| *** ducttape_ has quit IRC | 00:00 | |
| *** ducttape_ has joined #openstack-lbaas | 00:01 | |
| *** ducttape_ has quit IRC | 00:01 | |
| *** ducttape_ has joined #openstack-lbaas | 00:02 | |
| *** ducttape_ has quit IRC | 00:02 | |
| *** ducttape_ has joined #openstack-lbaas | 00:02 | |
| *** ShaneR_ has joined #openstack-lbaas | 00:17 | |
| *** catintheroof has quit IRC | 00:31 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Update Octavia v2 API for vip_network_id https://review.openstack.org/442879 | 01:04 |
|---|---|---|
| armax | johnsom: | 01:27 |
| armax | ping | 01:27 |
| johnsom | pong | 01:27 |
| armax | johnsom: was wondering if you’re aware of https://review.openstack.org/#/c/438682/7 | 01:28 |
| armax | as you guys use the same type of hook as neutron did | 01:28 |
| armax | you may want to tweak things in octavia/neutron-lbaas | 01:28 |
| armax | johnsom: I got neutronclient done: https://review.openstack.org/#/c/442884/ | 01:28 |
| johnsom | Yeah, I think someone already did that for us. | 01:28 |
| johnsom | Let me double check | 01:29 |
| armax | yamamoto | 01:29 |
| armax | I think | 01:29 |
| johnsom | Yeah | 01:29 |
| armax | but I am not sure if that goes deep enough | 01:29 |
| johnsom | https://review.openstack.org/438251 | 01:30 |
| johnsom | Yeah, I can take a deeper look. It looked ok to me, but maybe I am missing something. | 01:33 |
| johnsom | Thanks for the heads up though | 01:33 |
| armax | sure, these things are never too trivial :) | 01:34 |
| armax | I’ll double check with Ihar in the morning | 01:34 |
| johnsom | Question for you.... | 01:34 |
| armax | because I don’t have the full picture | 01:34 |
| armax | and I’ll get back to you | 01:34 |
| armax | shoot | 01:34 |
| johnsom | http://logs.openstack.org/20/439920/1/check/gate-neutron-lbaas-python27-ubuntu-xenial/159c759/console.html#_2017-03-02_00_12_49_439107 | 01:35 |
| johnsom | Did this replace_file change get backported to Ocata? | 01:35 |
| johnsom | Ihar said no on this patch: https://review.openstack.org/#/c/441255/ | 01:35 |
| johnsom | But the stable gate is saying otherwise | 01:36 |
| armax | there must be a dependency mismatch somewhere | 01:36 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Fix load balancer project_id handling on POST https://review.openstack.org/442865 | 01:36 |
| armax | let me look | 01:37 |
| armax | https://github.com/openstack/neutron-lbaas/blob/stable/ocata/tools/tox_install.sh#L17 | 01:38 |
| armax | that’s your problem-o | 01:38 |
| armax | you’re pulling from master-o | 01:38 |
| johnsom | What the..... | 01:38 |
| armax | you want me to continue the sentence? | 01:38 |
| johnsom | I know you already did | 01:39 |
| armax | btw | 01:39 |
| armax | https://github.com/openstack/neutron/blob/master/doc/source/policies/release-checklist.rst | 01:39 |
| armax | some of this stuff is captured in there | 01:39 |
| johnsom | Yep, it's there. Did not know anything about this... | 01:41 |
| johnsom | Thanks again | 01:42 |
| armax | sure thing | 01:42 |
| rm_work | johnsom: yes it did, i told them it would need to be updated after the network_id stuff | 01:45 |
| johnsom | Yeah, I have posts a dependent patch | 01:45 |
| rm_work | yeah reading it now | 01:45 |
| johnsom | https://review.openstack.org/#/c/442891 | 01:47 |
| johnsom | ^^ fixes the stable/ocata branch for neutron-lbaas | 01:47 |
| *** gongysh has joined #openstack-lbaas | 01:49 | |
| *** sanfern has quit IRC | 02:03 | |
| rm_work | lol yeah saw that discussion | 02:04 |
| rm_work | can only +1 :( | 02:04 |
| rm_work | I guess neutron stable team is different | 02:04 |
| johnsom | What? You should have +2 there | 02:05 |
| johnsom | Yeah, but you should be there too, different group. | 02:05 |
| johnsom | Let me look | 02:05 |
| johnsom | Yeah, you should be good | 02:06 |
| johnsom | Logged in? | 02:06 |
| rm_work | :/ | 02:06 |
| rm_work | yes | 02:06 |
| rm_work | I +1'd :P | 02:06 |
| johnsom | Yeah, sorry, end of day, too many things going on. | 02:07 |
| johnsom | https://review.openstack.org/#/admin/projects/openstack/neutron-lbaas,access | 02:07 |
| johnsom | That is the acls | 02:07 |
| johnsom | This is the list of folks: https://review.openstack.org/#/admin/groups/1662,members | 02:07 |
| johnsom | If I can you should be able to | 02:08 |
| rm_work | i'll log out and back in? | 02:08 |
| rm_work | lol well it got a lot of fails | 02:08 |
| rm_work | but still can't +2 | 02:08 |
| johnsom | Yeah maybe??? It's been that way for weeks now | 02:08 |
| rm_work | wtf http://logs.openstack.org/91/442891/1/check/gate-neutron-lbaasv2-dsvm-api-ubuntu-xenial/3b3a286/logs/devstacklog.txt.gz#_2017-03-08_01_55_50_679 | 02:09 |
| rm_work | we need barbican-tempest-plugin in PROJECTS now? | 02:10 |
| rm_work | don't know what to tell you though, only have +/- 1 | 02:11 |
| rm_work | i remember i had this issue before though and magically one day it worked :/ | 02:11 |
| johnsom | Yeah, I don't either | 02:11 |
| johnsom | The projects thing, yeah, they have been doing that, droping projects from the master projects list. I can fix that | 02:11 |
| rm_work | looks like my plans changed for today and i'll be looking at making centos amphora work in our cloud :/ | 02:13 |
| rm_work | ffffff | 02:13 |
| johnsom | Well, we made centos work in Ocata so.... | 02:13 |
| rm_work | yeah | 02:13 |
| rm_work | sometimes I wonder if i *abhor* centos for realistic reasons or just random bias | 02:15 |
| rm_work | so far, being forced to work with it more hasn't made me realize "oh, this is actually OK", more like "yeah there is a reason I hate this | 02:15 |
| rm_work | " | 02:15 |
| johnsom | It's the little differences that get you. However, since I used redhat tribe before ubuntu/debian it goes both ways.... | 02:16 |
| rm_work | centos7 doesn't work with devstack because the version of git it has is SOOOO old | 02:17 |
| rm_work | I have to do http://logs.openstack.org/91/442891/1/check/gate-neutron-lbaasv2-dsvm-api-ubuntu-xenial/3b3a286/logs/devstacklog.txt.gz#_2017-03-08_01_55_50_679 | 02:17 |
| rm_work | err | 02:17 |
| rm_work | https://github.com/rm-you/devstack_deploy/blob/centos/stackme.sh#L18-L20 | 02:17 |
| rm_work | that | 02:17 |
| johnsom | Yes, the fact that much of it is so old sucks. I commented to the wall of RedHat folks that came to our PTG session about the same | 02:17 |
| johnsom | Probably not the right folks, but I felt better afterwards.... | 02:18 |
| rm_work | well, it has to suck, it's free, right? | 02:19 |
| rm_work | can't have it be better than RHEL? | 02:19 |
| johnsom | Well, the centos goal is to track RHEL right? | 02:19 |
| rm_work | right but does it do that effectively? dunno | 02:20 |
| johnsom | I am so sick of projects that don't track stable | 02:27 |
| johnsom | Ok, leaving that stuff broken for tonight. I need to make dinner. | 02:29 |
| johnsom | It's this causing the problem: https://github.com/openstack/barbican/blob/stable/ocata/devstack/plugin.sh#L7 | 02:31 |
| johnsom | Or this: https://github.com/openstack/barbican/blob/stable/ocata/devstack/plugin.sh#L22 | 02:32 |
| johnsom | Lubosz was talking with them about that | 02:32 |
| *** ShaneDavidson has quit IRC | 02:35 | |
| rm_work | johnsom: is it a horrible idea to want some admin config flags *in the DB*, like "enable/disable health monitoring" etc, configurable via operator API? | 02:39 |
| rm_work | so we can like... diable HM during maintenances or such | 02:39 |
| *** bbzhao has joined #openstack-lbaas | 02:43 | |
| *** sanfern has joined #openstack-lbaas | 02:48 | |
| *** amotoki has quit IRC | 02:54 | |
| *** amotoki has joined #openstack-lbaas | 02:56 | |
| *** Guest4533 has joined #openstack-lbaas | 02:58 | |
| *** amotoki has quit IRC | 03:01 | |
| *** Guest4533 has quit IRC | 03:01 | |
| openstackgerrit | ZhaoBo proposed openstack/octavia master: Support vip bandwidth limitation https://review.openstack.org/441912 | 03:01 |
| *** ducttape_ has quit IRC | 03:07 | |
| *** ShaneDavidson has joined #openstack-lbaas | 03:13 | |
| *** chlong_ has joined #openstack-lbaas | 03:15 | |
| *** chlong has quit IRC | 03:15 | |
| *** ShaneDavidson has quit IRC | 03:18 | |
| *** ShaneDavidson has joined #openstack-lbaas | 03:19 | |
| *** ShaneDavidson has quit IRC | 03:27 | |
| *** oomichi has quit IRC | 03:30 | |
| *** oomichi has joined #openstack-lbaas | 03:33 | |
| armax | johnsom: ping | 03:47 |
| armax | what happened here? | 03:47 |
| armax | http://logs.openstack.org/91/442891/1/check/gate-neutron-lbaasv2-dsvm-api-ubuntu-xenial/3b3a286/logs/devstacklog.txt.gz#_2017-03-08_01_55_50_685 | 03:47 |
| armax | do you know? | 03:47 |
| *** amotoki has joined #openstack-lbaas | 03:56 | |
| *** amotoki has quit IRC | 04:01 | |
| rm_work | armax: yes | 04:01 |
| rm_work | we're looking at it. barbican's stuff doesn't track stable there... so we either need to add that to our project list, or get them to fix their devstack stuff in ocata to actually track ocata | 04:01 |
| rm_work | armax: [11:31:39] <johnsom>It's this causing the problem: https://github.com/openstack/barbican/blob/stable/ocata/devstack/plugin.sh#L7 | 04:02 |
| rm_work | [11:32:01] <johnsom>Or this: https://github.com/openstack/barbican/blob/stable/ocata/devstack/plugin.sh#L22 | 04:02 |
| rm_work | [11:32:14] <johnsom>Lubosz was talking with them about that | 04:02 |
| armax | rm_work: thanks | 04:02 |
| *** ducttape_ has joined #openstack-lbaas | 04:09 | |
| *** ducttape_ has quit IRC | 04:13 | |
| *** csomerville has joined #openstack-lbaas | 04:22 | |
| *** cody-somerville has quit IRC | 04:22 | |
| *** amotoki has joined #openstack-lbaas | 04:24 | |
| rm_work | anyone know where the console log ends up in centos? like, where can i look to see what happened during boot? | 04:31 |
| *** links has joined #openstack-lbaas | 04:41 | |
| *** ShaneDavidson has joined #openstack-lbaas | 04:42 | |
| *** ShaneDavidson has quit IRC | 05:04 | |
| rm_work | nm it looked like dmesg was missing stuff but it's there | 05:07 |
| *** ducttape_ has joined #openstack-lbaas | 05:09 | |
| *** ducttape_ has quit IRC | 05:14 | |
| *** gongysh has quit IRC | 05:31 | |
| *** gongysh has joined #openstack-lbaas | 05:32 | |
| *** yamamoto_ has joined #openstack-lbaas | 06:01 | |
| *** ducttape_ has joined #openstack-lbaas | 06:10 | |
| rm_work | hmm time to figure out | 06:12 |
| rm_work | ::ffff:10.32.60.187 - - [08/Mar/2017:06:05:58 +0000] "PUT /0.5/listeners/77fed5d8-0ea7-4e1f-a510-bded86ee03d2/0138db8b-0c46-4f8a-a8a1-0c3a88533357/haproxy HTTP/1.1" 400 417 "-" "Octavia HaProxy Rest Client/0.5 (https://wiki.openstack.org/wiki/Octavia)" | 06:12 |
| *** ducttape_ has quit IRC | 06:15 | |
| *** ShaneR_ has quit IRC | 06:27 | |
| *** ShaneR_ has joined #openstack-lbaas | 06:30 | |
| openstackgerrit | ZhaoBo proposed openstack/octavia master: Support vip bandwidth limitation https://review.openstack.org/441912 | 06:32 |
| *** yamamoto_ has quit IRC | 06:36 | |
| *** gcheresh_ has joined #openstack-lbaas | 06:40 | |
| *** armax has quit IRC | 06:40 | |
| *** armax has joined #openstack-lbaas | 06:40 | |
| *** armax has quit IRC | 06:41 | |
| *** pcaruana has joined #openstack-lbaas | 06:43 | |
| *** ShaneR_ has quit IRC | 06:46 | |
| rm_work | nmagnezi: around? | 07:00 |
| rm_work | ah nm got it i think | 07:03 |
| rm_work | was trying to remember the deal with user_group for centos amps | 07:03 |
| *** kobis has quit IRC | 07:03 | |
| *** ducttape_ has joined #openstack-lbaas | 07:11 | |
| *** ducttape_ has quit IRC | 07:16 | |
| johnsom | I think Nir is working on that | 07:18 |
| rm_work | yes | 07:18 |
| rm_work | that's why i pinged him :P | 07:18 |
| rm_work | but i figured it out i think | 07:19 |
| johnsom | nmagnezi can you give an update on the user_group patch | 07:19 |
| johnsom | NP, just checking in before sleep... | 07:20 |
| rm_work | things looking good | 07:20 |
| rm_work | though might have found a bug with _get_resource_by_filters | 07:21 |
| johnsom | Fyi, I would like to see the base v2 test patches merge so we have tests | 07:21 |
| rm_work | 2017-03-08 00:19:01.418 12191 ERROR octavia.network.drivers.neutron.base File "/opt/octavia/env/lib/python2.7/site-packages/octavia/network/drivers/neutron/base.py", line 169, in _get_resource_by_filters | 07:23 |
| rm_work | 2017-03-08 00:19:01.418 12191 ERROR octavia.network.drivers.neutron.base resource = resource['%ss' % resource_type][0] | 07:23 |
| rm_work | 2017-03-08 00:19:01.418 12191 ERROR octavia.network.drivers.neutron.base IndexError: list index out of range | 07:23 |
| rm_work | k yeah i know T_T i'll look today | 07:23 |
| *** yamamoto_ has joined #openstack-lbaas | 07:23 | |
| rm_work | so, that happens when failing over and there are *zero* ports for the user | 07:23 |
| rm_work | because this was the only amp, and it was deleted | 07:23 |
| rm_work | it is expecting to see neutron_client_exceptions.NotFound | 07:24 |
| rm_work | but it doesn't | 07:24 |
| *** rcernin has joined #openstack-lbaas | 07:40 | |
| *** belharar has joined #openstack-lbaas | 07:43 | |
| *** tesseract has joined #openstack-lbaas | 07:43 | |
| *** kobis has joined #openstack-lbaas | 07:56 | |
| *** belharar has quit IRC | 08:04 | |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Fix quota lock on single_create with auth enabled https://review.openstack.org/434518 | 08:07 |
| *** ducttape_ has joined #openstack-lbaas | 08:12 | |
| *** ducttape_ has quit IRC | 08:17 | |
| rm_work | johnsom: you meant this i assume? https://review.openstack.org/#/c/413697/ | 08:17 |
| *** chlong_ has quit IRC | 08:18 | |
| nmagnezi | johnsom, hey, I had some unexpected tasks on my plate the past few days, I will finalize the patch early next week. | 08:37 |
| rm_work | yeah for now I set "user_group = haproxy" and it seems to work :) | 08:38 |
| nmagnezi | rm_work, hi Adam :) | 08:38 |
| nmagnezi | rm_work, you never sleep, huh? :) | 08:38 |
| rm_work | heh | 08:38 |
| rm_work | I just work odd hours :P | 08:38 |
| rm_work | definitely sleep somewhere in there | 08:38 |
| nmagnezi | rm_work, same here :) | 08:39 |
| nmagnezi | rm_work, Adam do you happen to know if there was any work related to loadbalancers migration from the legacy haproxy in namespace to Octavia? | 08:41 |
| nmagnezi | i think this one is related: https://bugs.launchpad.net/octavia/+bug/1656390 | 08:41 |
| openstack | Launchpad bug 1656390 in octavia "Need to move the legacy namespace HAProxy driver from nlbaas to Octavia" [High,New] | 08:41 |
| rm_work | ha... ha... ha... | 08:41 |
| nmagnezi | but looks like no one handled it | 08:41 |
| rm_work | soooo | 08:41 |
| rm_work | sec | 08:41 |
| rm_work | http://lists.openstack.org/pipermail/openstack-dev/2017-March/113432.html | 08:42 |
| rm_work | nmagnezi: search for "Legacy Namespace HAProxy driver" | 08:43 |
| nmagnezi | rm_work, thanks for pointing this out (i dunno why i missed that email, I was waiting for it). anyhow my question is about a scenario in which an operator with existing loadbalancers that were created with haproxy in namespace | 08:49 |
| nmagnezi | and now that operator wants to migrate those loadbalancers to Octavia | 08:49 |
| rm_work | yeah, so... erm | 08:50 |
| rm_work | the API will be the same | 08:50 |
| rm_work | so theoretically they could just recreate them | 08:50 |
| rm_work | is that not a good option? :P | 08:50 |
| nmagnezi | lol | 08:50 |
| *** reedip_ has quit IRC | 08:50 | |
| nmagnezi | you know the answer :D | 08:50 |
| rm_work | T_T | 08:50 |
| rm_work | "someone" could write a tool to look at the DB for n-lbaas and do the creations automatically... | 08:51 |
| nmagnezi | operator are always happy to do manual work :) | 08:51 |
| rm_work | but IPs would change | 08:51 |
| rm_work | the tool would not be difficult to write | 08:51 |
| nmagnezi | yes but it would have to be designed in a way that will give a minimal downtime, plus an option to roll back in case something goes wrong | 08:52 |
| nmagnezi | so it is not that simple :) | 08:52 |
| rm_work | well | 08:52 |
| rm_work | you'd do the creates first | 08:52 |
| rm_work | and then let people swap over to the new IPs | 08:52 |
| rm_work | and then take down the old system after | 08:52 |
| rm_work | give them a couple weels | 08:52 |
| rm_work | *weeks | 08:53 |
| rm_work | look at traffic stats on the old LBs to see who isn't switching | 08:53 |
| nmagnezi | can the old driver and octavia co-exist for some time? | 08:53 |
| rm_work | I see no reason why not | 08:53 |
| rm_work | I mean | 08:53 |
| rm_work | basically what will happen | 08:53 |
| rm_work | is that you'll take the driver/agent offline for the old one | 08:53 |
| rm_work | and point it at octavia instead | 08:54 |
| rm_work | the old LBs will still *be running* | 08:54 |
| rm_work | just no API/agent to update them anymore | 08:54 |
| nmagnezi | yeah because the haproxy processes are still up | 08:54 |
| rm_work | then eventually just decommission the box | 08:54 |
| rm_work | right | 08:54 |
| rm_work | and using the DB, do scripted cleanup of the openstack resources (ports, FLIPs, whatever) | 08:55 |
| rm_work | because the DB will be left intact as well | 08:55 |
| rm_work | IMO that is the way to go | 08:55 |
| nmagnezi | sounds reasonable, maybe I should file a spec or something | 08:56 |
| rm_work | mayhaps | 08:56 |
| rm_work | I'd help fill it out i guess if you start something | 08:56 |
| rm_work | i've been thinking about it at least a little bit | 08:56 |
| nmagnezi | i think I will add this to the meeting agenda and we can start with your suggestion as a baseline, see what ppl have to say | 08:56 |
| rm_work | k | 08:56 |
| nmagnezi | rm_work, that would be awesome. | 08:57 |
| nmagnezi | rm_work, now all I need is to find the time :D | 08:57 |
| rm_work | yes | 08:57 |
| rm_work | that sounds familiar | 08:57 |
| nmagnezi | hah | 08:58 |
| *** reedip_ has joined #openstack-lbaas | 09:02 | |
| *** openstackgerrit has quit IRC | 09:03 | |
| *** openstackgerrit has joined #openstack-lbaas | 09:06 | |
| openstackgerrit | Merged openstack/octavia master: Switch to use stable data_utils https://review.openstack.org/442767 | 09:06 |
| *** ducttape_ has joined #openstack-lbaas | 09:13 | |
| *** ducttape_ has quit IRC | 09:17 | |
| *** reedip_ has quit IRC | 09:18 | |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Fix cascade delete flow (pool quota requirements missing) https://review.openstack.org/437200 | 09:30 |
| *** reedip_ has joined #openstack-lbaas | 09:31 | |
| *** yamamoto_ has quit IRC | 09:35 | |
| *** ducttape_ has joined #openstack-lbaas | 10:14 | |
| *** ducttape_ has quit IRC | 10:19 | |
| *** anilvenkata_afk is now known as anilvenkata | 10:34 | |
| *** yamamoto has joined #openstack-lbaas | 10:45 | |
| *** yamamoto has quit IRC | 10:48 | |
| *** yamamoto has joined #openstack-lbaas | 10:50 | |
| *** yamamoto has quit IRC | 10:51 | |
| *** reedip_ has quit IRC | 10:53 | |
| *** belharar has joined #openstack-lbaas | 10:54 | |
| *** reedip_ has joined #openstack-lbaas | 11:07 | |
| *** yamamoto has joined #openstack-lbaas | 11:15 | |
| *** ducttape_ has joined #openstack-lbaas | 11:15 | |
| *** yamamoto has quit IRC | 11:16 | |
| *** gongysh has quit IRC | 11:16 | |
| *** sanfern has quit IRC | 11:17 | |
| *** yamamoto has joined #openstack-lbaas | 11:18 | |
| *** yamamoto has quit IRC | 11:19 | |
| *** ducttape_ has quit IRC | 11:20 | |
| *** catintheroof has joined #openstack-lbaas | 11:20 | |
| *** reedip_ has quit IRC | 11:29 | |
| *** reedip_ has joined #openstack-lbaas | 11:42 | |
| *** eezhova has joined #openstack-lbaas | 11:51 | |
| *** amotoki has quit IRC | 12:00 | |
| *** eezhova has quit IRC | 12:12 | |
| *** ducttape_ has joined #openstack-lbaas | 12:16 | |
| *** belharar has quit IRC | 12:18 | |
| *** ducttape_ has quit IRC | 12:21 | |
| *** belharar has joined #openstack-lbaas | 12:25 | |
| *** catinthe_ has joined #openstack-lbaas | 12:29 | |
| *** catintheroof has quit IRC | 12:29 | |
| *** catintheroof has joined #openstack-lbaas | 12:30 | |
| *** catinthe_ has quit IRC | 12:31 | |
| *** sanfern has joined #openstack-lbaas | 12:31 | |
| *** eezhova has joined #openstack-lbaas | 12:47 | |
| *** eezhova has quit IRC | 13:00 | |
| *** gongysh has joined #openstack-lbaas | 13:06 | |
| *** ducttape_ has joined #openstack-lbaas | 13:17 | |
| openstackgerrit | Nir Magnezi proposed openstack/octavia master: Auto detect haproxy user_group https://review.openstack.org/429398 | 13:20 |
| *** ducttape_ has quit IRC | 13:21 | |
| *** links has quit IRC | 13:25 | |
| *** ducttape_ has joined #openstack-lbaas | 13:25 | |
| *** amotoki has joined #openstack-lbaas | 13:51 | |
| *** ducttape_ has quit IRC | 13:54 | |
| *** KeithMnemonic has joined #openstack-lbaas | 13:57 | |
| *** belharar has quit IRC | 14:02 | |
| *** gcheresh_ has quit IRC | 14:02 | |
| *** strigazi_AFK is now known as strigazi | 14:04 | |
| *** matt-borland has joined #openstack-lbaas | 14:08 | |
| xgerman | o/ | 14:15 |
| *** malos has joined #openstack-lbaas | 14:22 | |
| rm_work | o/ | 14:24 |
| rm_work | almost got this delete_cascade stuff finished | 14:24 |
| rm_work | working on fixing the tests | 14:24 |
| rm_work | and also including in this patch additional validation to not allow deleting things with children WITHOUT cascade | 14:25 |
| rm_work | because essentially it was allowing normal delete to cascade, but badly | 14:25 |
| *** ducttape_ has joined #openstack-lbaas | 14:27 | |
| *** belharar has joined #openstack-lbaas | 14:27 | |
| *** gcheresh_ has joined #openstack-lbaas | 14:42 | |
| *** gongysh has quit IRC | 14:42 | |
| *** anilvenkata has quit IRC | 14:42 | |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Fix cascade delete flow (pool quota requirements missing) https://review.openstack.org/437200 | 14:45 |
| rm_work | we'll see if that works better... :/ not sure how to effectively test tempest scenarios right now | 14:45 |
| rm_work | so letting the gate do it, which is ... slow, but works | 14:45 |
| *** Swami has joined #openstack-lbaas | 14:45 | |
| *** chlong_ has joined #openstack-lbaas | 14:49 | |
| *** gcheresh_ has quit IRC | 14:49 | |
| *** Swami_ has joined #openstack-lbaas | 14:52 | |
| *** Swami has quit IRC | 14:56 | |
| *** Swami__ has joined #openstack-lbaas | 14:56 | |
| *** Swami_ has quit IRC | 14:59 | |
| *** belharar has quit IRC | 15:11 | |
| *** belharar has joined #openstack-lbaas | 15:16 | |
| *** Swami_ has joined #openstack-lbaas | 15:20 | |
| *** Swami__ has quit IRC | 15:24 | |
| *** fnaval has joined #openstack-lbaas | 15:33 | |
| *** links has joined #openstack-lbaas | 15:36 | |
| *** amotoki has quit IRC | 15:37 | |
| *** belharar has quit IRC | 15:42 | |
| *** belharar has joined #openstack-lbaas | 15:42 | |
| *** belharar has quit IRC | 15:47 | |
| *** belharar has joined #openstack-lbaas | 15:48 | |
| rm_work | what the | 15:49 |
| rm_work | http://logs.openstack.org/00/437200/4/check/gate-octavia-v1-dsvm-scenario-ubuntu-xenial-nv/0802216/logs/devstacklog.txt.gz#_2017-03-08_15_14_54_378 | 15:49 |
| rm_work | fff what merged | 15:50 |
| rm_work | somehow now the tests aren't installing the dib reqs? | 15:53 |
| *** rcernin has quit IRC | 15:58 | |
| *** belharar has quit IRC | 15:59 | |
| rm_work | wtffff | 16:01 |
| rm_work | I see it but i don't understand how | 16:02 |
| rm_work | OK so it is this: https://review.openstack.org/#/c/393009/ | 16:07 |
| rm_work | which just merged back to master here: https://review.openstack.org/#/c/441723/ | 16:07 |
| rm_work | removing dib-utils from diskimage-builder requirements | 16:07 |
| rm_work | which we use and didn't require directly <_< | 16:08 |
| rm_work | working on iy | 16:08 |
| *** gcheresh_ has joined #openstack-lbaas | 16:08 | |
| *** gcheresh_ has quit IRC | 16:11 | |
| rm_work | note to self: babel does not properly require pytz | 16:13 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Look for diskimage-builder not dib-utils when building amps https://review.openstack.org/443220 | 16:18 |
| rm_work | ok time to see if i also need a project-config change | 16:19 |
| *** kobis has quit IRC | 16:21 | |
| rm_work | greghaynes: something with the requirements order in diskimage-builder makes installation fail with pip, because babel requires pytz and so does pbr but it gets installed AFTER pbr? or something | 16:40 |
| rm_work | it's kinda weird | 16:40 |
| rm_work | just make a new venv and pip install -r diskimage-builder/requirements.txt | 16:41 |
| rm_work | you'll see the issue | 16:41 |
| rm_work | really annoying | 16:41 |
| johnsom | Joy | 16:41 |
| johnsom | What is the project-config change about? | 16:41 |
| *** ducttape_ has quit IRC | 16:43 | |
| rm_work | yep, need project-config change | 16:43 |
| *** ducttape_ has joined #openstack-lbaas | 16:43 | |
| rm_work | diskimage-builder needs to be INSTALLED now | 16:43 |
| rm_work | not just present | 16:43 |
| *** ducttape_ has quit IRC | 16:44 | |
| *** ducttape_ has joined #openstack-lbaas | 16:44 | |
| rm_work | johnsom: working on it... :/ | 16:45 |
| rm_work | i loveeeee how constantly stuff breaks out from under us wrt DIB >_> | 16:45 |
| rm_work | johnsom: oh maybe it can be in our gate hook | 16:48 |
| rm_work | johnsom: where do you think is best to do an install on the local dib clone | 16:48 |
| rm_work | or should we just ... can we just add it to the projects list? | 16:48 |
| rm_work | and not clone it manually? | 16:48 |
| rm_work | oh it is in PROJECTS | 16:49 |
| johnsom | Yeah, well other projects aren't are friend either. | 16:49 |
| johnsom | So isn't it installed via requirements.txt? | 16:50 |
| johnsom | Yeah, it's in global requirements, why don't we just put it in requirements.txt now? | 16:51 |
| johnsom | diskimage-builder>=1.1.2,!=1.6.0,!=1.7.0,!=1.7.1 # Apache-2.0 | 16:51 |
| johnsom | rm_work ^^ | 16:51 |
| rm_work | lol k | 16:51 |
| rm_work | do we still need it cloned tho? | 16:52 |
| rm_work | I think so? | 16:52 |
| johnsom | I think there was a patch a long time ago that removed our cloned requirement | 16:52 |
| rm_work | lol... well, we still clone it | 16:52 |
| *** kobis has joined #openstack-lbaas | 16:52 | |
| rm_work | i guess i'll just... try removing that :P | 16:52 |
| johnsom | Yeah, they weren't ready | 16:52 |
| johnsom | https://review.openstack.org/340424 | 16:53 |
| rm_work | let's try this | 16:53 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Look for diskimage-builder not dib-utils when building amps https://review.openstack.org/443220 | 16:54 |
| rm_work | should do it hopefully | 16:54 |
| *** ducttape_ has quit IRC | 16:54 | |
| rm_work | can you take this over if it doesn't work? i'll prolly be off shortly :/ | 16:54 |
| johnsom | Probably some readme work too | 16:54 |
| rm_work | ugh | 16:55 |
| johnsom | Yeah, Np | 16:55 |
| * rm_work looks | 16:55 | |
| *** links has quit IRC | 16:55 | |
| johnsom | Wow, what is up with pep8??? | 17:03 |
| rm_work | ok, i fixed the readme | 17:06 |
| rm_work | waiting tho to see if the main part *worked* | 17:07 |
| johnsom | http://logs.openstack.org/20/443220/2/check/gate-octavia-pep8-ubuntu-xenial/a6b103a/console.html#_2017-03-08_16_59_53_857016 | 17:07 |
| johnsom | I totally don't get this... | 17:07 |
| rm_work | uhhh | 17:07 |
| rm_work | wut | 17:07 |
| rm_work | should api-ref be excluded from pep8? | 17:10 |
| johnsom | It runs fine normally | 17:11 |
| johnsom | Plus most of that isn't api-ref | 17:11 |
| rm_work | err, didn't we have a pylintrc file at some point? | 17:12 |
| rm_work | ah yeah true hold up | 17:12 |
| rm_work | i just saw the first thing | 17:12 |
| rm_work | this is a ton of stuff | 17:12 |
| johnsom | I don't think most of openstack runs that anymore | 17:12 |
| johnsom | I wonder if a new hacking came out or something... | 17:13 |
| rm_work | mostly E402 and W503 | 17:13 |
| rm_work | yeah | 17:13 |
| rm_work | something new must have appeared | 17:13 |
| rm_work | I'll smother it | 17:13 |
| rm_work | uhh, once i figure out where that went | 17:15 |
| rm_work | wtf happened to pylintrc | 17:15 |
| rm_work | is that not used anymore? | 17:15 |
| johnsom | I don't think we run pylint | 17:15 |
| johnsom | Pretty sure OpenStack dropped it | 17:15 |
| *** KeithMnemonic has quit IRC | 17:15 | |
| rm_work | ah | 17:16 |
| rm_work | where do exceptions go now | 17:16 |
| *** armax has joined #openstack-lbaas | 17:16 | |
| johnsom | Hmm, hacking didn't change | 17:17 |
| *** tesseract has quit IRC | 17:18 | |
| rm_work | blegh | 17:24 |
| rm_work | looks like it didn't fix the issue either | 17:24 |
| xgerman | my new enemy is py35 | 17:24 |
| *** matt-borland has quit IRC | 17:26 | |
| rm_work | johnsom: wtf, diskimage-builder is included in requirements.txt now, but it still says it's not installed? >_> | 17:26 |
| johnsom | Yeah, saw that | 17:26 |
| rm_work | checking | 17:27 |
| rm_work | yeah wtf it should be good | 17:30 |
| rm_work | ok: gate-octavia-v1-dsvm-scenario-ubuntu-xenial-nv: | 17:31 |
| rm_work | that one passed | 17:32 |
| rm_work | so it's just ... py3x that can't find it? | 17:32 |
| rm_work | gonna push up my readme change | 17:32 |
| rm_work | and then start rebasing stuff on this | 17:32 |
| rm_work | actually, gonna let the run finish so we get to keep these logs | 17:34 |
| rm_work | maybe dib isn't py3x compatible? | 17:35 |
| johnsom | Oh that would be horrible | 17:35 |
| rm_work | greghaynes: ^^ ? | 17:35 |
| greghaynes | im going to guess you installed it with py2 | 17:36 |
| greghaynes | but, dib certainly works with py3 | 17:36 |
| rm_work | hmm | 17:36 |
| rm_work | so, on py2 it works fine | 17:37 |
| rm_work | on py3, i can see where it says it installed it, and then it isn't found | 17:37 |
| greghaynes | what command installs it? | 17:37 |
| rm_work | oh | 17:38 |
| rm_work | fuck i think i see | 17:38 |
| rm_work | we're checking directly with `pip` | 17:38 |
| rm_work | need to use like | 17:38 |
| rm_work | env pip | 17:38 |
| johnsom | Ah | 17:38 |
| johnsom | Still that should be the right "pip" I would expect | 17:40 |
| rm_work | no | 17:42 |
| rm_work | well | 17:42 |
| rm_work | MAYBE but i am guessing no | 17:42 |
| rm_work | prolly is the py27 pip | 17:42 |
| rm_work | anywho, fixing | 17:42 |
| rm_work | actually, fixed, but waiting on these first tests to finish | 17:42 |
| rm_work | want to have a record | 17:42 |
| diltram | johnsom, rm_work: didn't had any problems with instlling diskimage-builder on py3 | 17:43 |
| rm_work | right, i figured it out | 17:44 |
| rm_work | our script was looking to see if it was installed using py2 pip, after it was installed with py3 pip | 17:44 |
| diltram | ok | 17:44 |
| rm_work | actually... how are we supposed to run python/pip in devstack? | 17:44 |
| rm_work | if we execute them directly, should we use ... env? or, is there like $DEVSTACK_PIP | 17:45 |
| rm_work | greghaynes: do you know? ^^ | 17:45 |
| greghaynes | I have no idea | 17:45 |
| greghaynes | but thats probably the right question to ask | 17:45 |
| greghaynes | there may be an env var? | 17:45 |
| greghaynes | qa folks likely know | 17:46 |
| rm_work | yeah asking there | 17:47 |
| rm_work | heh funny, before we were manually installing dib-utils with pip, and then manually checking with pip, and so it was fine | 17:48 |
| rm_work | and when we ran disk-image-create it was running under py2 | 17:48 |
| rm_work | johnsom: you figure out the pep8 thing? | 17:50 |
| johnsom | No, I checked hacking, but it hasn't changed. I'm in a meeting so splitting time | 17:51 |
| *** KeithMnemonic has joined #openstack-lbaas | 17:53 | |
| rm_work | yeah k looking at it now then | 17:53 |
| rm_work | seeing what got updated since my last good run | 17:53 |
| rm_work | I would really like an award for "first to run into pretty much every gate issue" | 17:54 |
| johnsom | Yes! We would totally win | 17:54 |
| rm_work | octavia is the forerunner of failure | 17:54 |
| *** ducttape_ has joined #openstack-lbaas | 17:55 | |
| rm_work | "We fail first, so you don't have to!" | 17:55 |
| johnsom | Maybe our mascot should be a canary..... | 17:55 |
| rm_work | T_T | 17:55 |
| *** Swami_ has quit IRC | 17:55 | |
| rm_work | wow | 17:56 |
| rm_work | pep8==1.5.7 -> pep8==1.7.0 | 17:56 |
| rm_work | it's actually *pep8* | 17:56 |
| rm_work | flake8 and hacking are unchanged | 17:56 |
| rm_work | and results are easily reproduced locally | 17:57 |
| rm_work | so ... how do we either: exclude things from pep8 checks; ignore certain checks | 17:57 |
| rm_work | i thought we were already ignoring some but i can't find that anymore | 17:57 |
| johnsom | There is an ignore in tox.ini | 17:57 |
| rm_work | where? | 17:59 |
| rm_work | well, i just added --ignore E402 | 17:59 |
| rm_work | to flake8 | 17:59 |
| rm_work | i don't see any other ones | 17:59 |
| *** ducttape_ has quit IRC | 17:59 | |
| johnsom | Yeah, that might be the only one. I'm trying to remember as well | 17:59 |
| rm_work | and W503 | 17:59 |
| rm_work | AH i see it | 18:00 |
| rm_work | k | 18:00 |
| rm_work | flake8 section | 18:00 |
| rm_work | got it | 18:00 |
| rm_work | rolling that into this same patch and just calling it "gatefix patch of the day" | 18:00 |
| rm_work | kk brb | 18:02 |
| *** ducttape_ has joined #openstack-lbaas | 18:02 | |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Look for diskimage-builder not dib-utils when building amps https://review.openstack.org/443220 | 18:02 |
| rm_work | damn thought that last job had finished already | 18:04 |
| rm_work | ah well | 18:04 |
| johnsom | I can throw a patch up that fixes those two, just grunt work | 18:04 |
| rm_work | seems like kinda arbitrary new rules :/ | 18:05 |
| rm_work | not sure how much we care | 18:06 |
| rm_work | also, do we have a plan for fixing this OVH kvm issue? :/ | 18:07 |
| rm_work | tired of "oh crap what did I break? oh it's just OVH" | 18:07 |
| johnsom | Yeah, the plan is to get containers going | 18:07 |
| rm_work | seems like every time it looks like our gate jobs might be actually stable, something else comes up T_T | 18:07 |
| rm_work | k so that's the *short term* plan even? :P | 18:08 |
| johnsom | Yes, exactly. Then people start giving me a hard time for unstable and non-voting gates | 18:08 |
| johnsom | Well, there is hope that whatever is broken gets fixed.... Grin | 18:09 |
| johnsom | Otherwise it would be trying to figure out if it's an ovh host and disabling kvm, which isn't great and leads to timeout failures... | 18:10 |
| rm_work | yeah i'm out for a bit | 18:10 |
| rm_work | if that doesn't pass... you can take it over? | 18:10 |
| johnsom | Yep, thanks! | 18:10 |
| rm_work | looking OK so far | 18:10 |
| rm_work | but the failures are like 30 minutes in so that doesn't mean much :P | 18:10 |
| *** ThiagoCMC has joined #openstack-lbaas | 18:17 | |
| ThiagoCMC | Hey guys! Does OpenStack's LBaaS does it supports HTTP/2.0? | 18:18 |
| ThiagoCMC | On Ubuntu 16.04, NGinx supports it... | 18:18 |
| ThiagoCMC | Also, looks like HAProxy 1.6 on Ubuntu might have it enabled, since it is compiled against libssl-dev, same one that nginx uses... | 18:18 |
| ThiagoCMC | But, what about OpenStack's LBaaS? | 18:18 |
| johnsom | If you are using the haproxy namespace driver or octavia we are using the HAProxy included with the distribution, 1.6 in the case of ubuntu 16.04 | 18:19 |
| ThiagoCMC | Oh, that is awesome! | 18:24 |
| ThiagoCMC | I just started to research about this, my LBaaS is not even deployed yet... | 18:25 |
| ThiagoCMC | Thing is that I have a bunch of NGinx as a reverse proxy / LB and I want to use OpenStack instead of my instances / VMs... | 18:25 |
| rm_work | nope it didn't like my `env pip` thing | 18:32 |
| rm_work | obviously need to figure out the actually recommended way | 18:32 |
| rm_work | but no one responded in -qa | 18:32 |
| johnsom | Yeah, I am going to push up a test that dumps the pip output. Just not sure why this changed all of a sudden. These gates were running fine with py3 before | 18:34 |
| rm_work | right | 18:34 |
| rm_work | but then we changed from installing the dib stuff with `pip install` | 18:35 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Look for diskimage-builder not dib-utils when building amps https://review.openstack.org/443220 | 18:35 |
| rm_work | which was installing it using the same (wrong) pip | 18:35 |
| rm_work | like i was saying, pretty sure that is going to be pip2 | 18:36 |
| rm_work | and we need pip3 for the py3 gates | 18:36 |
| *** chlong_ has quit IRC | 18:37 | |
| *** rcernin has joined #openstack-lbaas | 18:42 | |
| *** ducttape_ has quit IRC | 18:47 | |
| rm_work | johnsom: ok it's `get_pip_command` | 18:48 |
| johnsom | I am wondering if we even need that check \ | 18:48 |
| rm_work | lol | 18:48 |
| rm_work | maybe not? | 18:48 |
| johnsom | Now that it's in requirements, does it really need to check that | 18:48 |
| rm_work | like... maybe we just don't check | 18:48 |
| rm_work | and it just fails | 18:48 |
| rm_work | if it's not there | 18:48 |
| rm_work | well yeah | 18:48 |
| rm_work | it'll just... | 18:49 |
| rm_work | not find disk-image-builder | 18:49 |
| rm_work | (the binary won't be there) | 18:49 |
| rm_work | or else if it is, those packages are there | 18:49 |
| rm_work | yeah i think just removing it is the right call | 18:49 |
| johnsom | Right, it will get down to the disk-image-create command and bomb out | 18:49 |
| johnsom | Should I push that up or are you doing it? | 18:50 |
| rm_work | got it | 18:50 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Look for diskimage-builder not dib-utils when building amps https://review.openstack.org/443220 | 18:51 |
| rm_work | so that should be good | 18:51 |
| *** chlong_ has joined #openstack-lbaas | 18:53 | |
| *** pcaruana has quit IRC | 19:13 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Look for diskimage-builder not dib-utils when building amps https://review.openstack.org/443220 | 19:19 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Update for new pep8 rules E402 and W503 https://review.openstack.org/443298 | 19:30 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Update for new pep8 rules E402 and W503 https://review.openstack.org/443298 | 19:31 |
| *** cody-somerville has joined #openstack-lbaas | 19:33 | |
| johnsom | My ever so important contribution.... | 19:33 |
| *** csomerville has quit IRC | 19:34 | |
| *** gcheresh_ has joined #openstack-lbaas | 19:35 | |
| *** ducttape_ has joined #openstack-lbaas | 19:47 | |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Fix cascade delete flow (pool quota requirements missing) https://review.openstack.org/437200 | 19:58 |
| johnsom | Octavia meeting starting soon on #openstack-meeting-alt | 19:58 |
| rm_work | oh BTW this is good now (assuming you're OK with the change): https://review.openstack.org/#/c/434518/ | 19:58 |
| rm_work | took care of the last issue | 19:58 |
| johnsom | We still have a DIB issue with yaml now | 19:59 |
| rm_work | ughwut | 19:59 |
| rm_work | where | 19:59 |
| rm_work | ah | 20:00 |
| rm_work | is that a system yaml thing? | 20:00 |
| rm_work | where the bindep installs py2 only? | 20:01 |
| rm_work | johnsom: same issue | 20:09 |
| rm_work | py2 running when py3 should be | 20:09 |
| rm_work | https://github.com/openstack/diskimage-builder/blob/master/diskimage_builder/elements/package-installs/extra-data.d/99-squash-package-install | 20:09 |
| rm_work | greghaynes: ^^ I think this is not running with the devstack python | 20:11 |
| rm_work | sorry, dunno if this is relevant to you, somehow you became my go-to for DIB >_> | 20:12 |
| greghaynes | Haja | 20:12 |
| rm_work | http://logs.openstack.org/20/443220/6/check/gate-octavia-v1-dsvm-py3x-scenario-ubuntu-xenial-nv/b26cc80/logs/devstacklog.txt.gz#_2017-03-08_19_45_47_882 | 20:13 |
| greghaynes | I'm actually not doing much dib nowadays, but yes, that looks like it doesn't respect $DIB_PYTHON | 20:13 |
| rm_work | see there, yaml IS installed, but only in the "devstack py3" | 20:13 |
| rm_work | greghaynes: any chance you know how to fix it? :P | 20:13 |
| greghaynes | And you should make sure you set DIB_PYTHON to match what you want | 20:14 |
| rm_work | yeah I believe it is set correctly | 20:14 |
| greghaynes | Well that python_path var is pretty straightforward | 20:14 |
| rm_work | just need everything to respect it | 20:14 |
| rm_work | so we need to like... | 20:15 |
| greghaynes | It needs to take in to account that if dib_python is set that takes precedence | 20:15 |
| rm_work | set python_path=$DIB_PYTHON if it exists | 20:15 |
| rm_work | yeah k | 20:15 |
| rm_work | i wonder how soon we can get that in | 20:15 |
| rm_work | it's breaking our py3 gates | 20:15 |
| rm_work | johnsom: anyway, that's the fix, i seriously do have to go now | 20:15 |
| greghaynes | I think it needs to be command -v python$DIB_PYTHON | 20:15 |
| greghaynes | No idea, as I am not really up with the dib status lately | 20:16 |
| greghaynes | There is a dib channel you can ask in though | 20:16 |
| *** chlong_ has quit IRC | 20:16 | |
| *** chlong_ has joined #openstack-lbaas | 20:28 | |
| *** rcernin has quit IRC | 20:36 | |
| *** ducttape_ has quit IRC | 20:52 | |
| *** ducttape_ has joined #openstack-lbaas | 21:05 | |
| ThiagoCMC | Does new LBaaS (Octavia?) works when using OVN features? Specially on top of OVS+DPDK? | 21:09 |
| xgerman | yes, we don’t care what networking is used | 21:10 |
| xgerman | so won’t benefit from DPDK right now | 21:10 |
| johnsom | xgerman +1 | 21:10 |
| ThiagoCMC | Nice! That's ok... | 21:10 |
| johnsom | Well, we do benefit from DPDK at the lower levels | 21:10 |
| johnsom | We just don't optimize all the way up yet. | 21:11 |
| ThiagoCMC | Hmmm... I see... | 21:11 |
| johnsom | In theory if they have OVS+DPDK our bits will move faster | 21:11 |
| ThiagoCMC | That's what I'm thinking... | 21:11 |
| xgerman | well, if you have benchmarks please share | 21:11 |
| johnsom | It's just inside our vms we aren't implementing DPDK virtio yet | 21:12 |
| johnsom | It's something being investigated in Pike | 21:12 |
| ThiagoCMC | Wait, LBaaS are VMs? | 21:12 |
| xgerman | yep | 21:12 |
| xgerman | Octavia is | 21:13 |
| ThiagoCMC | Oh, I thought it was something like an Neutron Agent... | 21:13 |
| johnsom | Using the Octavia driver yes, the load balancers are implemented using VMs for scalability, etc. | 21:13 |
| johnsom | The old network namespace HAproxy driver uses network nodes and neutron agent | 21:13 |
| ThiagoCMC | Hmmm... | 21:13 |
| ThiagoCMC | Octavia doesn't use HAProxy anymore? | 21:14 |
| johnsom | But that is a pretty limited driver, especially from a scalability perspective | 21:14 |
| johnsom | The Octavia driver does use HAProxy, inside one or more VMs | 21:14 |
| johnsom | service vms | 21:14 |
| ThiagoCMC | Oh, cool! I'm looking for an LBaaS that supports http/2.0 and looks like that HAProxy on Ubuntu 16.04 supports it... | 21:15 |
| ThiagoCMC | So, time to deploy Octavia! :-D | 21:15 |
| *** yamamoto has joined #openstack-lbaas | 21:15 | |
| johnsom | Nice | 21:16 |
| *** yamamoto has quit IRC | 21:18 | |
| *** yamamoto has joined #openstack-lbaas | 21:20 | |
| *** kobis has quit IRC | 21:23 | |
| *** gcheresh_ has quit IRC | 21:47 | |
| *** yamamoto has quit IRC | 21:47 | |
| openstackgerrit | Ankur proposed openstack/octavia master: Add v2 load balancer type and controllers https://review.openstack.org/423543 | 22:04 |
| *** ducttape_ has quit IRC | 22:05 | |
| *** ducttape_ has joined #openstack-lbaas | 22:05 | |
| xgerman | rm_work I need some python3 assistamce | 22:15 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Look for diskimage-builder not dib-utils when building amps https://review.openstack.org/443220 | 22:23 |
| *** chlong_ has quit IRC | 22:24 | |
| xgerman | of course it’s all inside some neutron code where it bombs | 22:24 |
| johnsom | xgerman he is sleeping now | 22:24 |
| xgerman | gosh | 22:25 |
| xgerman | guess I will ask him tomorrow | 22:25 |
| johnsom | Something I can put you on the wrong path about? | 22:25 |
| *** armax has quit IRC | 22:29 | |
| xgerman | sure, what am I dojng. wrong: | 22:30 |
| xgerman | https://www.irccloud.com/pastebin/ByMGX5r3/ | 22:30 |
| xgerman | my assessment is that it’s all neutron’s fault | 22:31 |
| *** yamamoto has joined #openstack-lbaas | 22:31 | |
| johnsom | xgerman https://github.com/openstack/neutron-lib/blob/master/neutron_lib/exceptions/__init__.py#L72 | 22:38 |
| johnsom | So, Just "NotFound" isn't implemented in neutron-lib | 22:38 |
| johnsom | Well, I guess it inherits stuff. | 22:39 |
| xgerman | mmh,yep | 22:40 |
| johnsom | It seems like assigning in like this is a bit odd with that object: e.msg = r.content | 22:41 |
| xgerman | mmh, maybe | 22:42 |
| xgerman | but it saus it bombs on the status_code | 22:43 |
| *** catintheroof has quit IRC | 22:44 | |
| xgerman | but I can six it | 22:44 |
| *** ducttape_ has quit IRC | 22:55 | |
| *** ducttape_ has joined #openstack-lbaas | 22:55 | |
| xgerman | ok, that increased it’s happiness — will clean it up tomorrow… | 22:56 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Look for diskimage-builder not dib-utils when building amps https://review.openstack.org/443220 | 23:16 |
| *** fnaval has quit IRC | 23:26 | |
| *** catintheroof has joined #openstack-lbaas | 23:28 | |
| *** ducttape_ has quit IRC | 23:35 | |
| *** ducttape_ has joined #openstack-lbaas | 23:38 | |
| johnsom | Well, on our DIB issue I have discovered a few things.... | 23:46 |
| johnsom | cross-repo depends-on doesn't trump the requirements.txt so we can't test the DIB fix directly for our issue. | 23:47 |
| johnsom | The other is that DIB has the same problem: http://logs.openstack.org/36/443336/1/check/gate-dib-dsvm-functests-ubuntu-trusty/5bc215d/console.html#_2017-03-08_22_23_39_752811 | 23:47 |
| johnsom | When we fixed it over there, DIB starts failing. | 23:48 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!