| openstackgerrit | Michael Johnson proposed openstack/octavia master: Add support for scoped tokens and default roles https://review.opendev.org/c/openstack/octavia/+/775957 | 00:24 |
|---|---|---|
| *** luksky has quit IRC | 00:33 | |
| *** eandersson has joined #openstack-lbaas | 01:35 | |
| *** eandersson has quit IRC | 01:42 | |
| *** eandersson has joined #openstack-lbaas | 01:42 | |
| *** zzzeek has quit IRC | 02:47 | |
| *** zzzeek has joined #openstack-lbaas | 02:50 | |
| *** mchlumsky0 has joined #openstack-lbaas | 02:55 | |
| *** mchlumsky has quit IRC | 02:59 | |
| *** priteau has quit IRC | 02:59 | |
| *** nicolasbock has quit IRC | 02:59 | |
| *** mchlumsky0 is now known as mchlumsky | 02:59 | |
| *** bcafarel has quit IRC | 03:00 | |
| *** nicolasbock has joined #openstack-lbaas | 03:00 | |
| *** bcafarel has joined #openstack-lbaas | 03:25 | |
| *** ramishra has quit IRC | 03:27 | |
| *** rcernin has quit IRC | 03:36 | |
| *** psachin has joined #openstack-lbaas | 03:42 | |
| *** rcernin has joined #openstack-lbaas | 04:16 | |
| *** ramishra has joined #openstack-lbaas | 04:21 | |
| rm_work | johnsom: you aware of any time recently where the cent8 amp image build was broken? | 05:06 |
| rm_work | and/or cgoncalves / gthiemonge | 05:07 |
| rm_work | i remember someone was working on something cent8 breakage related | 05:07 |
| gthiemonge | rm_work: https://review.opendev.org/c/openstack/octavia/+/775419 | 05:20 |
| gthiemonge | rm_work: it works but it's ugly :D | 05:21 |
| rm_work | what was the symptom | 05:21 |
| rm_work | of the broken image | 05:21 |
| gthiemonge | cgoncalves is working on a fix in devstack | 05:21 |
| rm_work | did the image just NOT BUILD? | 05:21 |
| rm_work | or did it build but fail to boot? | 05:21 |
| gthiemonge | yep, it doesn't build, pip updates setuptools and then breaks some selinux tools that are used by DIB | 05:22 |
| gthiemonge | see the comment in https://review.opendev.org/c/openstack/octavia/+/775419/4/devstack/plugin.sh | 05:22 |
| *** psachin has quit IRC | 05:50 | |
| rm_work | ah k | 06:03 |
| rm_work | not my issue | 06:03 |
| rm_work | but, found my issue :D | 06:03 |
| *** gcheresh has joined #openstack-lbaas | 06:12 | |
| *** vishalmanchanda has joined #openstack-lbaas | 06:34 | |
| openstackgerrit | Gregory Thiemonge proposed openstack/octavia-tempest-plugin master: WIP Fix two-node job configuration https://review.opendev.org/c/openstack/octavia-tempest-plugin/+/773888 | 06:45 |
| *** eandersson has quit IRC | 06:53 | |
| *** eandersson has joined #openstack-lbaas | 06:54 | |
| *** eandersson has quit IRC | 06:56 | |
| *** eandersson has joined #openstack-lbaas | 06:56 | |
| cgoncalves | the centos 8 amphora image build is an issue only on devstack environments | 07:16 |
| cgoncalves | the nightly amphora image build job looks fine (non-devstack env): https://zuul.openstack.org/builds?job_name=publish-openstack-octavia-amphora-image-centos8 | 07:20 |
| *** rcernin has quit IRC | 07:58 | |
| openstackgerrit | Gregory Thiemonge proposed openstack/octavia-tempest-plugin master: WIP Fix two-node job configuration https://review.opendev.org/c/openstack/octavia-tempest-plugin/+/773888 | 08:00 |
| *** rcernin has joined #openstack-lbaas | 08:02 | |
| *** rcernin has quit IRC | 08:07 | |
| *** rpittau|afk is now known as rpittau | 08:14 | |
| *** luksky has joined #openstack-lbaas | 08:17 | |
| *** yamamoto has quit IRC | 08:49 | |
| *** rcernin has joined #openstack-lbaas | 08:49 | |
| *** rcernin has quit IRC | 08:53 | |
| *** yamamoto has joined #openstack-lbaas | 08:57 | |
| luksky | hi. I tried to upgrade octavia from queens to victoria version. Well... this time all goes bad :/ | 09:14 |
| luksky | I ended with deleteing all database, and installing fresh db schema from victoria (7.1.0) version | 09:15 |
| *** yamamoto has quit IRC | 09:16 | |
| luksky | but the amphoras are recreated | 09:16 |
| luksky | over and over again | 09:17 |
| luksky | http://paste.openstack.org/show/802768/ | 09:19 |
| luksky | I can access amphoras (by ssh) for the time they exists. Tried 4 build of amphoras. | 09:20 |
| gthiemonge | luksky: do you have any ERROR lines in the health-manager logs? | 09:25 |
| luksky | the last few lines for all logs | 09:27 |
| luksky | http://paste.openstack.org/show/802769/ | 09:29 |
| luksky | after cleaning DB - there is no ERROR in health-manager, was fighthing with this until 1:30 CET | 09:30 |
| luksky | I think this is more amphora/access to amphora related - checked the SG - but the looks OK | 09:32 |
| gthiemonge | do you mean that you don't have any amphorae/LBs in the DB? | 09:32 |
| luksky | they are recreated every few minutes | 09:33 |
| luksky | it looks like this: | 09:35 |
| luksky | http://paste.openstack.org/show/802771/ | 09:35 |
| gthiemonge | spare amphorae, right? | 09:38 |
| *** ccamposr has joined #openstack-lbaas | 09:38 | |
| luksky | yes, there is no LB now, creating any ends in ERROR staty anyway | 09:39 |
| luksky | but most anoing is this: | 09:39 |
| luksky | http://paste.openstack.org/show/802772/ | 09:39 |
| luksky | there is connection with amphora, she reutrns with 200, health manager says OK and.... recreate (why, o why ?? :) ) | 09:40 |
| gthiemonge | I think you can find the reason in health-manager.log | 09:43 |
| gthiemonge | it should expain why the HM triggers a failover and creates the amphorae | 09:43 |
| *** yamamoto has joined #openstack-lbaas | 09:44 | |
| luksky | well, my eyes may be tired. Maybe don't see some obvius thing here: | 09:47 |
| luksky | http://paste.openstack.org/show/802773/ | 09:47 |
| luksky | this the tail -f for few last seconds from health-manager.log | 09:48 |
| gthiemonge | Stale amphora's id is: cd8ff0c8-58eb-46a4-a568-1d7076568ade | 09:49 |
| luksky | ok, which was created few minutes ago | 09:49 |
| luksky | and deleted... | 09:50 |
| luksky | and over and over again | 09:50 |
| gthiemonge | so it means the HM detected that the amp is not sending heartbeat messages | 09:50 |
| *** yamamoto has quit IRC | 09:50 | |
| gthiemonge | and it considers that there's an error, and tries to failover it | 09:51 |
| gthiemonge | Received packet from ('10.99.99.41', 11488) | 09:51 |
| gthiemonge | but it receives an heartbeat message 1min before the failover | 09:51 |
| luksky | indeed | 09:52 |
| luksky | that is the part, I don't understand | 09:52 |
| gthiemonge | luksky: what is the heartbeat_interval value in octavia.conf? | 09:52 |
| gthiemonge | default value is 10 | 09:52 |
| luksky | it is set to heartbeat_interval = 30 | 09:54 |
| *** yamamoto has joined #openstack-lbaas | 09:59 | |
| luksky | in http://paste.openstack.org/show/802774/ - here is log from amphora | 10:01 |
| gthiemonge | luksky: can you check the output of 'sudo journalctl -lf' in a newly created amphora? I'd like to see if amphora-agent is throwing some errors there | 10:02 |
| gthiemonge | luksky: if you don't find anything there, I would use tcpdump on the management port of the amphora (or mangement interface of the controllers) to verify that it sends an heartbeat package every 30 seconds | 10:04 |
| luksky | hm.. | 10:06 |
| luksky | there is some error | 10:06 |
| luksky | http://paste.openstack.org/show/802775/ | 10:07 |
| luksky | unable to resolve host amphora-6b4c446e-def4-413b-8581-536472ab10d5 | 10:08 |
| gthiemonge | nothing important here | 10:08 |
| luksky | ok... | 10:08 |
| gthiemonge | just wait a few minutes until the amphora is deleted | 10:08 |
| luksky | I have one controller now - turned off all other, to have all connections to one controller | 10:09 |
| gthiemonge | hmmm | 10:13 |
| gthiemonge | I think that if you turn off the other controllers, some heartbeat messages will be lost | 10:14 |
| luksky | Hm... they can by not consumed on amqp servers ? | 10:15 |
| gthiemonge | basically the amphora sends a UDP packet to one controller at a time (in round-robin) | 10:15 |
| gthiemonge | so if you set the heartbeat_interval to 10, and heartbeat_timeout to 60, it should work with 3 controllers | 10:16 |
| gthiemonge | you need to ensure that (heartbeat_interval * nb_controller < heartbeat_timeout) | 10:19 |
| luksky | ok | 10:19 |
| luksky | will adjust config for one controller and check | 10:22 |
| gthiemonge | the new config will only affect newly created amp, existing amp will still have 30 sec heartbeat_interval, but they will eventually disappear with the failovers | 10:23 |
| *** yamamoto has quit IRC | 10:25 | |
| *** yamamoto has joined #openstack-lbaas | 10:27 | |
| luksky | hm.. but I already have heartbeat_interval = 30 and heartbeat_timeout = 60, so assuming (heartbeat_interval * nb_controller < heartbeat_timeout) = 30 * 1 < 60 | 10:28 |
| gthiemonge | luksky: do you have only one entry in the controller_ip_port_list setting in octavia.conf? | 10:32 |
| luksky | yes, NOW i have one entry, I checked on (turned off) octavia's controllers tcpdump | 10:33 |
| luksky | earlier it was 3 entries, I think this was the reason. Checking, and let You now if reducing number of entries in controller_ip_port_list from 3 to 1 solved the issue. | 10:34 |
| *** yamamoto has quit IRC | 10:36 | |
| *** yamamoto has joined #openstack-lbaas | 10:37 | |
| luksky | gthiemonge: thank You ! the amphoras aren't flapping now | 10:37 |
| *** yamamoto has quit IRC | 10:37 | |
| gthiemonge | luksky: cool! | 10:38 |
| luksky | checking if LB are created correctly :D | 10:38 |
| *** yamamoto has joined #openstack-lbaas | 10:46 | |
| *** yamamoto has quit IRC | 11:07 | |
| *** yamamoto has joined #openstack-lbaas | 11:30 | |
| *** priteau has joined #openstack-lbaas | 11:34 | |
| *** yamamoto has quit IRC | 11:49 | |
| *** yamamoto has joined #openstack-lbaas | 11:53 | |
| *** njohnston has quit IRC | 12:16 | |
| rm_work | you also uploaded a newly built amphora image before the upgrade, right? | 12:21 |
| rm_work | highly recommend that, in the case that anything did change that wasn't backwards compatible (we guarantee amp <-> controller will be compatable +/- one release either way, but you're doing further) | 12:22 |
| rm_work | but, worst case they will fail over, and if the new image is available, then everything will be fine and shouldn't even have any missed connections | 12:22 |
| openstackgerrit | Adrian Vladu proposed openstack/octavia master: Add support for building arm64 images https://review.opendev.org/c/openstack/octavia/+/770937 | 12:47 |
| openstackgerrit | Adrian Vladu proposed openstack/octavia master: Add support for building arm64 images https://review.opendev.org/c/openstack/octavia/+/770937 | 12:48 |
| *** yamamoto has quit IRC | 12:50 | |
| *** mzylowski has joined #openstack-lbaas | 12:57 | |
| *** yamamoto has joined #openstack-lbaas | 13:05 | |
| *** yamamoto has quit IRC | 13:05 | |
| *** yamamoto has joined #openstack-lbaas | 13:05 | |
| *** yamamoto has quit IRC | 13:10 | |
| *** haleyb has joined #openstack-lbaas | 13:30 | |
| *** yamamoto has joined #openstack-lbaas | 13:37 | |
| luksky | all is working correctly, I will try to downgrade to queens version and simulate once more time upgrade. | 13:42 |
| luksky | rm_wark: can You confirm that these steps are enough | 13:44 |
| luksky | 1. stop health-manager and other octavia services | 13:44 |
| luksky | 2. install new code | 13:44 |
| luksky | 3. octavia-db-manage upgrade head | 13:44 |
| luksky | 4. change tag on ampora image from old (queens) to new (victoria) version | 13:44 |
| luksky | 5. start health-manager (and other octavia services | 13:44 |
| luksky | rm_work: sorry for misspelling | 13:45 |
| mchlumsky | rm_work we had some stein amphoras running for almost a week in an Ussuri octavia environment without anyone complaining but I was anxious to get them all upgraded. | 13:48 |
| openstackgerrit | Gregory Thiemonge proposed openstack/octavia-tempest-plugin master: WIP Fix two-node job configuration https://review.opendev.org/c/openstack/octavia-tempest-plugin/+/773888 | 13:50 |
| mzylowski | Hello, I'm looking for a way to restrict some networks for Octavia. In other words, I want to make impossible to create loadbalancer for given network (or even better network type). Also I want to disallow creation pool members from those networks. I tried to achieve this by policy.json file, but no luck yet (looks like network fields in policies | 13:53 |
| mzylowski | are not taken into account). Any thoughts? | 13:53 |
| *** yamamoto has quit IRC | 13:54 | |
| *** sapd1 has joined #openstack-lbaas | 13:56 | |
| *** yamamoto has joined #openstack-lbaas | 14:08 | |
| openstackgerrit | Gregory Thiemonge proposed openstack/octavia master: Fix health manager messages about failed failovers https://review.opendev.org/c/openstack/octavia/+/776467 | 14:23 |
| rm_work | that should be good luksky | 14:26 |
| rm_work | mzylowski: yes this is possible | 14:26 |
| luksky | ok, let now how goes this time :P | 14:26 |
| luksky | let know* | 14:26 |
| rm_work | mzylowski: check the config option "valid_vip_networks" | 14:29 |
| rm_work | that will limit the creation of loadbalancers to a specific set of network_ids | 14:29 |
| rm_work | as far as not allowing members from specific networks... I think you'd need to do that using ACLs or something in your network fabric | 14:30 |
| rm_work | Octavia LBs will function with members with any routable address -- so you could add some random VM from AWS to your openstack octavia LB if you want, or your home server... | 14:31 |
| rm_work | ah, if you are using flavors, you can also set valid_vip_networks for each flavor individually | 14:32 |
| mzylowski | Okey I got it, but valid_vip_networks seems not be enough for me, because we want to allow our users to create custom private networks and those networks can be used with loadbalancers. I want to block some public network and allow all the rest, even those not created yet. | 14:37 |
| rm_work | If the users are not allowed to create ports on the network, i believe that will prevent it? So long as you don't have "allow invisible resource use" on | 14:39 |
| rm_work | Though... Hmm we might only check read access | 14:39 |
| rm_work | johnsom may know better | 14:39 |
| rm_work | He may be awake soon | 14:40 |
| mzylowski | Okey, great :) Thanks for Your help. | 14:40 |
| haleyb | rm_work: hey Adam, can you take a look at https://review.opendev.org/c/openstack/octavia/+/774426 - I think that should get out tempest gate green (it's the bottom of a depends-on chain) | 14:46 |
| rm_work | Sorry I'm in bed but I can look when I wake up | 14:46 |
| rm_work | Which is... Uhh | 14:46 |
| rm_work | In an hour and ten minutes, FML | 14:47 |
| rm_work | Back in a bit :D | 14:47 |
| haleyb | rm_work: thanks, i was just looking at my backlog and that was the blocker | 14:48 |
| *** yamamoto has quit IRC | 14:57 | |
| openstackgerrit | Gregory Thiemonge proposed openstack/octavia master: Fix health manager messages about failed failovers https://review.opendev.org/c/openstack/octavia/+/776467 | 15:06 |
| *** ihti[m] has joined #openstack-lbaas | 15:13 | |
| johnsom | luksky Yes, with a heartbeat_interval of 30 and heartbeat_timeout of 60 and two out of three controllers down, you are going to have the amphora marked as failed and a failover attempted. With the default interval of 10, it would have been fine, but with 30 you aren't giving it a chance to hit a healthy HM instance. | 15:28 |
| openstackgerrit | Takashi Kajinami proposed openstack/octavia-dashboard master: Disable Load Balancers panel when Octavias service is not deployed https://review.opendev.org/c/openstack/octavia-dashboard/+/766878 | 15:36 |
| openstackgerrit | Takashi Kajinami proposed openstack/octavia-dashboard master: Disable Load Balancers panel when Octavias service is not deployed https://review.opendev.org/c/openstack/octavia-dashboard/+/766878 | 15:37 |
| *** EmilienM has joined #openstack-lbaas | 15:57 | |
| EmilienM | hello everyone, I'm currently seeing https://bugs.launchpad.net/octavia/+bug/1706836 in my environment, I run Ussuri, any idea what it could be? | 15:58 |
| openstack | Launchpad bug 1706836 in octavia "Amphora agent 404 No suitable network interface found" [High,Triaged] | 15:58 |
| EmilienM | cgoncalves: ^ fyi | 15:58 |
| cgoncalves | gthiemonge, ^^ fyi :P | 15:59 |
| johnsom | Launchpad hasn't been used for Octavia for years | 15:59 |
| johnsom | https://bugs.launchpad.net/octavia | 15:59 |
| EmilienM | right | 16:00 |
| EmilienM | I just found the bug on search engine | 16:00 |
| johnsom | Yeah, the storyboard migration and launchpad doesn't close out the bugs when it's disabled in launchpad and you get migrated | 16:01 |
| *** vishalmanchanda has quit IRC | 16:02 | |
| EmilienM | any idea what it could be? | 16:03 |
| johnsom | I know there was a recent issue with NetworkManager doing bad things. It might be related if you are using a CentOS/RHEL image. | 16:04 |
| johnsom | It would be best if you could provide the amphora logs for the issue. | 16:04 |
| EmilienM | I use RHEL | 16:05 |
| EmilienM | ok I'll try to get them | 16:05 |
| *** mzylowski has quit IRC | 16:43 | |
| *** luksky has quit IRC | 17:05 | |
| EmilienM | johnsom: how do you get these logs? | 17:08 |
| johnsom | EmilienM If you are on OSP they are in with the other container logs. If it's upstream it is in log/controller/logs/octavia-amphora_log.txt | 17:43 |
| gthiemonge | johnsom: I'm working with EmilienM, it looks like the lookup of the mac address in the amphora doesn't work | 17:46 |
| gthiemonge | https://opendev.org/openstack/octavia/src/branch/master/octavia/amphorae/backends/agent/api_server/plug.py#L218-L228 | 17:46 |
| gthiemonge | ^ it fails but the interface exists | 17:46 |
| *** rpittau is now known as rpittau|afk | 17:47 | |
| johnsom | Oh! This was a regression in pyroute2 I bet. I reported that and got it fixed. One minute. | 17:47 |
| EmilienM | :-O | 17:47 |
| johnsom | https://github.com/svinota/pyroute2/issues/724 | 17:48 |
| gthiemonge | lol | 17:48 |
| EmilienM | https://github.com/openstack/octavia/commit/51b93c00223ef5529d98174ab422f99ac7c7570e | 17:48 |
| EmilienM | seems like you have a workaround | 17:48 |
| johnsom | 0.5.13 broken it all, 0.5.14 fixed it again | 17:48 |
| EmilienM | let me see what version of pyroute2 I have in the amphora | 17:49 |
| johnsom | 0.5.13 should be on a deny list somewhere. | 17:49 |
| EmilienM | yeah I have python3-pyroute2-0.5.13-1.el8ost.noarch | 17:51 |
| johnsom | That build is a train wreck. We need to fix that. | 17:51 |
| cgoncalves | https://github.com/openstack/requirements/blob/stable/train/global-requirements.txt#L254 | 17:52 |
| cgoncalves | though not that OSP would care for g-r.txt ... | 17:53 |
| *** luksky has joined #openstack-lbaas | 17:54 | |
| johnsom | Hmm, that is interesting. I could have sworn GR was updated to deny that build | 17:54 |
| johnsom | There was a long discussion about it in the requirements channel. | 17:54 |
| cgoncalves | EmilienM, how are you building the amphora image? | 18:06 |
| EmilienM | cgoncalves: I was using the OSP image that I got from rpm | 18:06 |
| EmilienM | but now I'm using the one from https://images.rdoproject.org/octavia/master/ | 18:06 |
| EmilienM | we'll see how that works :D | 18:07 |
| EmilienM | in OSP there is a package named octavia-amphora-image-x86_64, I was using that | 18:07 |
| EmilienM | the image was old from November of 2020, and it seems to have the wrong version of pyroute2 | 18:07 |
| cgoncalves | EmilienM, master amphora image running on a train octavia deployment is strongly not recommended and more likely to fail spectacularly | 18:08 |
| EmilienM | I run OSP17 | 18:08 |
| EmilienM | which is Ussuri I guess? | 18:09 |
| cgoncalves | master Victoria I would hope | 18:09 |
| EmilienM | I guess I can download an ussuri image if it fails | 18:09 |
| EmilienM | it's hard to tell which version of Octavia I run with OSP17 :/ | 18:09 |
| EmilienM | actually, I can tell on which commit was built the rpm | 18:10 |
| cgoncalves | sorry, I meant to say master Wallaby. OSP 17 will be based on Wallaby | 18:10 |
| EmilienM | openstack-octavia-api-6.1.0-0.20200828161906.0b1d8dd.el8ost.noarch | 18:10 |
| EmilienM | that's my version of Octavia | 18:10 |
| EmilienM | which is a commit in Aug 26th last year | 18:10 |
| EmilienM | so it's Victoria IIUC | 18:11 |
| EmilienM | https://releases.openstack.org/ | 18:11 |
| johnsom | 6.x.x is Ussuri | 18:12 |
| cgoncalves | OK. I really thought OSP 17 was tracking master since it will be based on Wallaby | 18:12 |
| johnsom | https://docs.openstack.org/releasenotes/octavia/ussuri.html | 18:12 |
| johnsom | 17 should be tracking Wallaby | 18:13 |
| haleyb | the tip of rhos-17.0-trunk-patches has a feb-2021 change | 18:13 |
| haleyb | Merge "Fix pools going into ERROR when updating the pool" | 18:13 |
| haleyb | and we're talking downstream on the upstream channel if it matters | 18:14 |
| EmilienM | python3-pyroute2-0.5.13-1.1.el8.noarch on master | 18:14 |
| EmilienM | which seems to have worked, my deployment is running fine | 18:15 |
| haleyb | EmilienM: i see octavia 7.1.0 numbered rpms downstream for 17 in the latest puddle, which was last November | 18:17 |
| *** ccamposr has quit IRC | 18:22 | |
| *** sapd1 has quit IRC | 18:28 | |
| gthiemonge | https://zuul.openstack.org/builds?job_name=openstack-tox-cover&project=openstack%2Foctavia | 18:36 |
| gthiemonge | I think that merging the SCTP commit has broken the cover job | 18:37 |
| johnsom | I think my patch on that list was just an issue with the oslo policy bump causing follow on issues. I can't speak for the others | 18:37 |
| gthiemonge | I'm checking | 18:38 |
| gthiemonge | johnsom: the SCTP commit passed the cover job when the threshold was 90% | 18:40 |
| gthiemonge | and the gates don't run openstack-tox-cover | 18:40 |
| johnsom | I see 91%, but... that is still below the 92% the other patch flagged. | 18:41 |
| johnsom | Oh, that is interesting.... So, SCTP passed check on Jan 29. 92% merged on Jan 30. Then SCTP merged Feb 17th. So, yeah, since cover isn't in the gate it didn't catch that it was now under the limit. | 18:45 |
| johnsom | Same on the cores that reviewed that for letting in un-covered code. Grin. But, maybe we need to add cover to the gate too. | 18:46 |
| *** xarlos has joined #openstack-lbaas | 18:47 | |
| johnsom | Neat, so gthiemonge Can you spin a patch to fix the coverage in SCTP? | 18:47 |
| gthiemonge | sure, I have some ideas ;-) | 18:50 |
| johnsom | I haven't read through all of this yet, but ... This promises to be an adventure: https://docs.sqlalchemy.org/en/14/changelog/migration_20.html | 20:52 |
| *** dougwig has quit IRC | 21:08 | |
| *** dougwig has joined #openstack-lbaas | 21:09 | |
| *** gcheresh has quit IRC | 21:18 | |
| openstackgerrit | Gregory Thiemonge proposed openstack/octavia master: Add test coverage for SCTP health checker script https://review.opendev.org/c/openstack/octavia/+/776535 | 21:20 |
| johnsom | Wow, thanks Greg | 21:20 |
| haleyb | someone's up late | 21:21 |
| gthiemonge | ;-) | 21:22 |
| johnsom | OMG, I just don't know what to say. | 21:38 |
| johnsom | https://www.irccloud.com/pastebin/180NtWOE/ | 21:39 |
| johnsom | An import, a long constants string, all for "W" | 21:39 |
| johnsom | Even if I created a "WALLABY" constant in our constants file it would be less bytes | 21:40 |
| haleyb | shouldn't it be the other way around? W = "WALLABY" crikey | 21:53 |
| johnsom | https://github.com/openstack/oslo.log/blob/master/oslo_log/versionutils.py#L157 | 21:54 |
| johnsom | I think the deprecation message would read better with Wallaby than just "w" | 21:55 |
| johnsom | But.... whatever | 21:55 |
| haleyb | you should propose a patch for X = 'Xena' to add the sanity back | 21:55 |
| johnsom | I have other bike sheds | 21:56 |
| haleyb | :) and you must have shaved your yak already | 21:56 |
| johnsom | Already sent the yak down the river in the enchanted canoe. | 22:08 |
| haleyb | might need to get that yak back to fix the docs job failure, we just can't win | 22:11 |
| johnsom | Really? lol | 22:13 |
| johnsom | Well, my checked out version is rendering my docs, so I guess I will not rebase | 22:13 |
| haleyb | i'm looking at it but it's weird, and will have to merge with greg's change to make the gate green | 22:16 |
| *** yamamoto has joined #openstack-lbaas | 22:18 | |
| haleyb | the file in question hasn't changed in a year, doc/source/install/install-ubuntu.rst so sphinx or something moved forward. probably just a trivial formatting error that went from warning to error | 22:18 |
| johnsom | haleyb Hmmm, might need this: https://review.opendev.org/c/x/wsme/+/763422 | 22:22 |
| johnsom | I can rage merge that if you decide you need it | 22:23 |
| johnsom | Oh, it looks like it is just the "console", so yeah, change it to bash will probably work | 22:24 |
| haleyb | johnsom: there's other occurences of console in that file too, :-/ | 22:25 |
| haleyb | doc/source/install/install-ubuntu.rst:269:Could not lex literal_block as "console". Highlighting skipped. | 22:25 |
| haleyb | maybe just change them all to bash | 22:28 |
| rm_work | ok what did i need to look at, slept through by accident | 22:42 |
| haleyb | rm_work: oh, the one i had asked about was https://review.opendev.org/c/openstack/octavia/+/774426 but i doubt it will merge now, damn gate | 22:44 |
| johnsom | I will have another RBAC spin here soon-ish | 22:45 |
| haleyb | johnsom: changing that to bash seems to fix it, strange since vim isn't showing all those blocks correctly so i figured it was some line above there. | 22:46 |
| haleyb | usually the '.. foo::' are in yellow | 22:46 |
| * haleyb skeedaddles | 22:47 | |
| rm_work | hmm k | 22:48 |
| *** rcernin has joined #openstack-lbaas | 23:03 | |
| *** mchlumsky has quit IRC | 23:09 | |
| *** rcernin has quit IRC | 23:11 | |
| *** rcernin has joined #openstack-lbaas | 23:12 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Add support for scoped tokens and default roles https://review.opendev.org/c/openstack/octavia/+/775957 | 23:17 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Add support for scoped tokens and default roles https://review.opendev.org/c/openstack/octavia/+/775957 | 23:32 |
| johnsom | weee, lower constraints chains | 23:32 |
| johnsom | I really should spend the time to build a VM that is clean enough to run that locally. | 23:32 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Add support for scoped tokens and default roles https://review.opendev.org/c/openstack/octavia/+/775957 | 23:39 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Add support for scoped tokens and default roles https://review.opendev.org/c/openstack/octavia/+/775957 | 23:49 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!