishg has quit IRC | 06:00 | |
| *** Shijia_ has joined #openstack-lbaas | 06:01 | |
| *** Shijia has quit IRC | 06:04 | |
| *** Alex_Stef has joined #openstack-lbaas | 06:25 | |
| *** Alex_Stef is now known as Alex_Staf | 06:25 | |
| rm_work | whelp, the tests are running... waiting on first VMs to spin up | 06:30 |
| * rm_work crosses fingers | 06:30 | |
| sbalukoff | Good luck! | 06:34 |
| rm_work | looking promising? some tests passing on at lease one of them | 06:34 |
| rm_work | *least | 06:34 |
| sbalukoff | By the way, I'm trying out your script for standing up a new Octavia devstack up front. Are you using vagrant or some other tool to house your devstack? | 06:34 |
| rm_work | sbalukoff: i should be | 06:35 |
| rm_work | i had it in my list of TODOs to set up ansible to spin up and deploy the script | 06:35 |
| sbalukoff | Anything in particular I should know about your vagrant environment? (ie. how much RAM / CPU do you throw at it? Any particular special networking settings?) | 06:35 |
| rm_work | but as of now, i just manually spin (or more commonly, just re-image from a clean ubuntu 14.04) RAX public cloud instances | 06:35 |
| rm_work | wget the script | 06:35 |
| rm_work | and run it | 06:35 |
| rm_work | I usually use 8GB because that is what the gate uses | 06:35 |
| rm_work | but it would probably(?) run on 4G | 06:36 |
| sbalukoff | Ok. | 06:36 |
| rm_work | it also works in HP instances (tested at LEAST once, like two weeks ago :P) | 06:36 |
| sbalukoff | Well, I'm gathering notes on different ways to run this stuff in preparation for writing docs that I'm guessing will ultimately be showcased in the lab in Tokyo. So, it's good to know the details of everyone's functional devstacks, eh. :) | 06:37 |
| sbalukoff | So far, this isn't looking too bad. I'm guessing the docs will be fairly short. Which is a good thing, obviously. | 06:38 |
| rm_work | yes | 06:38 |
| rm_work | That script I run has served me really well | 06:38 |
| sbalukoff | I'll probably spend more time on documenting the current component configuration (so people understand all the bits and pieces of Octavia) than I will have to on actually spinning up a stack. | 06:38 |
| rm_work | there's an official method to use for debugging dsvm gates which is in the README for one of the infra projects | 06:38 |
| sbalukoff | Yeah, it looks handy, eh! | 06:38 |
| rm_work | devstack-gate possibly | 06:38 |
| rm_work | but that one is ... let's just say I was possibly the first person to actually TRY to use it | 06:39 |
| rm_work | it works, kinda, with some tweaks | 06:39 |
| sbalukoff | HAH! | 06:39 |
| rm_work | but also takes forever to spin up and if you run it on RAX images it locks you out of them permanently :P so lulz | 06:39 |
| sbalukoff | I've got a couple minor tweaks to the barbican devstack stuff that smoothed things over a bit for my attempts to get it running. I'll probably make a CR of that tomorrow. | 06:40 |
| rm_work | (by default it flips ssh to "no-root" and RAX images don't set up the default "ubuntu" user) | 06:40 |
| sbalukoff | Right now, I'm just watching this devstack build churn and watching a Dr. Who marathon as I nurse a cold. | 06:40 |
| rm_work | sbalukoff: ah, anything of note? | 06:40 |
| rm_work | other than, python-barbicanclient is bugged for dealing with auth | 06:40 |
| *** numans has quit IRC | 06:40 | |
| sbalukoff | rm_work: Just some assumptions made about roles and accounts that fail if they already exist. I'm adding checks to avoid that noise. It's pretty minor, actually. | 06:40 |
| sbalukoff | Yeah.... I notice the CLI was... special. | 06:41 |
| sbalukoff | noticed. | 06:41 |
| sbalukoff | Anyway. | 06:41 |
| rm_work | https://review.openstack.org/#/c/220459/ | 06:43 |
| rm_work | not getting traction | 06:43 |
| rm_work | but it's a minor annoyance | 06:44 |
| rm_work | just pass in the keystone version as an arg | 06:44 |
| rm_work | instead of relying on the environment | 06:44 |
| rm_work | hmm, scenario tests still failing due to a LB stuck in PENDING for too long | 06:45 |
| rm_work | https://jenkins05.openstack.org/job/gate-neutron-lbaasv2-dsvm-scenario/30/console | 06:45 |
| rm_work | but whatever | 06:45 |
| rm_work | non-voting | 06:45 |
| *** evgenyf has joined #openstack-lbaas | 06:47 | |
| rm_work | ...... cmon | 06:50 |
| sbalukoff | That's the reason your patch above isn't passing? Or are you talking about an Octavia patch | 06:50 |
| rm_work | different context altogether, yeah | 06:51 |
| rm_work | the jenkins log is a neutron-lbaas scenario run | 06:51 |
| rm_work | for making Octavia the default | 06:51 |
| sbalukoff | Gotcha. | 06:52 |
| rm_work | it is so weird having other people be awake at the same time as me <_< | 06:53 |
| sbalukoff | I'm actually usually awake at this time. | 06:54 |
| sbalukoff | It's rare to see me online before noon. XD | 06:54 |
| rm_work | heh | 06:55 |
| *** manishg has joined #openstack-lbaas | 06:56 | |
| *** manishg has quit IRC | 07:01 | |
| *** Shijia__ has joined #openstack-lbaas | 07:02 | |
| *** Shijia_ has quit IRC | 07:03 | |
| *** Shijia_ has joined #openstack-lbaas | 07:04 | |
| *** apuimedo_ has joined #openstack-lbaas | 07:06 | |
| *** Shijia__ has quit IRC | 07:07 | |
| *** vjay10 has quit IRC | 07:10 | |
| *** vjay11 has joined #openstack-lbaas | 07:10 | |
| *** numans has joined #openstack-lbaas | 07:14 | |
| *** akamyshnikova has joined #openstack-lbaas | 07:16 | |
| *** chlong has quit IRC | 07:17 | |
| *** Shijia_ has quit IRC | 07:23 | |
| *** vjay12 has joined #openstack-lbaas | 07:24 | |
| *** vjay11 has quit IRC | 07:25 | |
| *** Shijia has joined #openstack-lbaas | 07:26 | |
| *** Varun_Lodaya has quit IRC | 07:32 | |
| *** Shijia_ has joined #openstack-lbaas | 07:36 | |
| *** Shijia has quit IRC | 07:40 | |
| rm_work | these are taking ... a while | 07:42 |
| rm_work | i really hope they don't STILL timeout | 07:42 |
| rm_work | only one left: https://jenkins04.openstack.org/job/gate-neutron-lbaasv2-dsvm-pool/1/console | 07:56 |
| rm_work | crap, pools might spin too many VMs still <_< | 07:57 |
| *** manishg has joined #openstack-lbaas | 07:57 | |
| *** manishg has quit IRC | 08:02 | |
| rm_work | fff | 08:10 |
| rm_work | timed out | 08:10 |
| rm_work | rechecking one more time | 08:10 |
| *** apuimedo has quit IRC | 08:18 | |
| *** vjay12 has quit IRC | 08:21 | |
| *** apuimedo_ has quit IRC | 08:22 | |
| *** apuimedo has joined #openstack-lbaas | 08:22 | |
| *** vjay12 has joined #openstack-lbaas | 08:29 | |
| *** chlong has joined #openstack-lbaas | 08:43 | |
| *** Shijia has joined #openstack-lbaas | 08:47 | |
| openstackgerrit | Brandon Logan proposed openstack/neutron-lbaas: Use admin context in octavia driver, improve logs https://review.openstack.org/225542 | 08:48 |
| openstackgerrit | Brandon Logan proposed openstack/neutron-lbaas: Set Octavia as the reference LBaaSv2 driver https://review.openstack.org/223265 | 08:49 |
| *** Shijia_ has quit IRC | 08:49 | |
| *** Shijia has left #openstack-lbaas | 08:50 | |
| *** dougwig has quit IRC | 08:53 | |
| *** manishg has joined #openstack-lbaas | 08:58 | |
| *** manishg has quit IRC | 09:02 | |
| *** dougwig has joined #openstack-lbaas | 09:16 | |
| *** vjay12 has quit IRC | 09:37 | |
| *** evgenyf has quit IRC | 09:48 | |
| *** manishg has joined #openstack-lbaas | 09:59 | |
| *** manishg has quit IRC | 10:03 | |
| *** kiran-r has joined #openstack-lbaas | 10:03 | |
| *** evgenyf has joined #openstack-lbaas | 10:20 | |
| *** bharathm has quit IRC | 10:40 | |
| *** manishg has joined #openstack-lbaas | 10:59 | |
| *** manishg has quit IRC | 11:04 | |
| *** bradjones has joined #openstack-lbaas | 11:05 | |
| *** bradjones has quit IRC | 11:05 | |
| *** bradjones has joined #openstack-lbaas | 11:05 | |
| *** bradjones has quit IRC | 11:08 | |
| *** bradjones has joined #openstack-lbaas | 11:08 | |
| *** bradjones has quit IRC | 11:08 | |
| *** bradjones has joined #openstack-lbaas | 11:08 | |
| *** bradjones has quit IRC | 11:14 | |
| *** bradjones has joined #openstack-lbaas | 11:14 | |
| *** bradjones has quit IRC | 11:14 | |
| *** bradjones has joined #openstack-lbaas | 11:14 | |
| *** Guest68836 has quit IRC | 11:46 | |
| *** manishg has joined #openstack-lbaas | 12:00 | |
| *** manishg has quit IRC | 12:05 | |
| *** Brian has joined #openstack-lbaas | 12:32 | |
| *** Brian is now known as Guest84075 | 12:32 | |
| *** Guest84075 has quit IRC | 12:37 | |
| *** blallau has joined #openstack-lbaas | 12:49 | |
| *** blallau has quit IRC | 12:51 | |
| *** blallau has joined #openstack-lbaas | 12:59 | |
| *** blallau has quit IRC | 13:00 | |
| *** manishg has joined #openstack-lbaas | 13:01 | |
| *** blallau has joined #openstack-lbaas | 13:05 | |
| *** manishg has quit IRC | 13:05 | |
| *** Alex_Staf has quit IRC | 13:13 | |
| *** blallau has quit IRC | 13:20 | |
| *** blallau has joined #openstack-lbaas | 13:21 | |
| *** woodster_ has joined #openstack-lbaas | 13:33 | |
| *** bharathm has joined #openstack-lbaas | 13:41 | |
| *** bharathm has quit IRC | 13:45 | |
| *** blallau has quit IRC | 13:46 | |
| mestery | lbaas folks: Looks like https://blueprints.launchpad.net/neutron/+spec/neutron-flavor-framework-templates isn't going to make Liberty | 13:47 |
| mestery | I see 3 patches out, none wiht reviews yet other than the one from armax | 13:47 |
| mestery | I'm likely to bump this one to Mitaka at this point. | 13:47 |
| mestery | dougwig blogan xgerman: ^^^ | 13:47 |
| xgerman | Ok. Has been a long shot to begin with... | 13:48 |
| xgerman | And every time we did something the scope increased. | 13:50 |
| *** manishg has joined #openstack-lbaas | 14:02 | |
| *** manishg has quit IRC | 14:05 | |
| *** blallau has joined #openstack-lbaas | 14:08 | |
| *** Brian has joined #openstack-lbaas | 14:10 | |
| *** Brian is now known as Guest27392 | 14:10 | |
| *** vivek-ebay has joined #openstack-lbaas | 14:11 | |
| *** Guest27392 has quit IRC | 14:14 | |
| *** kiran-r has quit IRC | 14:15 | |
| *** vivek-ebay has quit IRC | 14:21 | |
| *** bharathm has joined #openstack-lbaas | 14:23 | |
| openstackgerrit | Monty Taylor proposed openstack/octavia: Change ignore-errors to ignore_errors https://review.openstack.org/225774 | 14:34 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/neutron-lbaas: Updated from global requirements https://review.openstack.org/225227 | 14:38 |
| *** denghui has quit IRC | 14:38 | |
| *** pck is now known as pckizer | 14:43 | |
| *** nmagnezi has quit IRC | 14:49 | |
| *** minwang2 has joined #openstack-lbaas | 14:57 | |
| *** albertom has quit IRC | 14:59 | |
| *** devlaps has joined #openstack-lbaas | 15:02 | |
| *** blallau has quit IRC | 15:03 | |
| *** albertom has joined #openstack-lbaas | 15:05 | |
| *** Brian has joined #openstack-lbaas | 15:11 | |
| *** Brian is now known as Guest82988 | 15:12 | |
| *** manishg has joined #openstack-lbaas | 15:15 | |
| *** Guest82988 has quit IRC | 15:17 | |
| *** devlaps has quit IRC | 15:19 | |
| *** ganeshna has joined #openstack-lbaas | 15:27 | |
| *** evgenyf has quit IRC | 15:28 | |
| *** devlaps has joined #openstack-lbaas | 15:28 | |
| *** Varun_Lodaya has joined #openstack-lbaas | 15:29 | |
| *** mlavalle has joined #openstack-lbaas | 15:30 | |
| *** blallau has joined #openstack-lbaas | 15:30 | |
| *** KunalGandhi has joined #openstack-lbaas | 15:35 | |
| johnsom | blogan Any thoughts on the sqlalchemy error: http://logs.openstack.org/65/223265/15/check/gate-neutron-lbaasv2-dsvm-loadbalancer/f7e4a90/logs/screen-q-svc.txt.gz | 15:36 |
| johnsom | It's near the bottom, search for | 15:36 |
| johnsom | RuntimeError: reentrant call inside <_io.BufferedReader name=18> | 15:36 |
| *** KunalGandhi has quit IRC | 15:39 | |
| *** manishg has quit IRC | 15:43 | |
| *** numans has quit IRC | 15:43 | |
| *** KunalGandhi has joined #openstack-lbaas | 15:46 | |
| *** manishg has joined #openstack-lbaas | 15:48 | |
| *** Varun_Lodaya has quit IRC | 15:56 | |
| *** eranra has quit IRC | 16:02 | |
| *** diogogmt has joined #openstack-lbaas | 16:02 | |
| *** alejandrito has joined #openstack-lbaas | 16:03 | |
| *** apuimedo has quit IRC | 16:06 | |
| *** blallau has quit IRC | 16:09 | |
| *** apuimedo has joined #openstack-lbaas | 16:13 | |
| *** minwang2 has quit IRC | 16:16 | |
| dougwig | morning all. is the octavia job stable, or did we recheck hammer it in? | 16:18 |
| xgerman | I think it’s still running | 16:18 |
| xgerman | mmh, it says one job failed: http://status.openstack.org/zuul/ | 16:21 |
| xgerman | not sure if that’s the one we skip | 16:21 |
| xgerman | rm_work did bologna take today off again? | 16:23 |
| xgerman | blogan | 16:23 |
| *** vivek-ebay has joined #openstack-lbaas | 16:32 | |
| *** vivek-ebay has quit IRC | 16:33 | |
| *** vivek-ebay has joined #openstack-lbaas | 16:34 | |
| *** minwang2 has joined #openstack-lbaas | 16:47 | |
| *** chlong has quit IRC | 16:50 | |
| *** mlavalle_ has joined #openstack-lbaas | 17:00 | |
| *** mlavalle has quit IRC | 17:01 | |
| *** chlong has joined #openstack-lbaas | 17:02 | |
| *** TrevorV|Home has joined #openstack-lbaas | 17:04 | |
| *** Brian has joined #openstack-lbaas | 17:07 | |
| *** Brian is now known as Guest55031 | 17:07 | |
| *** Guest55031 has quit IRC | 17:12 | |
| *** devlaps has quit IRC | 17:12 | |
| *** macjack has joined #openstack-lbaas | 17:19 | |
| *** macjack has left #openstack-lbaas | 17:21 | |
| *** bana_k has joined #openstack-lbaas | 17:22 | |
| mestery | xgerman blogan: https://review.openstack.org/#/c/223265/ is dieing in the gate again, looks another multi-hour recheck :( | 17:22 |
| xgerman | :-( | 17:23 |
| mestery | xgerman: Also, the fact you called blogan bologna is pure #awesomesauce | 17:23 |
| mestery | :) | 17:23 |
| blogan | johnsom: not sure on that one, i pushed up a patch that might give more insight and migth actually fix it, but not sure | 17:23 |
| blogan | mestery: grrrr | 17:23 |
| xgerman | bologna | 17:23 |
| blogan | mestery: xgerman always has awesome names for everyone, ptoohill is pothole for that reason | 17:23 |
| mestery | rofl | 17:24 |
| *** TrevorV|Home has quit IRC | 17:24 | |
| xgerman | it’s my bad typing combined with a spell checker | 17:24 |
| johnsom | I still prefer towgan | 17:24 |
| mestery | lol | 17:24 |
| pc-pothole | :) | 17:25 |
| xgerman | blogan is your fix being pulled? | 17:29 |
| blogan | xgerman: https://review.openstack.org/#/c/225542/ | 17:30 |
| blogan | xgerman: no, but i dont have hopes it'll fix the problem, there is a possibility though | 17:30 |
| blogan | xgerman: it'll at least give better logging for troubleshooting | 17:30 |
| blogan | xgerman: though this error seems quite odd | 17:31 |
| xgerman | yeah, this whole sqlalchemy py-something is odd — we should have used JDBC | 17:31 |
| blogan | well every orm is going to have some hard to troubleshoot bugs, sqlalchemy is well known in openstack | 17:32 |
| blogan | seemed like a no brainer | 17:32 |
| xgerman | anyhow… so we ned +2s pon that one | 17:33 |
| xgerman | dougwig... | 17:34 |
| dougwig | looking | 17:34 |
| *** manishg has quit IRC | 17:35 | |
| xgerman | but probably won’t make the gate… | 17:35 |
| xgerman | ? | 17:35 |
| blogan | pc-pothole: https://review.openstack.org/#/c/225542 | 17:35 |
| xgerman | ah, we still have harpy in the gate.. so we should be good | 17:35 |
| xgerman | .. | 17:35 |
| *** manishg has joined #openstack-lbaas | 17:36 | |
| *** manishg has quit IRC | 17:36 | |
| blogan | xgerman: what? | 17:36 |
| *** manishg has joined #openstack-lbaas | 17:37 | |
| xgerman | I was thinking we couldn’t get that merged since our gate jobs fail but then remembered we were still running the namespace driver | 17:37 |
| *** manishg has quit IRC | 17:37 | |
| *** manishg has joined #openstack-lbaas | 17:37 | |
| xgerman | som it should be good | 17:37 |
| *** crc32 has joined #openstack-lbaas | 17:45 | |
| *** harlowja has joined #openstack-lbaas | 17:46 | |
| *** vjay10 has joined #openstack-lbaas | 17:46 | |
| *** ganeshna has quit IRC | 17:57 | |
| *** ajmiller has joined #openstack-lbaas | 18:00 | |
| *** KunalGandhi has quit IRC | 18:01 | |
| openstackgerrit | Phillip Toohill proposed openstack/octavia: Updates for containers functionality https://review.openstack.org/199954 | 18:06 |
| *** ajmiller has quit IRC | 18:09 | |
| *** vjay10 has quit IRC | 18:11 | |
| *** vivek-ebay has quit IRC | 18:14 | |
| *** KunalGandhi has joined #openstack-lbaas | 18:19 | |
| *** Varun_Lodaya has joined #openstack-lbaas | 18:28 | |
| *** manishg has quit IRC | 18:35 | |
| johnsom | The last round in the merge gate seems to have failed because we timed out attempting to remove the security groups. They were still in use. | 18:36 |
| johnsom | What do we think about increasing that retry count? | 18:36 |
| *** manishg has joined #openstack-lbaas | 18:41 | |
| *** blallau has joined #openstack-lbaas | 18:43 | |
| *** blallau has quit IRC | 18:44 | |
| *** blallau has joined #openstack-lbaas | 18:45 | |
| *** vivek-ebay has joined #openstack-lbaas | 18:45 | |
| *** blallau has quit IRC | 18:46 | |
| *** blallau has joined #openstack-lbaas | 18:46 | |
| *** minwang2 has quit IRC | 18:47 | |
| *** manishg has quit IRC | 19:00 | |
| *** manishg has joined #openstack-lbaas | 19:01 | |
| *** barclaac|2 has joined #openstack-lbaas | 19:01 | |
| blogan | johnsom: probably makes sense | 19:02 |
| *** manishg has quit IRC | 19:02 | |
| blogan | johnsom: triple it to make it more successful, haevnt run into that yet, but glad i made it configurable | 19:02 |
| *** manishg has joined #openstack-lbaas | 19:03 | |
| *** barclaac has quit IRC | 19:04 | |
| *** bradjones has quit IRC | 19:04 | |
| *** blallau has quit IRC | 19:09 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia: Increase the number of network call retries https://review.openstack.org/225989 | 19:11 |
| blogan | johnsom, xgerman: is that controller logging review somethign we want to get in today? | 19:11 |
| johnsom | xgerman blogan rm_work https://review.openstack.org/225989 | 19:11 |
| xgerman | blogan — I don’t think it’s critical but I gave it +2 for low risk (logging review) | 19:12 |
| blogan | johnsom: you ahve the .conf file have max_retries = 10, but the config.py says default is 15 | 19:13 |
| blogan | i prefer the 15 btw | 19:13 |
| xgerman | misdirection... | 19:13 |
| openstackgerrit | Michael Johnson proposed openstack/octavia: Increase the number of network call retries https://review.openstack.org/225989 | 19:13 |
| blogan | these are not the max retries you are looking fo | 19:13 |
| blogan | for | 19:13 |
| johnsom | Sorry, I was toying with 20, but didn't fully undo that change | 19:13 |
| johnsom | fixed ^^ | 19:13 |
| *** vivek-ebay has quit IRC | 19:15 | |
| blogan | i do worry about getting rate limited, but its a minor concern, and usually 1/s isn't terrible | 19:17 |
| xgerman | done + done | 19:17 |
| openstackgerrit | Merged openstack/octavia: Increase the number of network call retries https://review.openstack.org/225989 | 19:28 |
| *** minwang2 has joined #openstack-lbaas | 19:29 | |
| blogan | that was quick | 19:29 |
| xgerman | blogan recheck? | 19:35 |
| blogan | xgerman: like to get that other one merged first | 19:36 |
| blogan | the one in neutron-lbaas | 19:36 |
| xgerman | k, I will let you do the honors | 19:36 |
| blogan | ajmiller still on vacation? and johnsom doesn't have +2 honors yet right? | 19:36 |
| johnsom | For neutron-lbaas yes | 19:37 |
| openstackgerrit | Merged openstack/octavia: Update controller logging https://review.openstack.org/225543 | 19:39 |
| *** blallau has joined #openstack-lbaas | 19:45 | |
| blogan | johnsom: yes you dont? or yes you do? | 19:46 |
| blogan | lol | 19:46 |
| johnsom | Yes, ajmiller is on vacation. No, I don't have the +2 yets | 19:47 |
| blogan | i asked it in one of those awkward negative ways | 19:47 |
| *** vivek-ebay has joined #openstack-lbaas | 19:47 | |
| *** blallau has quit IRC | 20:04 | |
| *** blallau has joined #openstack-lbaas | 20:05 | |
| *** manishg has quit IRC | 20:06 | |
| xgerman | just assign pothole a number ron your speed dial | 20:10 |
| *** crc32 has quit IRC | 20:10 | |
| *** vivek-ebay has quit IRC | 20:15 | |
| *** barclaac|2 has quit IRC | 20:15 | |
| rm_work | ugh, last night it LOOKED like all were passing except pools timed out | 20:16 |
| rm_work | but now there's errors too? | 20:16 |
| *** manishg has joined #openstack-lbaas | 20:16 | |
| *** blallau has quit IRC | 20:24 | |
| blogan | rm_work: welcome to the day! | 20:25 |
| rm_work | lol | 20:25 |
| rm_work | also -- anyone know how LIBS_FROM_GIT works? | 20:25 |
| johnsom | I called it pretty close, around noon time | 20:25 |
| *** blallau has joined #openstack-lbaas | 20:25 | |
| blogan | rm_work: doesn't that mean use git to install the clients isntead of pip? | 20:25 |
| rm_work | so it will be on master? | 20:26 |
| johnsom | yes it works on master | 20:26 |
| rm_work | I assume so | 20:26 |
| rm_work | well, i will probably ask again later when we aren't scrambling to get our shit working | 20:27 |
| johnsom | blogan - Can I give you a proxy +2 +A? It seems like we are stuck | 20:27 |
| rm_work | because it is related to a barbican issue | 20:27 |
| xgerman | johnsom you can | 20:27 |
| xgerman | mestery saved us!! | 20:27 |
| blogan | johnsom: got mestery to do it | 20:27 |
| xgerman | what about dougwig? slacking? | 20:28 |
| mestery | blogan: This one is still being a pesky little thing: https://review.openstack.org/#/c/223265/ | 20:28 |
| mestery | xgerman johnsom: ^^^^ :( | 20:28 |
| mestery | Need that to merge today yet | 20:28 |
| *** blallau has quit IRC | 20:28 | |
| blogan | mestery: yeah we're hoping that review you just +A'ed will help out on that | 20:28 |
| mestery | Are those failures real? | 20:28 |
| mestery | blogan: Cool! | 20:28 |
| johnsom | Yep | 20:28 |
| mestery | blogan: OK, so lets wait to recheck until that one merges? | 20:28 |
| xgerman | yep | 20:28 |
| mestery | nice | 20:28 |
| johnsom | And some more timeout adjustments that are already merged | 20:28 |
| mestery | OK, I'll defer to you folks to recheck that bad boy | 20:28 |
| blogan | mestery: they're one of them failures that doesnt happen all the time it seems, hoping the context change will fix it | 20:28 |
| mestery | cool | 20:29 |
| mestery | cool | 20:29 |
| rm_work | wtf last night loadbalancers passed but pools timed out | 20:29 |
| xgerman | yeah, it’s very random | 20:29 |
| xgerman | we *need* more stable code | 20:29 |
| blogan | and loadbalancer is almost taking the entire 2 hr limit | 20:29 |
| xgerman | and faster gates | 20:29 |
| mestery | yikes | 20:29 |
| mestery | yeah | 20:29 |
| mestery | faster == goodness | 20:29 |
| rm_work | erg | 20:30 |
| rm_work | you guys got this handled? ping me if you need anything? I REALLY have to work on some internal stuff today | 20:30 |
| blogan | btw this weekend i tried to get teh tests to only spin up one lb to test for, taht doesn't work so well when each test class creates its own user | 20:30 |
| blogan | so i hit a point where it was just going ot be a huge change | 20:30 |
| johnsom | Yeah, we REALLY need to get nested virtualization turned on | 20:30 |
| *** blallau has joined #openstack-lbaas | 20:30 | |
| rm_work | johnsom: i have no idea if that's even feasible -- i poked at some people and there is some interest internally, but ... it's not something we can lightly do for our whole public cloud | 20:31 |
| blogan | containers! | 20:31 |
| blogan | we need a magnum compute driver, but that'll ahve to happen in M | 20:31 |
| xgerman | +1 | 20:32 |
| *** crc32 has joined #openstack-lbaas | 20:32 | |
| *** blallau has quit IRC | 20:35 | |
| *** vivek-ebay has joined #openstack-lbaas | 20:57 | |
| openstackgerrit | Phillip Toohill proposed openstack/octavia: Updates for containers functionality https://review.openstack.org/199954 | 21:05 |
| *** barclaac|2 has joined #openstack-lbaas | 21:13 | |
| *** vivek-ebay has quit IRC | 21:21 | |
| *** madhu_ak has joined #openstack-lbaas | 21:25 | |
| madhu_ak | #opensatck-rally | 21:25 |
| madhu_ak | sorry | 21:25 |
| openstackgerrit | Phillip Toohill proposed openstack/octavia: Updates for containers functionality https://review.openstack.org/199954 | 21:30 |
| dougwig | let's chat about where octavia is at (blogan mestery johnsom xgerman rm_work others) | 21:37 |
| johnsom | o/ | 21:37 |
| xgerman | o/ | 21:37 |
| dougwig | things are moving furiously, which is great. are we certain we have stable jobs, and we're not just hammer in intermittent issues via recheck? | 21:37 |
| dougwig | i've been out sick today, so haven't been watching closely. | 21:37 |
| rm_work | o/ | 21:38 |
| johnsom | Current state is we are waiting for this to merge: https://review.openstack.org/#/c/225542/ | 21:38 |
| rm_work | dougwig: i don't think we are certain of that | 21:38 |
| rm_work | YET | 21:38 |
| rm_work | yeah | 21:38 |
| johnsom | Then, combined with some additional timeout increases that have merged we want to do another recheck | 21:38 |
| rm_work | with that things might be clearer | 21:38 |
| * mestery hops in | 21:38 | |
| mestery | Hey. dougwig! blogan thought maybe a bear got you ;) | 21:39 |
| xgerman | so it might be a night of rechecks until the gate submits | 21:39 |
| dougwig | no, something much smaller, and virus-like. | 21:39 |
| *** vivek-ebay has joined #openstack-lbaas | 21:39 | |
| mestery | lol | 21:39 |
| dougwig | what we *really* want to avoid is getting something in via recheck, only to have it intermittent and thus make the general neutron gate non-deterministic. that'd be BAD. | 21:39 |
| xgerman | indeed | 21:39 |
| mestery | not just bad | 21:40 |
| mestery | REALLY REALLY bad | 21:40 |
| johnsom | Agreed | 21:40 |
| mestery | Catastrophically bad | 21:40 |
| mestery | Thus my concern over the rechecks so far | 21:40 |
| dougwig | i want to push to get octavia in L, but we also need an honest assessment here of if we have enough time remaining to get to some level of comfort in its stability. | 21:40 |
| rm_work | I thought the rechecks last night were due to timing | 21:40 |
| rm_work | but even if they are, that is probably bad | 21:40 |
| rm_work | but it looks like we're seeing other stuff too? | 21:40 |
| johnsom | I worry that the variability in performance on the gate instances will impact our test stability | 21:40 |
| mestery | ++ | 21:41 |
| xgerman | ++ | 21:41 |
| rm_work | possibly the two long-running ones (load_balancers and pools) could have a test or two clipped out? | 21:41 |
| rm_work | not a great long-term option, but' | 21:41 |
| johnsom | blogan took a chainsaw to pools already | 21:41 |
| dougwig | let's also consider the impact that bumping neutron's check and gate queues to 2 hours each is going to cause. | 21:41 |
| mestery | Hint: It won't make people happy | 21:42 |
| mestery | :) | 21:42 |
| dougwig | (yes, i'm being the devil's advocate now.) | 21:42 |
| xgerman | people are unhappy anyway :-) | 21:42 |
| mestery | dougwig: It's a fair point | 21:42 |
| dougwig | mestery: what's the remaining release timeline look like? how much wiggle room do we have? | 21:42 |
| mestery | So, the options are the 2 hour run time, or breaking it down futther and using more instances per test run? | 21:42 |
| mestery | dougwig: We | 21:43 |
| mestery | dougwig: We're cutting RC1 Wednesday | 21:43 |
| mestery | And if it isn't in then | 21:43 |
| mestery | It's out | 21:43 |
| mestery | :( | 21:43 |
| dougwig | mestery: break it down, make one that is semi fast (a 'smoke' if you will), and only putting the fast one in the neutron queues. | 21:43 |
| xgerman | dougwig +1 | 21:43 |
| mestery | dougwig: Ack | 21:43 |
| xgerman | and mid term get some 3rd party CI which can run on VT_X and does the checks | 21:43 |
| dougwig | mestery: will there be an RC2 ? | 21:43 |
| xgerman | there always is | 21:44 |
| mestery | dougwig: Our track record indicates yes! | 21:44 |
| dougwig | mestery: if we get it sort in by RC1, can we use until RC2 to get it stable? or is that playing too fast and loose? | 21:44 |
| dougwig | /sort in/sort of in/ | 21:44 |
| * mestery thinks | 21:45 | |
| mestery | IMHO, it's playing fast and loose, and I know you folks don't want to hear that :( | 21:45 |
| mestery | How confident are we the current patches blogan has will make the races better? | 21:45 |
| *** vivek-ebay has quit IRC | 21:46 | |
| xgerman | we have a lot of confidence in the timeput | 21:46 |
| dougwig | overall, how confident are folks that we are ready for L (not want, but is ready) ? i get the feeling that we need another 2-3 weeks here, as much as i don't want to say that. | 21:46 |
| johnsom | Yeah, I have had good success local with the new timeouts. I can't talk to the change blogan made. | 21:46 |
| dougwig | lots of crickets. :) | 21:48 |
| mestery | So, seems like perhaps this sobering discussion is leading us to the fact Octavia may not make Liberty? | 21:48 |
| mestery | dougwig: ^^^^ ??? | 21:48 |
| xgerman | well, let’s say we pass the gate | 21:48 |
| dougwig | mestery: i think it's on the knife edge, and i'd like the folks that have been busting ass the most to weigh in, because i want to find a way to have it make it. | 21:48 |
| mestery | dougwig: Ack | 21:48 |
| mestery | xgerman: Continue please | 21:48 |
| mestery | If we pass the gate ... then what? | 21:48 |
| xgerman | so we know Octavia itself is pretty table since RAX is going live with it soon | 21:48 |
| mestery | Yup | 21:49 |
| xgerman | so our main problem is with the driver | 21:49 |
| *** barclaac has joined #openstack-lbaas | 21:49 | |
| xgerman | I think if we pass the gate we should be at least as good/bad as the namespace driver | 21:49 |
| dougwig | xgerman: i think our main problem is with running VMs inside VMs in the gate, right? | 21:49 |
| xgerman | gate times we need to address though | 21:49 |
| johnsom | The nested virt issue is my biggest concern. qemu gets so slow... | 21:50 |
| rm_work | yeah there is already the framework in place for the neutron gate only running a "minimal" set | 21:50 |
| rm_work | we just need to define that | 21:50 |
| johnsom | rm_work +1 | 21:50 |
| xgerman | rm_work +! | 21:50 |
| xgerman | and we can put the heavy lifting into third party on bare metal | 21:50 |
| rm_work | yes | 21:50 |
| xgerman | and I think that’s independent of the release date | 21:51 |
| pc-pothole | I would agree with the statement about being as good/bad as namespace. I've got things set up in a lab and have been testing things with good outcomes. | 21:51 |
| rm_work | not sure how long that will take to set up though | 21:51 |
| dougwig | forgive my meetbot abuse here. | 21:51 |
| dougwig | #startmeeting neutron lbaas | 21:51 |
| openstack | Meeting started Mon Sep 21 21:51:40 2015 UTC and is due to finish in 60 minutes. The chair is dougwig. Information about MeetBot at http://wiki.debian.org/MeetBot. | 21:51 |
| openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 21:51 |
| rm_work | are we voting? :P\ | 21:51 |
| openstack | The meeting name has been set to 'neutron_lbaas' | 21:51 |
| dougwig | #startvote is octavia-as-ref close enough to go for RC1? Yes No | 21:51 |
| openstack | Begin voting on: is octavia-as-ref close enough to go for RC1? Valid vote options are Yes, No. | 21:51 |
| openstack | Vote using '#vote OPTION'. Only your last vote counts. | 21:51 |
| xgerman | #vote Yes | 21:52 |
| rm_work | so does this have anything to do with the testing times? | 21:52 |
| pc-pothole | #vote Yes | 21:52 |
| sbalukoff | #vote Yes | 21:52 |
| rm_work | or JUST the quality of the service | 21:52 |
| dougwig | rm_work: yep, it's all part of it. | 21:52 |
| *** barclaac|2 has quit IRC | 21:52 | |
| pc-pothole | eh, im a bit unsure about the testing times | 21:52 |
| dougwig | we have to sort out test times, get a partial into the neutron queues, swap the default in the conf, etc. | 21:52 |
| mestery | #vote No | 21:52 |
| rm_work | I don't know if we can solve the test time issues by Wednesday without just taking a chainsaw to them and not testing | 21:52 |
| rm_work | the split was ... the big hope | 21:53 |
| rm_work | but it isn't even enough apparently | 21:53 |
| dougwig | lbaas v2 does have about 5x as many tests as v1, btw. | 21:53 |
| johnsom | #vote No | 21:53 |
| dougwig | rm_work: haven't seen you vote yet. :) | 21:53 |
| johnsom | Yes, we have a ton of good tests. | 21:53 |
| mestery | dougwig: you either | 21:53 |
| rm_work | that's because i haven't voted yet | 21:53 |
| pc-pothole | changing to #vote No because of tests | 21:53 |
| johnsom | That is what is slowing us down so much | 21:54 |
| rm_work | yeah so | 21:54 |
| rm_work | if we move to test parity with v1 | 21:54 |
| rm_work | IE, delete 85% of our tests | 21:54 |
| rm_work | we could go :P | 21:54 |
| sbalukoff | Then we pass | 21:54 |
| pc-pothole | :) | 21:54 |
| dougwig | #endvote | 21:54 |
| openstack | Voted on "is octavia-as-ref close enough to go for RC1?" Results are | 21:54 |
| openstack | Yes (3): xgerman, sbalukoff, pc-pothole | 21:54 |
| openstack | No (2): mestery, johnsom | 21:54 |
| dougwig | #endmeeting | 21:54 |
| openstack | Meeting ended Mon Sep 21 21:54:31 2015 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 21:54 |
| openstack | Minutes: http://eavesdrop.openstack.org/meetings/neutron_lbaas/2015/neutron_lbaas.2015-09-21-21.51.html | 21:54 |
| openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/neutron_lbaas/2015/neutron_lbaas.2015-09-21-21.51.txt | 21:54 |
| openstack | Log: http://eavesdrop.openstack.org/meetings/neutron_lbaas/2015/neutron_lbaas.2015-09-21-21.51.log.html | 21:54 |
| rm_work | well ptoohill changed his vote non-bindingly, and i was about to vote No (I think), for what it's worth | 21:55 |
| rm_work | what was your vote dougwig? | 21:55 |
| sbalukoff | Honestly, we'll go with Octavia as soon as we think it's capable, either way. I'm hoping sooner rather than later-- I saw an internal product built off v1 recently, and I've been giving that group a (deservedly) hard time about it. | 21:55 |
| dougwig | from everything i've heard, it's no, i just really really don't want it to be. | 21:55 |
| rm_work | same | 21:56 |
| rm_work | the testing is frustrating | 21:56 |
| dougwig | sbalukoff: i bet you could make a v1 octavia driver, too. :) | 21:56 |
| johnsom | dougwig same here | 21:56 |
| rm_work | because we are holding ourselves to a higher standard than v1 | 21:56 |
| rm_work | and that is screwing us | 21:56 |
| xgerman | well, most projects don’t have tests to begin with | 21:56 |
| sbalukoff | dougwig: Careful. I have access to matches and I hear you're flammable. | 21:56 |
| pc-pothole | :/ | 21:56 |
| dougwig | well, what's the impact if it misses L? the operators here can move forward anyway. devstack will still switch, likely before the summit. it's mostly around messaging and adoption from the smaller ops? | 21:57 |
| rm_work | if we lowered our testing standard we'd be ok | 21:57 |
| sbalukoff | rm_work: In some ways it screws us, in others... well, at least we won't be getting people excited about something we don't feel is ready, right? | 21:57 |
| *** manishg has quit IRC | 21:57 | |
| xgerman | dougwig — we like to deprecate V1 | 21:57 |
| rm_work | yeah | 21:57 |
| rm_work | dougwig: yeah i think deprecation is the big step | 21:57 |
| rm_work | though | 21:57 |
| rm_work | we can still do that without Octavia | 21:57 |
| rm_work | right? | 21:57 |
| pc-pothole | need the new ref | 21:57 |
| rm_work | we just can't deprecate the namespace driver | 21:57 |
| xgerman | well, then you need to support namespace for another year | 21:57 |
| rm_work | right | 21:58 |
| rm_work | we can deprecate lbaasv1 but not namespace | 21:58 |
| xgerman | (if you make v2 non-experimental) | 21:58 |
| rm_work | because lbaasv2 is *working* | 21:58 |
| rm_work | just not with Octavia | 21:58 |
| xgerman | yes, and i don’t like to have to deal with namespace for another year | 21:58 |
| rm_work | yeah personally i don't have a huge problem with the namespace driver sticking around, but I'm not the one who has to maintain it so far :P | 21:59 |
| sbalukoff | It don't think it's a terrible idea to carry the namespace driver for another year. Octavia is far more complicated than the namespace driver, and as long as we get people off v1... well, from a tenant perspective at least, going from v2 to Octavia is simple. | 21:59 |
| rm_work | sbalukoff +1 | 21:59 |
| rm_work | i do worry some about having a super complex system as the reference | 21:59 |
| dougwig | it's another cycle, not another year. :) | 21:59 |
| xgerman | well, we would need to implement everything in two places… | 21:59 |
| rm_work | when the idea of a reference is that it is simple and easy to understand/deploy | 21:59 |
| xgerman | not everybody has bottomless development resources like IBM | 22:00 |
| rm_work | right, one ADDITIONAL cycle | 22:00 |
| dougwig | easier to go from namespace v2 to octavia v2, since the providers can run in parallel. more of a hassle for v1 to v2, btw. | 22:00 |
| rm_work | technically we have to keep it for one year anyway | 22:00 |
| sbalukoff | xgerman: True. Any additions to v2, yes. | 22:00 |
| rm_work | now we need it for 1.5 years | 22:00 |
| sbalukoff | So that means... what... L7? | 22:00 |
| rm_work | because one additional cycle until deprecation | 22:00 |
| rm_work | right? | 22:00 |
| sbalukoff | Because we don't have to implement HA or anything not in the API in the namespace driver. | 22:00 |
| sbalukoff | And I basically had the code "close" on L7 before mestery abandoned that CR. | 22:01 |
| * sbalukoff stares at mestery | 22:01 | |
| sbalukoff | ;) | 22:01 |
| * rm_work glares at mestery | 22:01 | |
| johnsom | We can still cut Octavia 0.5 and ship it in the Liberty cycle right? It's just making it the reference driver we are considering at risk I think. | 22:01 |
| rm_work | right | 22:01 |
| rm_work | literally all it costs us is one cycle extra on the namespace driver support | 22:01 |
| xgerman | well, al I heard was we have until Wednesday | 22:01 |
| sbalukoff | Let's plan on that. | 22:01 |
| rm_work | true | 22:01 |
| sbalukoff | v0.5 has lost a few features in the interest of trying to the the "as ref driver" done. | 22:02 |
| rm_work | so we can still try to fix it if we want? | 22:02 |
| xgerman | whole operating systems have been programmed in 48 hours | 22:02 |
| rm_work | but | 22:02 |
| sbalukoff | But these move easily to v1.0 | 22:02 |
| rm_work | we won't freak out if we can't | 22:02 |
| *** barclaac|2 has joined #openstack-lbaas | 22:02 | |
| dougwig | we won't be able to fit other things in if octavia-as-ref slips. we're already well past FF here. but it'd be a pretty good place to be in l7 + horizon + octavia as ref were all in at the very start of mitaka, giving all of them a cycle to harden up. i'd prefer to get oar (octavia-as-ref) in now, but. | 22:02 |
| johnsom | sbalukoff From what I remember, we hit 0.5 | 22:02 |
| dougwig | johnsom: yes to the driver. | 22:03 |
| sbalukoff | johnsom: https://wiki.openstack.org/wiki/Octavia/Roadmap L7 and HA used to be a part of 0.5 | 22:03 |
| johnsom | sbalukoff Revisionist's history! | 22:03 |
| xgerman | +1 | 22:04 |
| dougwig | the beauty of wiki's. | 22:04 |
| sbalukoff | johnsom: Actually this was in the ROADMAN.rst file since... about the beginning. ;) | 22:04 |
| sbalukoff | ROADMAP.rst | 22:04 |
| xgerman | anyhow, let’s focus | 22:04 |
| sbalukoff | Right. | 22:04 |
| johnsom | sbalukoff I thought we had HA in 1.0 | 22:04 |
| sbalukoff | johnsom: HA for the control layer. | 22:04 |
| *** vivek-ebay has joined #openstack-lbaas | 22:05 | |
| sbalukoff | HA for the service delivery was supposed to be part of 0.5 | 22:05 |
| dougwig | i think we have two options here: 1) call this missing L now, and deal with it, or 2) continue our push until tomorrow night, at which point we make a decision. i'm game for either. i'm a little leery of #2, just because I worry about de-stabilizing the gate in general. | 22:05 |
| *** barclaac has quit IRC | 22:05 | |
| *** alejandrito has quit IRC | 22:05 | |
| xgerman | yeah, I would like to try the gate | 22:06 |
| sbalukoff | Well... let's assume we go with #1. What else do we want in by RC1? Anything else that needs to be shored up? | 22:06 |
| xgerman | 48 hours/ | 22:06 |
| xgerman | ? | 22:06 |
| johnsom | rm_work the gate change you did for upstream neutron is just the lb tests? | 22:06 |
| sbalukoff | xgerman: I'm not against it, eh. | 22:06 |
| *** manishg has joined #openstack-lbaas | 22:06 | |
| rm_work | johnsom: yes | 22:06 |
| sbalukoff | Note that I do need to run off in about 5 minutes. | 22:07 |
| xgerman | well, I think right now we have WYSIWYG | 22:07 |
| xgerman | I don’t think we can put more features in L | 22:07 |
| sbalukoff | Right. The next obvious choice would be L7. But we deprecated work on that to get OAR done. | 22:07 |
| sbalukoff | :P | 22:07 |
| rm_work | https://github.com/openstack/neutron-lbaas/blob/master/neutron_lbaas/tests/contrib/post_test_hook.sh#L22 | 22:07 |
| rm_work | "minimal" is what the neutron gate runs now | 22:08 |
| johnsom | We cannot merge anything new. That would be insane. | 22:08 |
| sbalukoff | If y'all wanted to do pool sharing... well, *I* think my neutron-lbaas patch is ready. ;) | 22:08 |
| rm_work | lol | 22:08 |
| sbalukoff | johnsom: I know. | 22:08 |
| sbalukoff | I guess I'm asking: Any significant bugfixes we want to get in before Wednesday? | 22:09 |
| johnsom | So, the upstream neutron gate for lbaas would run about an hour and a half to two hours as it sits. | 22:09 |
| johnsom | dougwig is that in line with other gate tests that need to run or is that too much? | 22:10 |
| dougwig | johnsom: for neutron-lbaas, we can deal with that while we sort it out. for neutron in general, which runs our api jobs, that'd lead to a palace revolt. | 22:10 |
| johnsom | Yeah, that was my question | 22:10 |
| dougwig | johnsom: current timings: https://review.openstack.org/#/c/224634/ | 22:10 |
| dougwig | grenade at 52m is the longest. | 22:10 |
| xgerman | vpnaas disabled their gate for a while | 22:10 |
| johnsom | So, how fast can we cherry pick a new subset for the neutron gate? rm_work thoughts? | 22:11 |
| rm_work | no idea | 22:11 |
| rm_work | i don't know what we'd want | 22:12 |
| rm_work | I just did the gate, ya'll who wrote the tests probably need to decide that :P | 22:12 |
| rm_work | basically i'd say | 22:12 |
| rm_work | we have 20m to spin up devstack | 22:12 |
| johnsom | On a good day | 22:12 |
| rm_work | which leaves about 30m of wiggle | 22:12 |
| blogan | good god i missed a lot of conversation | 22:13 |
| rm_work | that's... like 4 tests | 22:13 |
| blogan | looks like some major decisions were made | 22:13 |
| rm_work | so we could run FOUR tests | 22:13 |
| dougwig | blogan: please scroll back and weigh in. | 22:13 |
| rm_work | pick your favorite 4 | 22:13 |
| rm_work | 5 would push it to JUST under 60m on an average run, so that might be ok | 22:13 |
| rm_work | when i say 4-5 tests, i mean test classes | 22:14 |
| xgerman | yeah, I think we just pick some and then start working on some 3rd party CI with more tests and/or container | 22:14 |
| rm_work | also i saw a few tests that are set to be skipped currently | 22:16 |
| xgerman | yeah, tests are a mess | 22:16 |
| rm_work | if there's any test classes that have ONLY skipped tests, those are wasting a ton of time for no reason | 22:16 |
| rm_work | don't know if that is true | 22:16 |
| rm_work | but if it is, it'll spin up the VM before it notices that all tests are skipped | 22:16 |
| xgerman | I have some distracting meeting | 22:17 |
| blogan | yeah going through the list | 22:17 |
| sbalukoff | Ok, I've got to run off. Will catch up when I'm back. | 22:18 |
| blogan | dougwig: so my vote would be No its nto ready, bc of the tests | 22:18 |
| xgerman | most projects don’;t have tests so I would;’t hold that against us | 22:18 |
| xgerman | and the tests even pass but take too long | 22:19 |
| blogan | dougwig: also, keeping the namespace driver around has some pros, its easier to implement features in that than octavia | 22:19 |
| blogan | for one | 22:19 |
| blogan | xgerman: its mroe the length of time it takes for them to run and how that affects neutron (and us too) | 22:19 |
| blogan | so keeping the namespace driver around for a deprecation cycle may actually make us appreciate its simplicity more and it being the ref impl, i'm trying to be optimistic here | 22:21 |
| blogan | is the big thing about OAR that the namespace driver can just be removed? | 22:28 |
| xgerman | yep | 22:29 |
| xgerman | and I don;t hnave to argue with my product management about including Octavia | 22:30 |
| blogan | xgerman: and the big thing about not having the namespace driver around is not having to impelement new features twice? | 22:30 |
| xgerman | yep | 22:30 |
| blogan | xgerman: including octavia in what? | 22:30 |
| xgerman | our internal stuff | 22:30 |
| xgerman | it’s HP politics | 22:30 |
| blogan | xgerman: do yall plan to use Octavia internally before M release? | 22:31 |
| xgerman | yep | 22:31 |
| blogan | xgerman: i know its politics, but isn't that reason enough? | 22:31 |
| blogan | i suppose i answered my own question | 22:32 |
| blogan | but you nkow, its a rhetorical question! | 22:32 |
| dougwig | rm_work: don't do that TODO in the post_test_hook yet. i think if we decide to postpone OAR, we'll want to restore the old api job, though i'd like to see the split jobs all running with an 'octavia' in the name as well, for just our project. | 22:32 |
| rm_work | hmm | 22:33 |
| rm_work | i wasn't going to touch anything until someone told me to actually :P | 22:33 |
| rm_work | i tried to be too quick on doing that with the barbican gate and got a lot of pushback | 22:34 |
| * blogan resists | 22:34 | |
| dougwig | rm_work: yes, let's not until post rc1, then which direction we need to go should be clear. | 22:34 |
| *** diogogmt has quit IRC | 22:35 | |
| blogan | dougwig: has a direction been decided on? | 22:35 |
| dougwig | blogan: i believe folks want to continue working on the jobs through tomorrow, and the general feel is hopeful but unlikely. | 22:36 |
| rm_work | dougwig: so what you're telling me is... "do nothing" | 22:36 |
| rm_work | dougwig: I THINK I can manage that | 22:36 |
| dougwig | if anyone disagrees with that summary, feel free. :) | 22:36 |
| dougwig | rm_work: heh. | 22:36 |
| rm_work | it'll be tough, but i'll stick it out | 22:36 |
| rm_work | wish me luck | 22:36 |
| blogan | i can't say it doesnt hurt to push until tomorrow | 22:37 |
| blogan | but won't that 2 hour job piss neutron off? | 22:37 |
| rm_work | yeah we might want to push the patch updating which "minimal" jobs run | 22:37 |
| xgerman | they are pissed off anyway — I have never seen Salvatore happy | 22:37 |
| rm_work | before we merged the AOR patch | 22:37 |
| blogan | he's only happy when he's making fun of himself | 22:37 |
| rm_work | *OAR | 22:37 |
| rm_work | no matter what | 22:38 |
| rm_work | again, it's right here: https://github.com/openstack/neutron-lbaas/blob/master/neutron_lbaas/tests/contrib/post_test_hook.sh#L22 | 22:38 |
| rm_work | so whoever wants to define what "minimal" is | 22:38 |
| rm_work | "test_subset" gets passed to tox as a positional arg | 22:38 |
| rm_work | which is the test filter | 22:38 |
| johnsom | I could take a stab at that. I just need to figure out how to define those subset names | 22:38 |
| rm_work | so you can make a huge list of individual tests if you want even | 22:38 |
| rm_work | it's essentially a regex | 22:39 |
| rm_work | it looks simple because I put "load_balanacer" | 22:39 |
| rm_work | what that does is look in the defined test_env directory | 22:39 |
| rm_work | and the format for each test expands to: | 22:39 |
| rm_work | <path.to.test.file.with.dots.instead.of.slashes.and.no.file.extension>.<TestClass>.<test_name> | 22:40 |
| rm_work | and then it does a regex for "load_balancer" | 22:40 |
| rm_work | and picks all the ones that match | 22:40 |
| rm_work | you can test it by doing | 22:41 |
| rm_work | tox -e apiv2 -- <test_thing_here> | 22:41 |
| rm_work | even not in devstack, which is actually easier, because they all fail but it shows you instantly which ones it selected :) | 22:41 |
| rm_work | it falls out of pretty_tox.sh and the .testr config | 22:42 |
| rm_work | if you really want to trace it | 22:42 |
| blogan | the gate queue is soooo long | 22:44 |
| johnsom | Yes | 22:44 |
| xgerman | Yeah... | 22:44 |
| dougwig | it was >100 a few weeks ago. :) | 22:45 |
| rm_work | heh | 22:45 |
| rm_work | we are NOT helping, lol | 22:45 |
| xgerman | see nobody cares... | 22:45 |
| dougwig | now imagine if *we* doubled the time of that gate, since we're in the integrated gate. | 22:45 |
| dougwig | xgerman: you're on a big 'no one cares' 'everybody hates us anyway' kick today. you need some coffee or something? | 22:45 |
| xgerman | no worries - rough day | 22:47 |
| xgerman | and it’s hot in San Diego!! | 22:47 |
| dougwig | we'll blame the heat. | 22:49 |
| blogan | xgerman: do yall have air conditioning? or is it cool enough year round to not worry about it? | 22:49 |
| xgerman | some of us do… and I do but I got a thermostat as a gift from the [power company which turned out to be a trojan horse not letting me turn on air conditioning during the day on a hot day in an effort to save power | 22:50 |
| blogan | xgerman: lol they try that down here too, luckily someone told me about it and i refused | 22:51 |
| johnsom | Ouch. I have heard of those. Apply hammer firmly | 22:51 |
| blogan | of course not being able to turn on the AC down here in the middle of teh summer can be a death sentence | 22:52 |
| dougwig | texas is uninhabitable at the best of times. of course you need AC. | 22:52 |
| blogan | at least we dont have to deal with snow | 22:52 |
| xgerman | :-) | 22:53 |
| johnsom | Nothing wrong with a little snow | 22:53 |
| blogan | well yeah a little snow is awesome, but shoveling snow every day just to walk to your house or get to work, that sounds awful | 22:53 |
| johnsom | That's what the switch on the dash is for.... | 22:54 |
| blogan | then again ive never been around that much snow | 22:54 |
| *** mlavalle_ has quit IRC | 22:54 | |
| blogan | imagine texas drivers trying to drive in snow | 22:54 |
| xgerman | https://s-media-cache-ak0.pinimg.com/236x/e4/2f/71/e42f716b1529c67ca7e5cc44a5676cc5.jpg | 22:54 |
| blogan | well south texas drivers | 22:54 |
| blogan | lol | 22:55 |
| blogan | is that a bunch of beer cans and bottles? | 22:55 |
| dougwig | idaho drivers suck in snow. we don't get enough of it. it's just ginormous 4x4 trucks hitting everything, since none of them understand that acceleration traction != braking traction. | 22:55 |
| xgerman | yep, that’s what snow is good for :-) | 22:55 |
| blogan | oh lol i didnt realize they were using it to keep them cool, i thought teh snow rolled in with a bunch of trash | 22:56 |
| johnsom | We don't really get enough either, so you give the honda's and toyota's half a day to be safely parked in the ditch | 22:56 |
| blogan | dougwig: sounds like an adventure driving in adaho in the winter! | 22:56 |
| xgerman | probably summer, too — I have seen his picture of the blown tire | 22:57 |
| dougwig | blogan: i just don't do it. we eat what's left in the freezer and rent a movie. | 22:57 |
| dougwig | xgerman: haha. | 22:57 |
| blogan | the very rare times we get any kind of snow dusting, or icy weather, i like to go driving around, there are always many cars in the ditches | 22:58 |
| blogan | bbl | 22:59 |
| johnsom | My favorite is watching those cars try to get up the hill. | 22:59 |
| johnsom | rm_work Question for someone else that knows x509. I want to fix that darn "subjectAltName" warning for the amp certs. | 23:02 |
| rm_you | lol | 23:03 |
| rm_you | is it saying it's missing? | 23:03 |
| johnsom | We can create a uuid OID for our identifier using the OtherName option or be rogue and make up our own prefix with more meaning. | 23:03 |
| johnsom | Thoughts? | 23:03 |
| rm_you | what is the actual error | 23:03 |
| xgerman | I was hoping anchor would fix that | 23:03 |
| johnsom | Yeah, it's not being specified in the current code | 23:03 |
| rm_you | i mean | 23:03 |
| rm_you | we can just add one | 23:04 |
| rm_you | just set it to the CN i think | 23:04 |
| johnsom | SecurityWarning: Certificate has no `subjectAltName`, falling back to check for a `commonName` for now. | 23:04 |
| rm_you | yeah lol | 23:04 |
| rm_you | just in the generatore | 23:04 |
| rm_you | *in the generator, set subjectAltName to the same value as CN | 23:04 |
| johnsom | Well, it has a type in the SAN. That is what I have been thinking about. | 23:04 |
| rm_you | it's technically supposed to be in all certs now | 23:04 |
| johnsom | once we decide this it's a one line change | 23:04 |
| rm_you | but i never bothered because it is really optional in practice | 23:04 |
| rm_you | wait | 23:05 |
| rm_you | what is in SAN now? | 23:05 |
| johnsom | Yeah, it's just a ton of log spew | 23:05 |
| rm_you | i thought nothing | 23:05 |
| johnsom | Yeah, nothing | 23:05 |
| rm_you | so | 23:05 |
| johnsom | Thus, the warning | 23:05 |
| rm_you | just set it? I am not seeing the issue :P | 23:05 |
| *** vivek-ebay has quit IRC | 23:06 | |
| *** vivek-ebay has joined #openstack-lbaas | 23:10 | |
| *** Brian has joined #openstack-lbaas | 23:12 | |
| *** Brian is now known as Guest28591 | 23:13 | |
| *** vivek-ebay has quit IRC | 23:14 | |
| *** Guest28591 has quit IRC | 23:17 | |
| *** minwang2 has quit IRC | 23:26 | |
| *** minwang2 has joined #openstack-lbaas | 23:31 | |
| *** vivek-ebay has joined #openstack-lbaas | 23:44 | |
| *** manishg has quit IRC | 23:48 | |
| *** denghui has joined #openstack-lbaas | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
