| *** goldyfruit_ has joined #openstack-lbaas | 00:00 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Fix openstack port show content handling https://review.opendev.org/682327 | 00:08 |
|---|---|---|
| johnsom | There is an awk script for you..... lol | 00:09 |
| johnsom | Ok, not only is that backward compatible with the old OSC, it gets us one step closer to devstack working with IPv6 lb-mgmt-net | 00:09 |
| johnsom | Here is the awk script version of it which is easier to read: | 00:11 |
| johnsom | https://www.irccloud.com/pastebin/PSV7QZwv/ | 00:11 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Fix the amphora no-op driver https://review.opendev.org/682420 | 00:13 |
| johnsom | I wonder if our backport vote counts towards stackalytics..... https://review.opendev.org/#/c/682171 | 00:17 |
| johnsom | It looks like the answer is no, which is good | 00:19 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: WIP: fix plugging member subnets on existing networks https://review.opendev.org/665402 | 00:27 |
| rm_work | lol well that's a thing | 00:28 |
| rm_work | nice awk | 00:28 |
| johnsom | Thanks | 00:28 |
| rm_work | i thought I handled that in one of my scripts a simpler way tho >_> | 00:28 |
| rm_work | but i guess that's fine | 00:28 |
| rm_work | err i mean, in one of my ipv6 related patches | 00:28 |
| johnsom | Well, OSC just changed the output format, so half that is being backwards compatible. | 00:29 |
| rm_work | https://review.opendev.org/#/c/665103/4 | 00:29 |
| johnsom | I can't wait to hear everyone scream over this change.... So many scripts are going to break | 00:29 |
| * rm_work shrugs | 00:30 | |
| rm_work | did you know i had that patch up for a while? :D | 00:30 |
| rm_work | it might have gotten lost | 00:30 |
| johnsom | I knew someone was looking at the IPv6 goal. I was potentially going to poke at it this week. | 00:30 |
| rm_work | which is funny because it's after you mentioned i should be using dustin's patch :D | 00:30 |
| rm_work | they changed how `-f value` displays? | 00:32 |
| johnsom | rm_work https://review.opendev.org/#/c/594078/ | 00:32 |
| johnsom | Yes | 00:33 |
| johnsom | https://www.irccloud.com/pastebin/R5QfYUzr/ | 00:33 |
| johnsom | The old output: | 00:34 |
| johnsom | https://www.irccloud.com/pastebin/UNEsKcET/ | 00:34 |
| rm_work | is that OSC or is that neutron? | 00:34 |
| johnsom | OSC | 00:34 |
| rm_work | because i thought that *is* the value from OSC, but because that's how neutron formats it inside the json value | 00:34 |
| rm_work | and it's not "double parsing" it | 00:34 |
| rm_work | did OSC decide to like, detect that and double-parse the neutron responses? lol | 00:35 |
| johnsom | This patch changed it: https://review.opendev.org/#/c/657907/ | 00:35 |
| rm_work | oh nm | 00:35 |
| johnsom | Greg found this issue... | 00:35 |
| johnsom | Like I said, pretty big change to the OSC output | 00:35 |
| rm_work | i see, it's changed from print('\n'.join(fixed_ips)) to print(fixed_ips) | 00:36 |
| rm_work | oh, no, even more complex | 00:36 |
| rm_work | they've parsed it? | 00:36 |
| rm_work | or ... are we sure that's not neutron's return structure changing? lol | 00:36 |
| rm_work | because it's a sub-field thing | 00:36 |
| rm_work | i don't know if neutron used to return it as a \n joined string, and now they're returning it as a real structure? or what | 00:37 |
| rm_work | i always thought it was odd before | 00:37 |
| rm_work | what does it look like if you use `-f json`? | 00:37 |
| johnsom | https://www.irccloud.com/pastebin/c9qSPURQ/ | 00:38 |
| johnsom | Not sure how the version of OSC I have installed in the venv would change the neutron API format..... | 00:39 |
| rm_work | huh. | 00:39 |
| rm_work | yeah | 00:39 |
| johnsom | Pretty sure this is OSC change.... | 00:39 |
| rm_work | but notice | 00:39 |
| rm_work | in the old one, "fixed_ips" is the field, and that whole thing is the value | 00:39 |
| rm_work | \n separated (from neutron) | 00:39 |
| rm_work | so in the new version, OSC must actually be parsing the field and json-ifying it magically | 00:40 |
| rm_work | which is ... GREAT. but, breaking >_< | 00:40 |
| johnsom | Yep, well, with the above awk magic, our devstack plugin will parse both.... | 00:42 |
| rm_work | k. well, there's way more stuff that needs to be fixed too -- see my patch | 00:42 |
| rm_work | are we going to do them one-off or try to batch some of them? | 00:42 |
| johnsom | Umm, which patch? | 00:43 |
| rm_work | it might be good to just merge them as we go I guess, since otherwise we just get what has continued to happen, which is patches doing the same thing piling up but none of them merging because they don't handle "EVERYTHING" | 00:43 |
| rm_work | the one I linked | 00:43 |
| rm_work | https://review.opendev.org/#/c/665103/ | 00:43 |
| johnsom | Ah, ipv6 | 00:43 |
| rm_work | i guess this isn't an ipv6-related one technically | 00:44 |
| johnsom | I will try to poke at that and consolidate on yours unless you are motivated | 00:44 |
| rm_work | IDC, abandon mine and dustins and add the important bits to yours | 00:44 |
| rm_work | as long as they're all captured | 00:44 |
| rm_work | want to get those things merged ASAP | 00:44 |
| rm_work | closer and closer to IPv6 | 00:44 |
| johnsom | Well, mine is an OSC gate fix that just happens to handle IPv6 better. It's not intended to be the "ipv6" patch | 00:44 |
| rm_work | kk then I can +2 that when I see it pass | 00:45 |
| rm_work | and you can do more on the one i put up, which was really just a newer version of dustin's | 00:45 |
| johnsom | Yeah, I was thinking I would work on the IPv6 goal a bit this week. | 00:45 |
| rm_work | or you can backport my changes to dustin's :D | 00:45 |
| rm_work | i think i just decided to start over because his was going to be confusing to rebase... | 00:45 |
| rm_work | or i just forgot | 00:45 |
| rm_work | but i am also pretty sure i pulled some stuff from his so *shrug* | 00:45 |
| rm_work | i personally give zero shits about LoC credit :D | 00:46 |
| johnsom | lol, well, it is easy enough to be polite too.... | 00:46 |
| rm_work | ah yeah, i mean personally | 00:46 |
| rm_work | I generally try to be polite to *others* | 00:47 |
| johnsom | Did you see this: https://review.opendev.org/#/c/682171 | 00:47 |
| rm_work | but do whatever you want with my stuff here | 00:47 |
| rm_work | did not see it yet | 00:47 |
| rm_work | need to go through and look for new reviews sometime this week, maybe tomorrow | 00:47 |
| johnsom | +1 to backport the master release notes.... | 00:47 |
| rm_work | lolol | 00:47 |
| rm_work | yeah ... maybe it shouldn't have been at the top | 00:48 |
| rm_work | i wonder if we can control the order | 00:48 |
| rm_work | i've seen people do it by accident and catch it and remove it | 00:48 |
| rm_work | but some prolly don't notice | 00:48 |
| johnsom | It is alphabetical | 00:48 |
| johnsom | I think | 00:48 |
| rm_work | ok so... | 00:48 |
| rm_work | zBackport-Candidate | 00:48 |
| rm_work | :D | 00:48 |
| rm_work | i am pretty sure that person meant to +1 the patch not the backport | 00:48 |
| rm_work | it's caught *me* before | 00:49 |
| johnsom | Really I think that was just a "backport the world" action as I saw a few of those recently | 00:49 |
| rm_work | i give benefit of the doubt here | 00:49 |
| rm_work | because i really have seen it catch people up, including myself | 00:49 |
| johnsom | https://review.opendev.org/#/c/667249/ | 00:49 |
| rm_work | OR it could be +1 scripts that aren't smart so they just pick the first field :D | 00:49 |
| johnsom | That could be | 00:50 |
| rm_work | could make a field "aIMABot" | 00:51 |
| rm_work | and see who +1s it | 00:51 |
| johnsom | That would be fun | 00:55 |
| *** hongbin has joined #openstack-lbaas | 01:03 | |
| *** baffle has quit IRC | 02:18 | |
| lxkong | johnsom: may i know how do you trigger the job `publish-openstack-octavia-amphora-image-bionic`? | 02:20 |
| lxkong | rm_work ^^ | 02:20 |
| johnsom | It is periodic, it runs once a day | 02:21 |
| lxkong | johnsom: where is that defined? | 02:21 |
| johnsom | Octavia/zuul/jobs.yaml | 02:22 |
| lxkong | johnsom: yeah, i can only see the job definition itself, i mean where `runs once a day` is defined? | 02:23 |
| lxkong | is it the behavior of `publish-openstack-artifacts` job? | 02:23 |
| johnsom | Oh, projects.yaml, it is at the bottom | 02:23 |
| lxkong | ah, i see | 02:23 |
| lxkong | johnsom: ta | 02:25 |
| johnsom | Np | 02:25 |
| *** baffle has joined #openstack-lbaas | 02:29 | |
| openstackgerrit | Merged openstack/python-octaviaclient master: Update master for stable/train https://review.opendev.org/682171 | 03:04 |
| rm_work | cgoncalves: you around to +A https://review.opendev.org/#/c/682327/ ? | 03:12 |
| rm_work | would like gates working :D | 03:12 |
| rm_work | or johnsom if he wants ;) | 03:12 |
| johnsom | Since the PTL told me too | 03:16 |
| johnsom | rm_work On https://review.opendev.org/#/c/682420/ Are you sure those aren't in the "client" part of the REST driver? or are you arguing we change all of them? | 03:18 |
| *** psachin has joined #openstack-lbaas | 03:22 | |
| *** ricolin has joined #openstack-lbaas | 03:22 | |
| *** ramishra has joined #openstack-lbaas | 03:33 | |
| rm_work | well you're changing: | 03:36 |
| rm_work | octavia/amphorae/drivers/driver_base.py | 03:36 |
| rm_work | octavia/amphorae/drivers/noop_driver/driver.py | 03:36 |
| rm_work | and this is | 03:36 |
| rm_work | octavia/amphorae/drivers/haproxy/rest_api_driver.py | 03:36 |
| rm_work | but | 03:37 |
| rm_work | within that file... | 03:37 |
| rm_work | hmm, maybe you're right, maybe that's why it got conflated to begin with? | 03:37 |
| rm_work | but it might not be a bad idea to update | 03:38 |
| rm_work | just to make things less confusing | 03:38 |
| rm_work | update_agent_config is called from update_amphora_agent_config lol | 03:38 |
| rm_work | yeah I might vote to update them all, but technically yours is the only one that needs to be updated to fix the bug, so I'll +2 | 03:39 |
| *** hongbin has quit IRC | 04:01 | |
| *** hongbin has joined #openstack-lbaas | 04:08 | |
| *** hongbin has quit IRC | 04:10 | |
| openstackgerrit | Merged openstack/octavia master: Fix openstack port show content handling https://review.opendev.org/682327 | 05:15 |
| *** pcaruana has joined #openstack-lbaas | 05:16 | |
| *** goldyfruit_ has quit IRC | 05:19 | |
| *** gcheresh has joined #openstack-lbaas | 05:23 | |
| *** pcaruana has quit IRC | 05:29 | |
| *** gcheresh has quit IRC | 05:45 | |
| *** ccamposr has joined #openstack-lbaas | 06:06 | |
| *** sapd1_x has joined #openstack-lbaas | 06:19 | |
| *** mjozefcz has joined #openstack-lbaas | 06:21 | |
| *** eandersson has quit IRC | 06:28 | |
| *** gcheresh has joined #openstack-lbaas | 06:34 | |
| *** eandersson has joined #openstack-lbaas | 06:38 | |
| *** gcheresh has quit IRC | 06:44 | |
| cgoncalves | such a beautiful awk line | 07:03 |
| openstackgerrit | Maciej Józefczyk proposed openstack/octavia master: Validate supported LB algorithm in Amphora provider drivers https://review.opendev.org/672477 | 07:05 |
| *** gtema has joined #openstack-lbaas | 07:19 | |
| gtema | Hey Octavia guys. Since few days devstack jobs in OpenStackSDK with octavia enabled are failing with following devstack setup message: argument --bind_ip: Invalid IPAddress value: | 07:20 |
| gtema | i.e. https://zuul.opendev.org/t/openstack/build/00faec04eed849ae85995325cc135fc3/log/job-output.txt#36243 | 07:20 |
| gtema | any ideas why could this happen? I do not see any recent suspicious changes | 07:21 |
| gthiemonge | gtema: a fix was merged 2h ago: https://review.opendev.org/#/c/682327/ | 07:21 |
| *** ivve has joined #openstack-lbaas | 07:21 | |
| gtema | oj, that's great. Thanks a lot | 07:21 |
| *** trident has quit IRC | 07:22 | |
| *** rpittau|afk is now known as rpittau | 07:28 | |
| *** trident has joined #openstack-lbaas | 07:31 | |
| *** trident has quit IRC | 07:36 | |
| *** trident has joined #openstack-lbaas | 07:46 | |
| *** gcheresh has joined #openstack-lbaas | 07:47 | |
| *** rcernin has quit IRC | 07:47 | |
| *** gcheresh has quit IRC | 08:03 | |
| *** tkajinam has quit IRC | 08:03 | |
| *** ataraday has joined #openstack-lbaas | 08:20 | |
| *** luksky has joined #openstack-lbaas | 08:44 | |
| *** frickler has joined #openstack-lbaas | 08:50 | |
| *** sapd1_x has quit IRC | 08:53 | |
| *** pcaruana has joined #openstack-lbaas | 09:19 | |
| openstackgerrit | Merged openstack/octavia stable/stein: Fixed pool and members status with UDP loadbalancers https://review.opendev.org/675362 | 09:57 |
| openstackgerrit | Gregory Thiemonge proposed openstack/octavia master: Fix healthmonitor message v2 for UDP listeners https://review.opendev.org/682365 | 09:57 |
| openstackgerrit | Merged openstack/octavia stable/stein: Fixed down server issue after reloading keepalived https://review.opendev.org/675363 | 09:58 |
| *** ramishra_ has joined #openstack-lbaas | 10:07 | |
| *** pcaruana has quit IRC | 10:07 | |
| *** ramishra has quit IRC | 10:08 | |
| openstackgerrit | Gregory Thiemonge proposed openstack/octavia-tempest-plugin master: Added tests for creation of listeners on same port https://review.opendev.org/668176 | 10:08 |
| openstackgerrit | Gregory Thiemonge proposed openstack/octavia-tempest-plugin master: Add tests for mixed IP networks UDP members https://review.opendev.org/668619 | 10:08 |
| openstackgerrit | Gregory Thiemonge proposed openstack/octavia-tempest-plugin master: Add UDP test scenario https://review.opendev.org/656515 | 10:08 |
| *** gtema has left #openstack-lbaas | 10:10 | |
| *** sapd1_x has joined #openstack-lbaas | 10:25 | |
| *** luksky has quit IRC | 10:29 | |
| lxkong | hi octavia-core, anyone could please review https://review.opendev.org/682240? | 10:31 |
| *** pcaruana has joined #openstack-lbaas | 10:55 | |
| *** luksky has joined #openstack-lbaas | 11:05 | |
| *** goldyfruit_ has joined #openstack-lbaas | 11:07 | |
| *** rouk has quit IRC | 11:14 | |
| *** sapd1_x has quit IRC | 11:15 | |
| *** boden has joined #openstack-lbaas | 11:45 | |
| *** Conqueror has joined #openstack-lbaas | 11:49 | |
| *** takamatsu has joined #openstack-lbaas | 11:53 | |
| *** goldyfruit_ has quit IRC | 12:02 | |
| *** pcaruana has quit IRC | 12:27 | |
| *** openstackstatus has quit IRC | 12:28 | |
| *** openstack has joined #openstack-lbaas | 12:32 | |
| *** ChanServ sets mode: +o openstack | 12:32 | |
| *** pcaruana has joined #openstack-lbaas | 12:59 | |
| *** goldyfruit_ has joined #openstack-lbaas | 13:09 | |
| *** yamamoto has joined #openstack-lbaas | 13:20 | |
| *** pcaruana has quit IRC | 13:28 | |
| *** goldyfruit_ has quit IRC | 13:39 | |
| *** psachin has quit IRC | 13:43 | |
| *** gthiemonge has quit IRC | 13:55 | |
| *** gthiemonge has joined #openstack-lbaas | 13:55 | |
| *** goldyfruit_ has joined #openstack-lbaas | 13:59 | |
| *** pcaruana has joined #openstack-lbaas | 13:59 | |
| *** rpittau is now known as rpittau|afk | 14:10 | |
| *** goldyfruit___ has joined #openstack-lbaas | 14:17 | |
| *** goldyfruit_ has quit IRC | 14:20 | |
| *** Vorrtex has joined #openstack-lbaas | 14:27 | |
| *** pcaruana has quit IRC | 14:55 | |
| cgoncalves | help needed reviewing backport patches: https://review.opendev.org/#/q/status:open+project:openstack/octavia+NOT+branch:master+AND+Workflow%253E%253D0 | 14:57 |
| openstackgerrit | Carlos Goncalves proposed openstack/octavia stable/rocky: Fix l7rule API handling of None updates https://review.opendev.org/678551 | 15:01 |
| openstackgerrit | Carlos Goncalves proposed openstack/octavia stable/queens: Fix l7rule API handling of None updates https://review.opendev.org/678557 | 15:03 |
| *** Vorrtex has quit IRC | 15:09 | |
| *** luksky has quit IRC | 15:11 | |
| *** sapd1 has quit IRC | 15:14 | |
| *** sapd1 has joined #openstack-lbaas | 15:15 | |
| *** ivve has quit IRC | 15:27 | |
| *** nmagnezi has joined #openstack-lbaas | 15:35 | |
| *** goldyfruit_ has joined #openstack-lbaas | 15:50 | |
| *** goldyfruit___ has quit IRC | 15:53 | |
| *** boden has quit IRC | 16:09 | |
| *** AustinR has quit IRC | 16:11 | |
| *** AustinR has joined #openstack-lbaas | 16:12 | |
| *** yamamoto has quit IRC | 16:14 | |
| *** boden has joined #openstack-lbaas | 16:23 | |
| *** goldyfruit___ has joined #openstack-lbaas | 16:40 | |
| *** goldyfruit_ has quit IRC | 16:43 | |
| *** luksky has joined #openstack-lbaas | 16:50 | |
| *** pcaruana has joined #openstack-lbaas | 16:55 | |
| *** ramishra_ has quit IRC | 17:08 | |
| *** yamamoto has joined #openstack-lbaas | 17:40 | |
| *** yamamoto has quit IRC | 17:48 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Allow IPv6 health network in devstack https://review.opendev.org/665103 | 18:00 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Allow IPv6 health network in devstack https://review.opendev.org/665103 | 18:07 |
| openstackgerrit | Merged openstack/octavia stable/stein: Prevent UDP LBs to use different IP protocol versions in amphora driver https://review.opendev.org/675364 | 18:07 |
| openstackgerrit | Merged openstack/octavia stable/rocky: Add support for monitor_{address,port} in UDP members https://review.opendev.org/675351 | 18:07 |
| openstackgerrit | Merged openstack/octavia stable/rocky: Fixed pool and members status with UDP loadbalancers https://review.opendev.org/675575 | 18:07 |
| openstackgerrit | Merged openstack/octavia stable/rocky: Fixed down server issue after reloading keepalived https://review.opendev.org/675576 | 18:07 |
| openstackgerrit | Merged openstack/octavia stable/rocky: Prevent UDP LBs to use different IP protocol versions in amphora driver https://review.opendev.org/675577 | 18:07 |
| johnsom | Looks like we got through some of those backport reviews.... | 18:08 |
| *** ccamposr has quit IRC | 18:24 | |
| *** ccamposr has joined #openstack-lbaas | 18:27 | |
| *** ricolin has quit IRC | 18:28 | |
| *** Vorrtex has joined #openstack-lbaas | 18:30 | |
| *** goldyfruit___ has quit IRC | 18:31 | |
| *** henriqueof has joined #openstack-lbaas | 18:38 | |
| cgoncalves | still 12 to go | 18:41 |
| *** mjozefcz has quit IRC | 19:05 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/octavia-tempest-plugin master: [train][goal] Define new 'octavia-v2-dsvm-noop-api-ipv6-only' job https://review.opendev.org/682726 | 19:37 |
| *** gmann has joined #openstack-lbaas | 19:39 | |
| *** dtruong has quit IRC | 19:40 | |
| *** goldyfruit___ has joined #openstack-lbaas | 19:40 | |
| *** dtruong has joined #openstack-lbaas | 19:40 | |
| openstackgerrit | Ghanshyam Mann proposed openstack/octavia-tempest-plugin master: [train][goal] Define new 'octavia-v2-dsvm-noop-api-ipv6-only' job https://review.opendev.org/682726 | 19:41 |
| johnsom | Well, that was a collision in work... | 19:42 |
| gmann | johnsom: if you have patch ready and it take time to rebase or merge with mine then I can abandon mine. whatever easy for you | 19:45 |
| johnsom | gmann I can run with this, no problem. The base_vars is new to me though. It looks like zuul doesn't like them either. | 19:46 |
| gmann | ok. i thought that can work but we can copy paste the *base_vars_2 and*base_vars_3 | 19:48 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Allow IPv6 health network in devstack https://review.opendev.org/665103 | 19:49 |
| colin- | was i mistaken about the health monitor api supporting a string search for HTTP type checks? | 19:51 |
| johnsom | colin- Correct, not implemented yet | 19:51 |
| colin- | understood | 19:51 |
| *** goldyfruit_ has joined #openstack-lbaas | 20:15 | |
| *** goldyfruit___ has quit IRC | 20:18 | |
| *** boden has quit IRC | 20:20 | |
| *** mjozefcz has joined #openstack-lbaas | 20:43 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia-tempest-plugin master: [train][goal] Define new 'octavia-v2-dsvm-noop-api-ipv6-only' job https://review.opendev.org/682726 | 20:44 |
| zigo | Hi. | 20:47 |
| zigo | Will there be a train release of octavia-lib ? | 20:47 |
| zigo | I'm currently going through all the Train libs pacakging for Debian... | 20:48 |
| openstackgerrit | Michael Johnson proposed openstack/octavia-tempest-plugin master: [train][goal] Define new 'octavia-v2-dsvm-noop-api-ipv6-only' job https://review.opendev.org/682726 | 20:48 |
| *** trident has quit IRC | 20:48 | |
| johnsom | zigo It is already out, 1.4.0: https://releases.openstack.org/train/index.html#octavia-lib | 20:48 |
| zigo | johnsom: It's not mirrored in github, then? | 20:48 |
| johnsom | Should be | 20:49 |
| johnsom | zigo https://github.com/openstack/octavia-lib/tree/1.4.0 | 20:49 |
| johnsom | It is there | 20:49 |
| johnsom | pypi too: https://pypi.org/project/octavia-lib/#history | 20:50 |
| zigo | Oh, I did it already in fact ! :) | 20:51 |
| zigo | Just didn't upload ... | 20:51 |
| zigo | I got to tell, we learned how to love Octavia in our company. | 20:54 |
| zigo | Thanks for it ! | 20:54 |
| zigo | It's really convenient ... | 20:54 |
| openstackgerrit | Merged openstack/octavia stable/stein: Add support for oslo_middleware http_proxy_to_wsgi https://review.opendev.org/674467 | 20:55 |
| *** mjozefcz has quit IRC | 20:55 | |
| *** luksky has quit IRC | 20:55 | |
| zigo | Just one thing, I don't get why our load balancers never leave operating_status OFFLINE ... | 20:55 |
| zigo | Though it does work ... | 20:55 |
| zigo | We're still on Rocky in production though. | 20:56 |
| zigo | Is this a known bug? | 20:56 |
| zigo | johnsom: ^ | 20:57 |
| *** pcaruana has quit IRC | 20:58 | |
| *** trident has joined #openstack-lbaas | 21:00 | |
| johnsom | zigo No, that likely means your lb-mgmt-net is not working. The health heartbeat messages are not getting back to the controllers for some reason. | 21:02 |
| zigo | johnsom: The network is ok, the load balancer system does work ... | 21:03 |
| johnsom | Check this config setting: https://docs.openstack.org/octavia/latest/configuration/configref.html#health_manager.controller_ip_port_list | 21:03 |
| johnsom | Then check that the health manager is getting the heartbeat packets on UDP 5555 | 21:03 |
| zigo | Oh... | 21:04 |
| zigo | johnsom: Typically, that'd be the IP of the API ? | 21:04 |
| johnsom | zigo You can still load balance if the lb-mgmt-net is not getting it's heartbeats. We just won't update stats, status, or be able to do failovers | 21:04 |
| zigo | Ok, good to know. | 21:05 |
| johnsom | No, it's a list of the health manager endpoints on the lb-mgmt-net | 21:05 |
| zigo | Hum... | 21:05 |
| zigo | So, the health manager needs to have a network interface on the lb-mgnt-net then? | 21:05 |
| johnsom | Yes | 21:05 |
| johnsom | Technically the worker, health manager, and housekeeping should have an interface on that network | 21:06 |
| johnsom | API does not need it | 21:06 |
| zigo | That's not how I designed it ... :( | 21:06 |
| zigo | Why can't it just be the API IP ? | 21:06 |
| johnsom | Well, for one thing, the amphroa don't have access to keystone for a token | 21:06 |
| zigo | Our workers don't have an IP on the lb-mgmt-net, and it still works ... | 21:07 |
| zigo | On our setup, the amphora could ... | 21:07 |
| zigo | It's routed. | 21:07 |
| zigo | Probably we shouldn't do that. | 21:07 |
| zigo | But I got the point. | 21:08 |
| zigo | Thanks. | 21:08 |
| openstackgerrit | Michael Johnson proposed openstack/octavia-tempest-plugin master: [train][goal] Define new 'octavia-v2-dsvm-noop-api-ipv6-only' job https://review.opendev.org/682726 | 21:14 |
| johnsom | Sorry for the patch noise, need to figure out this ansible/yaml syntax. As always, ansible is "lightly documented" | 21:14 |
| openstackgerrit | Michael Johnson proposed openstack/octavia-tempest-plugin master: [train][goal] Define new 'octavia-v2-dsvm-noop-api-ipv6-only' job https://review.opendev.org/682726 | 21:16 |
| openstackgerrit | Merged openstack/octavia stable/stein: Fix template that generates vrrp check script https://review.opendev.org/679753 | 21:16 |
| *** Vorrtex has quit IRC | 21:21 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia-tempest-plugin master: [train][goal] Define new 'octavia-v2-dsvm-noop-api-ipv6-only' job https://review.opendev.org/682726 | 21:23 |
| openstackgerrit | Michael Johnson proposed openstack/octavia-tempest-plugin master: [train][goal] Define new 'octavia-v2-dsvm-noop-api-ipv6-only' job https://review.opendev.org/682726 | 21:24 |
| openstackgerrit | Michael Johnson proposed openstack/octavia-tempest-plugin master: [train][goal] Define new 'octavia-v2-dsvm-noop-api-ipv6-only' job https://review.opendev.org/682726 | 21:29 |
| *** ccamposr has quit IRC | 21:34 | |
| *** ccamposr has joined #openstack-lbaas | 21:38 | |
| lxkong | johnsom, rm_work may i have your review to the patch https://review.opendev.org/#/c/682240/? We need that in our scenario test | 22:04 |
| openstackgerrit | Michael Johnson proposed openstack/octavia-tempest-plugin master: [train][goal] Define new 'octavia-v2-dsvm-noop-api-ipv6-only' job https://review.opendev.org/682726 | 22:06 |
| johnsom | lxkong I saw you mention last night. I need to understand the issue | 22:11 |
| lxkong | johnsom: it's in our prod, when running that test case, i found the VIP port `device_owner ` was changed to `compute:None` | 22:13 |
| johnsom | Yeah, I guess that is a nova / neutron race condition. | 22:14 |
| lxkong | and we are using old version for both Nova and Neutron | 22:15 |
| lxkong | :-( | 22:15 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Allow IPv6 health network in devstack https://review.opendev.org/665103 | 22:19 |
| *** ccamposr__ has joined #openstack-lbaas | 22:23 | |
| *** ccamposr has quit IRC | 22:26 | |
| rm_work | Yeah I'll look lxkong | 22:58 |
| *** tkajinam has joined #openstack-lbaas | 23:04 | |
| lxkong | rm_work: ty | 23:14 |
| *** goldyfruit_ has quit IRC | 23:17 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Allow IPv6 health network in devstack https://review.opendev.org/665103 | 23:29 |
| *** rcernin has joined #openstack-lbaas | 23:44 | |
| *** yamamoto has joined #openstack-lbaas | 23:46 | |
| *** yamamoto has quit IRC | 23:51 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!