| *** ramishra has joined #openstack-lbaas | 02:19 | |
| *** psachin has joined #openstack-lbaas | 03:35 | |
| *** ricolin has joined #openstack-lbaas | 04:12 | |
| *** ricolin_ has joined #openstack-lbaas | 04:15 | |
| *** ricolin has quit IRC | 04:19 | |
| *** gcheresh_ has joined #openstack-lbaas | 06:27 | |
| *** pcaruana has joined #openstack-lbaas | 06:43 | |
| *** maciejjozefczyk has joined #openstack-lbaas | 06:46 | |
| *** gcheresh_ has quit IRC | 06:51 | |
| openstackgerrit | Vishal Manchanda proposed openstack/octavia-dashboard master: Switch to official Ussuri jobs https://review.opendev.org/688336 | 07:02 |
|---|---|---|
| *** tesseract has joined #openstack-lbaas | 07:04 | |
| *** rcernin has quit IRC | 07:11 | |
| *** rpittau|afk is now known as rpittau | 07:40 | |
| *** ivve has joined #openstack-lbaas | 07:46 | |
| *** ricolin_ is now known as ricolin | 08:41 | |
| frickler | is anyone still fixing py3 issues in stable neutron-lbaas versions like this one in rocky? http://paste.openstack.org/show/783587/ | 08:43 |
| cgoncalves | frickler, resources shifted to octavia long ago. barely a handful of folks still looking after neutron-lbaas and even less py3 in rocky. nevertheless, bug fixes are welcome to stein and earlier versions | 08:51 |
| openstackgerrit | Merged openstack/octavia-dashboard master: Switch to official Ussuri jobs https://review.opendev.org/688336 | 08:51 |
| frickler | cgoncalves: o.k., I created a story for it as the first step. we'll probably also discuss speeding up the migration to octavia internally https://storyboard.openstack.org/#!/story/2006714 | 08:54 |
| cgoncalves | frickler, thanks for capturing this in storyboard! | 08:55 |
| cgoncalves | frickler, some resources you might find helpful for the migration: | 08:57 |
| cgoncalves | https://wiki.openstack.org/wiki/Neutron/LBaaS/Deprecation | 08:57 |
| cgoncalves | https://www.youtube.com/watch?v=jj4KMJPA0Pk | 08:57 |
| *** trident has quit IRC | 09:23 | |
| *** trident has joined #openstack-lbaas | 09:28 | |
| *** salmankhan has joined #openstack-lbaas | 09:35 | |
| *** salmankhan has quit IRC | 09:37 | |
| *** ricolin has quit IRC | 09:41 | |
| openstackgerrit | Maciej Józefczyk proposed openstack/octavia-tempest-plugin master: Add an option to reuse connections https://review.opendev.org/672976 | 09:50 |
| openstackgerrit | Maciej Józefczyk proposed openstack/octavia-tempest-plugin master: Handle NotImplemented flavor exception https://review.opendev.org/676135 | 09:54 |
| openstackgerrit | Jens Harbott (frickler) proposed openstack/neutron-lbaas stable/rocky: WIP: Misc Python 3.6 fixes https://review.opendev.org/688379 | 10:32 |
| *** maciejjozefczyk has quit IRC | 10:40 | |
| *** maciejjozefczyk has joined #openstack-lbaas | 10:41 | |
| *** gcheresh_ has joined #openstack-lbaas | 10:41 | |
| openstackgerrit | Ann Taraday proposed openstack/octavia master: Add option to set default ssl ciphers in haproxy https://review.opendev.org/685337 | 10:46 |
| *** ioni has joined #openstack-lbaas | 10:51 | |
| ioni | hi guys | 10:51 |
| ioni | openstack ansible uses http://tarballs.openstack.org/octavia/test-images/test-only-amphora-x64-haproxy-ubuntu-bionic.qcow2 as image | 10:51 |
| ioni | is this image alright for production ? | 10:51 |
| jrosser | ioni: openstack-ansible uses that image for CI tests | 11:06 |
| ioni | for production, what should i use? | 11:06 |
| jrosser | you shoukd probably build your own | 11:07 |
| ioni | is there any documentation related to that? | 11:08 |
| jrosser | from an OSA perspective there is this https://docs.openstack.org/openstack-ansible-os_octavia/latest/configure-octavia.html#building-octavia-images | 11:09 |
| jrosser | but i think that quite likley needs updating | 11:10 |
| jrosser | i'm sure there are octavia docs too | 11:10 |
| ioni | jrosser, thanks | 11:11 |
| *** gcheresh_ has quit IRC | 11:23 | |
| cgoncalves | ioni, https://docs.openstack.org/octavia/latest/admin/amphora-image-build.html | 11:26 |
| ioni | cgoncalves, thanks | 11:28 |
| *** pcaruana has quit IRC | 11:34 | |
| *** pcaruana has joined #openstack-lbaas | 11:41 | |
| openstackgerrit | Carlos Goncalves proposed openstack/octavia master: DNM CentOS7 gate test https://review.opendev.org/665464 | 11:46 |
| openstackgerrit | Jens Harbott (frickler) proposed openstack/neutron-lbaas stable/stein: Fix lb stats model https://review.opendev.org/688391 | 11:52 |
| frickler | cgoncalves: ^^ the fix is easy it seems, but adding a unit test I discovered that the verification was also broken for py2 :-/ | 11:53 |
| openstackgerrit | Gregory Thiemonge proposed openstack/octavia-tempest-plugin master: WIP Use VM as client in traffic operation scenario tests https://review.opendev.org/687562 | 11:53 |
| sapd1_ | Hi. Anyone help me. I don't know how to change the query in Listener Models after add type column to listener_cidr table. | 12:01 |
| frickler | cgoncalves: does the migration guide mentioned in https://wiki.openstack.org/wiki/Neutron/LBaaS/Deprecation#How_do_I_migrate_to_Octavia.3F actually exist? I cannot find it in the current octavia docs | 12:03 |
| cgoncalves | frickler, whoops! IIRC we do not have a doc there, no. perhaps that promised document is the README file in https://github.com/openstack/neutron-lbaas/tree/stable/stein/tools/nlbaas2octavia | 12:09 |
| cgoncalves | frickler, one other migration tool is https://github.com/nmagnezi/nlbaas2octavia-lb-replicator | 12:10 |
| frickler | cgoncalves: o.k., I'll look at both, thanks for the links | 12:12 |
| cgoncalves | frickler, the former tool is periodically tested in CI -- http://zuul.openstack.org/builds?pipeline=periodic&project=openstack%2Fneutron-lbaas | 12:15 |
| frickler | cgoncalves: yes, I saw that job when looking into updating for py36 testing earlier. it does look a bit unstable, but will be a good guide for internal testing when we tackle the migration | 12:27 |
| openstackgerrit | Carlos Goncalves proposed openstack/octavia master: DNM CentOS7 gate test https://review.opendev.org/665464 | 12:44 |
| openstackgerrit | Jens Harbott (frickler) proposed openstack/neutron-lbaas stable/stein: Fix lb stats model https://review.opendev.org/688391 | 12:46 |
| openstackgerrit | Ann Taraday proposed openstack/octavia master: [WIP] Jobboard based controller https://review.opendev.org/647406 | 12:57 |
| *** Guest31368 has joined #openstack-lbaas | 12:58 | |
| *** Guest31368 is now known as redrobot_ | 13:00 | |
| *** redrobot_ is now known as redrobot | 13:01 | |
| *** AustinR has quit IRC | 13:01 | |
| *** AustinR has joined #openstack-lbaas | 13:15 | |
| *** openstackstatus has quit IRC | 13:15 | |
| *** AustinR has quit IRC | 13:22 | |
| *** AustinR has joined #openstack-lbaas | 13:23 | |
| *** gcheresh_ has joined #openstack-lbaas | 13:33 | |
| openstackgerrit | Carlos Goncalves proposed openstack/octavia master: DNM CentOS7 gate test https://review.opendev.org/665464 | 13:34 |
| openstackgerrit | Merged openstack/octavia master: Clean up requirements.txt from tox.ini https://review.opendev.org/687978 | 13:43 |
| *** gcheresh_ has quit IRC | 14:00 | |
| *** gcheresh_ has joined #openstack-lbaas | 14:01 | |
| *** ricolin has joined #openstack-lbaas | 14:13 | |
| *** gcheresh_ has quit IRC | 14:24 | |
| *** maciejjozefczyk has quit IRC | 15:05 | |
| *** ivve has quit IRC | 15:05 | |
| *** openstackstatus has joined #openstack-lbaas | 16:03 | |
| *** ChanServ sets mode: +v openstackstatus | 16:03 | |
| *** rpittau is now known as rpittau|afk | 16:06 | |
| *** ramishra has quit IRC | 16:36 | |
| *** AlexStaf has joined #openstack-lbaas | 16:59 | |
| *** AustinR has quit IRC | 17:03 | |
| *** schaney has quit IRC | 17:07 | |
| *** AlexStaf has quit IRC | 17:12 | |
| *** AlexStaf has joined #openstack-lbaas | 17:14 | |
| guilhermesp | hey there! Is this fixed? http://paste.openstack.org/show/783716/ | 17:16 |
| guilhermesp | this is stable/queens | 17:16 |
| johnsom | rm_work FYI, https://storyboard.openstack.org/#!/story/2006719 | 17:17 |
| guilhermesp | when users attempet to list lb from horizon, it spanws this in horizon logs repeatedly | 17:17 |
| guilhermesp | I can list lbs from cli btw | 17:17 |
| johnsom | guilhermesp Check the version of openstacksdk you have installed on your horizon system. It seems like there is a bad version of openstacksdk installed. | 17:18 |
| guilhermesp | thanks for the reply johnsom I will take a look! | 17:20 |
| johnsom | It's a guess. I haven't seen that error before. | 17:21 |
| guilhermesp | yeah, at this point, i think is the best clue tbh | 17:22 |
| johnsom | Hmm, it also kind of looks like you have the master version of the dashboard installed instead of the queens version. | 17:23 |
| johnsom | guilhermesp Yeah, I think that might be your problem. You have master/train octavia-dashboard. So it probably doesn't align to the openstacksdk version installed. | 17:25 |
| *** gmann is now known as gmann_afk | 17:25 | |
| *** tesseract has quit IRC | 17:25 | |
| johnsom | Looks like it could be the stein version too, that line hasn't changed in a while. | 17:26 |
| guilhermesp | yeah I'm cheking right now the exact version we are running | 17:27 |
| *** ricolin has quit IRC | 17:33 | |
| guilhermesp | yeah sorry johnsom our hjorizon instances runs stein | 17:34 |
| guilhermesp | let me check the openstacksdk version running in it | 17:34 |
| johnsom | Should be 0.27.0 according to the releases page: https://releases.openstack.org/stein/index.html#openstacksdk | 17:35 |
| jrosser | guilhermesp: you are probably still using the queens constraints file even though moving the horizon service sha forward | 17:36 |
| guilhermesp | openstacksdk==0.26.0 yeah we are running the wrong one | 17:36 |
| johnsom | That is pretty close though. Not sure I would expect an error with that combo. hmmm | 17:37 |
| guilhermesp | good point jrosser | 17:37 |
| jrosser | you can probably make a config override just for the horizon role to fix that | 17:37 |
| guilhermesp | yeah that would do it jrosser | 17:38 |
| guilhermesp | thanks in advance guys! | 17:38 |
| *** AustinR has joined #openstack-lbaas | 17:40 | |
| *** maciejjozefczyk has joined #openstack-lbaas | 17:40 | |
| guilhermesp | argh, seems that upgrading the sdk didn't fix | 17:45 |
| *** maciejjozefczyk has quit IRC | 17:49 | |
| johnsom | Yeah, I was not convinced that 0.26.0->0.27.0 would have a fix. Let me look up the patch that changed that line to see if that gives any clues as to why you are seeing that. | 17:52 |
| *** ivve has joined #openstack-lbaas | 17:53 | |
| guilhermesp | appreciate that jrosser | 17:53 |
| guilhermesp | ops | 17:53 |
| guilhermesp | johnsom ( and jrosser too ) :P | 17:53 |
| rm_work | ahh johnsom i'll look at that today if i have time | 17:55 |
| rm_work | kinda lame >_< | 17:55 |
| rm_work | would expect that to work | 17:56 |
| johnsom | Yeah, not sure what is up with that. Just moved the bug over | 17:56 |
| rm_work | woah that bug isn't new? old enough to migrate from launchpad? >_> | 17:56 |
| rm_work | just noticed that | 17:56 |
| johnsom | Someone opened it in launchpad. There are tricks that still allow it unfortunately. Like open it against another project, then re-assign. | 17:57 |
| johnsom | guilhermesp I am puzzled... https://github.com/openstack/openstacksdk/blob/0.27.0/openstack/load_balancer/v2/_proxy.py#L68 | 18:00 |
| guilhermesp | not sure but we enabling a flat to be able to use octavia v1 api | 18:01 |
| johnsom | Well, v2 must be enabled for horizon to work | 18:02 |
| johnsom | v1 was only used by neutron-lbaas | 18:02 |
| *** gcheresh_ has joined #openstack-lbaas | 18:03 | |
| guilhermesp | i will try to toggle that off and see what happens | 18:04 |
| *** psachin has quit IRC | 18:05 | |
| johnsom | guilhermesp Do you have both octavia-dashboard and neutron-lbaas-dashboard installed in horizon? | 18:06 |
| rm_work | ugh I did a really bad QA job on batch members, that was an obvious edge case and I just didn't test for it | 18:13 |
| rm_work | oh, wait, no | 18:15 |
| rm_work | I take that back, i totally did test for this and it should work fine | 18:15 |
| rm_work | though I could have named the test better | 18:15 |
| rm_work | ah i see how that error could happen | 18:22 |
| rm_work | if you update with [] on a pool that already has no members | 18:27 |
| rm_work | oh, nm, more common than that. | 18:28 |
| guilhermesp | seems that we dont johnsom http://paste.openstack.org/show/783718/ | 18:42 |
| guilhermesp | but I found this in neutron.conf, which I think we dont need | 18:43 |
| guilhermesp | service_plugins=router,neutron_vpnaas.services.vpn.plugin.VPNDriverPlugin,neutron_lbaas.services.loadbalancer.plugin.LoadBalancerPluginv2 | 18:43 |
| guilhermesp | specifically neutron_lbaas.services.loadbalancer.plugin.LoadBalancerPluginv2 | 18:43 |
| *** gcheresh_ has quit IRC | 18:45 | |
| rm_work | right, that shouldn't be there if you're using Octavia directly | 19:05 |
| openstackgerrit | Adam Harwell proposed openstack/octavia-lib master: Batch member update needs pool_id explicitly https://review.opendev.org/688546 | 19:33 |
| johnsom | guilhermesp I see you have 3.0.0 deployed for octavia-dashboard. You will really want to grab 3.0.1 as it has an important patch for a 403 error on creating load balancers. | 19:36 |
| openstackgerrit | Brian Haley proposed openstack/octavia master: Stop testing python 2 https://review.opendev.org/687370 | 19:43 |
| guilhermesp | johnsom: seems that we were missing that commit https://github.com/openstack/octavia/commit/aa4f3aba179020c0e9e42e8c01e97d30749787e1 | 19:45 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Fix batch member update error on empty change list https://review.opendev.org/688548 | 19:47 |
| rm_work | johnsom: ^^ | 19:48 |
| rm_work | not sure if we can backport the fix :/ | 19:48 |
| rm_work | can we do backports at all on octavia-lib? | 19:48 |
| johnsom | I will look after lunch | 19:49 |
| johnsom | Sure, lib has the same stable rules | 19:49 |
| zigo | Hi guys. After thinking a lot about the Amphora image, I found that a lot of it can just go into the octavia-amphora package ... | 19:50 |
| rm_work | zigo: ah, i had on my list for this week to look at the debian image building in DIB -- looks like it shouldn't be too bad to make it build based on the official debian cloud image as a base | 19:58 |
| zigo | rm_work: Looking at things, really, there's almost nothing that belong to image building, everything can be packaged. | 19:59 |
| rm_work | yeah, should be very little | 19:59 |
| zigo | The only thing that belongs there is "update-rc.d haproxy disable" | 20:00 |
| rm_work | just make sure the DIB reqs are installed and run the image build command | 20:00 |
| zigo | Everything else, I've pushed it to the octavia-agent package itself. | 20:00 |
| zigo | I uploaded this packaging change to Experimental, then this will be pushed to Sid this week, when Train gets released. | 20:01 |
| *** gmann_afk is now known as gmann | 20:01 | |
| zigo | Then I'll push a merge request to the Debian image builder... | 20:01 |
| zigo | rm_work: https://salsa.debian.org/openstack-team/services/octavia/commit/fe6be9d39974c4c1d76c3094c4d83c94922d4271 | 20:03 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Fix batch member update error on empty change list https://review.opendev.org/688548 | 20:03 |
| rm_work | yeah it'd really be much better to not have you manage our elements separately and duplicate that work / have to keep up with changes | 20:03 |
| johnsom | rm_work Why are you passing pool_id here? https://review.opendev.org/#/c/688548/2/octavia/api/drivers/amphora_driver/v2/driver.py | 20:25 |
| zigo | rm_work: Don't you think you'd better ask each and every distro provider to package octavia-agent properly? | 20:25 |
| *** AustinR has quit IRC | 20:25 | |
| rm_work | johnsom: to solve the bug | 20:25 |
| rm_work | zigo: no i do not | 20:26 |
| *** pcaruana has quit IRC | 20:26 | |
| johnsom | rm_work can't you just no-op if there are no members to pull the pool_id off of? | 20:26 |
| *** AustinR has joined #openstack-lbaas | 20:26 | |
| rm_work | johnsom: at that point it's not yet a noop if members is empty | 20:26 |
| zigo | rm_work: ?!? | 20:26 |
| rm_work | johnsom: deleted members don't go in that list, so it might need to delete some, we don't know yet | 20:26 |
| rm_work | that' | 20:26 |
| rm_work | *that's determined by the next stage | 20:27 |
| johnsom | Hmm, seems like we should sent it pre-calculated | 20:27 |
| johnsom | Yeah, ok, let me think about this.... | 20:27 |
| rm_work | zigo: i believe that no distro should diverge from the officially proscribed build method, and duplicate a ton of work just to result in a more fragile build process | 20:28 |
| rm_work | we maintain all of this stuff in elements and using a flexible multi-os build system for a reason | 20:28 |
| zigo | rm_work: What you should be doing is provide the things of the elements to the distro so that they package stuff. | 20:29 |
| rm_work | i guess we agree to disagree | 20:29 |
| zigo | rm_work: Because of what you did, I didn't even realize that these things are belonging to a package. | 20:29 |
| zigo | Now I do understand my mistake. | 20:29 |
| zigo | Trying to solve it at the amphora image build is the wrong approach. | 20:29 |
| rm_work | if you weren't trying to replicate the whole build process outside of our build system, it wouldn't be a problem either | 20:30 |
| zigo | It would ... | 20:30 |
| rm_work | trying to replicate all of our elements in your own build system is the wrong approach | 20:30 |
| zigo | Things that belongs in a package belongs in a package, that's it ! :) | 20:30 |
| rm_work | i think we fundamentally disagree | 20:31 |
| rm_work | so i guess i'll leave you to it, good luck | 20:31 |
| zigo | :) | 20:31 |
| zigo | rm_work: What you are doing in your elements really is packaging work, you just don't realize it. | 20:34 |
| zigo | Then it becomes a big hack... | 20:34 |
| johnsom | I think I disagree as well. | 20:46 |
| *** rcernin has joined #openstack-lbaas | 20:57 | |
| *** goldyfruit has joined #openstack-lbaas | 21:20 | |
| *** vesper11- has joined #openstack-lbaas | 21:53 | |
| *** vesper11 has quit IRC | 21:53 | |
| *** rcernin has quit IRC | 21:59 | |
| openstackgerrit | Brian Haley proposed openstack/octavia master: Stop testing python 2 https://review.opendev.org/687370 | 22:00 |
| *** trident has quit IRC | 22:04 | |
| *** trident has joined #openstack-lbaas | 22:07 | |
| *** ivve has quit IRC | 22:49 | |
| openstackgerrit | Merged openstack/octavia master: Use bandit block in tox.ini https://review.opendev.org/687963 | 22:50 |
| *** rcernin has joined #openstack-lbaas | 23:07 | |
| *** tkajinam has joined #openstack-lbaas | 23:17 | |
| *** tkajinam_ has joined #openstack-lbaas | 23:17 | |
| *** tkajinam has quit IRC | 23:21 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!