| *** sapd1 has quit IRC | 00:16 | |
| rm_work | Baggypants12000: also here: https://developer.openstack.org/api-ref/load-balancer/v2/index.html?expanded=create-listener-detail#create-listener | 00:24 |
|---|---|---|
| rm_work | the API for create listener lists the fields necessary | 00:25 |
| rm_work | `default_tls_container_ref `: The URI of the key manager service secret containing a PKCS12 format certificate/key bundle for TERMINATED_TLS listeners. | 00:25 |
| rm_work | ah sorry, that's roukoswarf | 00:25 |
| rm_work | ^^ | 00:25 |
| *** abaindur has quit IRC | 00:26 | |
| rm_work | Baggypants12000: wires crossed, was also thinking of a response for you -- I think you could use something like certbot or dogtag to handle the internal certs as well, plus cron... but, that CA needs to live longer than any existing amphora cert (for validation purposes) and we don't support multiple CAs? or actually maybe we do, if that's standard cert stuff to be able to put multiple CAs in one PEM | 00:27 |
| *** fnaval has joined #openstack-lbaas | 00:38 | |
| *** fnaval has quit IRC | 00:46 | |
| *** fnaval has joined #openstack-lbaas | 00:49 | |
| *** fnaval has quit IRC | 00:49 | |
| *** sapd1 has joined #openstack-lbaas | 00:50 | |
| openstackgerrit | Merged openstack/octavia master: Fix health monitor exception https://review.openstack.org/641008 | 01:29 |
| *** abaindur has joined #openstack-lbaas | 01:48 | |
| *** abaindur has quit IRC | 02:17 | |
| *** hongbin has joined #openstack-lbaas | 02:42 | |
| *** ramishra has joined #openstack-lbaas | 02:55 | |
| *** abaindur has joined #openstack-lbaas | 04:25 | |
| *** abaindur has quit IRC | 04:26 | |
| *** abaindur has joined #openstack-lbaas | 04:33 | |
| *** Dinesh_Bhor has quit IRC | 05:29 | |
| *** Dinesh_Bhor has joined #openstack-lbaas | 05:33 | |
| *** hongbin has quit IRC | 05:42 | |
| *** sapd1 has quit IRC | 05:42 | |
| *** sapd1 has joined #openstack-lbaas | 05:52 | |
| *** abaindur has quit IRC | 06:23 | |
| *** sapd1 has quit IRC | 06:44 | |
| *** ivve has joined #openstack-lbaas | 06:49 | |
| *** ccamposr has joined #openstack-lbaas | 06:49 | |
| *** ramishra has quit IRC | 07:10 | |
| *** luksky has joined #openstack-lbaas | 07:17 | |
| *** ramishra has joined #openstack-lbaas | 07:25 | |
| *** pcaruana has joined #openstack-lbaas | 07:26 | |
| *** gthiemonge has joined #openstack-lbaas | 07:37 | |
| *** rpittau|afk is now known as rpittau | 08:06 | |
| *** pcaruana has quit IRC | 09:12 | |
| *** pcaruana has joined #openstack-lbaas | 09:27 | |
| *** rcernin has quit IRC | 09:47 | |
| *** ramishra has quit IRC | 10:05 | |
| *** mkuf has quit IRC | 10:09 | |
| *** mkuf has joined #openstack-lbaas | 10:13 | |
| Baggypants12000 | rm_work: dogtag seems like the best shout here, thanks for the pointer! | 10:18 |
| *** salmankhan has joined #openstack-lbaas | 10:27 | |
| *** yamamoto has quit IRC | 10:43 | |
| *** ramishra has joined #openstack-lbaas | 10:48 | |
| *** luksky has quit IRC | 11:06 | |
| rm_work | Baggypants12000: yeah np, i think probably one of the main reasons we haven't really gotten into it is because it's so opinionated and everyone has their own way of dealing with certs | 11:13 |
| *** yamamoto has joined #openstack-lbaas | 11:18 | |
| *** yamamoto has quit IRC | 11:25 | |
| *** yamamoto has joined #openstack-lbaas | 11:26 | |
| *** luksky has joined #openstack-lbaas | 11:36 | |
| *** gcheresh_ has joined #openstack-lbaas | 11:37 | |
| *** gcheresh_ has quit IRC | 11:58 | |
| *** willksm has joined #openstack-lbaas | 12:51 | |
| *** henriqueof has quit IRC | 13:06 | |
| *** ivve has quit IRC | 13:08 | |
| *** ramishra has quit IRC | 13:27 | |
| *** yamamoto has quit IRC | 13:27 | |
| openstackgerrit | Merged openstack/octavia master: Updates Octavia to support octavia-lib https://review.openstack.org/613709 | 13:29 |
| *** pcaruana has quit IRC | 13:30 | |
| cgoncalves | ^ we're ready to cut a stein branch :) | 13:53 |
| *** yamamoto has joined #openstack-lbaas | 13:54 | |
| *** pcaruana has joined #openstack-lbaas | 14:37 | |
| *** gthiemonge has quit IRC | 15:15 | |
| *** gthiemonge has joined #openstack-lbaas | 15:18 | |
| *** kmadac2 has joined #openstack-lbaas | 15:18 | |
| kmadac2 | Hi folks, I have octavia installed with openstack-ansible. I'm able to create LBs with openstack cli client, but dashboard in horizon does not work. I can see Load Balancers menu, but whe I click on it, horizon starts to reload the page again and again indefinitely. No errors is found in horizon log files. Do you have any hint where to look to find the root cause? | 15:22 |
| kmadac2 | OpenStack version is Rocky | 15:23 |
| johnsom | kmadac2: Most likely issue is that both neutron-lbaas-dashboard and octavia-dashboard are installed in horizon. Removing neutron-lbaas-dashboard should solve your problem. | 15:25 |
| johnsom | We have a known bug when they are both installed. | 15:25 |
| kmadac2 | @johnsom: thanks, I had both of them installed, but then I disable in neutron-lbaas-dashboard in deployment configuration, but problem still persist. What is the best way to check if it is really removed? | 15:29 |
| johnsom | kmadac2: In the horizon container and directory there is a folder local/enabled or something similar. Make sure __1482... is the only lbaas file there. The other might be 1481. | 15:31 |
| johnsom | Also you can re-run the manage.py collectstatic and compress commands. The. Restart apache. | 15:32 |
| johnsom | If that doesn’t work you can clear the horizon static folder or rebuilt the horizon containers using ansible. | 15:33 |
| kmadac2 | johnsom: I will probably rebuild the containers from scratch, thank you :) | 15:34 |
| johnsom | Np | 15:35 |
| *** yamamoto has quit IRC | 15:48 | |
| *** yamamoto has joined #openstack-lbaas | 15:49 | |
| *** yamamoto has quit IRC | 15:51 | |
| *** luksky has quit IRC | 15:51 | |
| *** yamamoto has joined #openstack-lbaas | 16:01 | |
| *** yamamoto has quit IRC | 16:07 | |
| *** gthiemonge has quit IRC | 16:12 | |
| *** gthiemonge has joined #openstack-lbaas | 16:12 | |
| cgoncalves | folks, any last-minute feature patch you'd like to see merged in for Stein? | 16:20 |
| xgerman | love to see logging but we never agreed on how to do that | 16:25 |
| cgoncalves | yeah, pity. I'd have liked to have in that too. last time I checked logs were not being saved in gate (https://review.openstack.org/#/c/626406/) | 16:27 |
| xgerman | well, and then johnsom didn’t like the design (I think he wanted two hosts to send logs to) | 16:32 |
| xgerman | but if that’s all we want we can try FFE | 16:32 |
| *** pcaruana has quit IRC | 17:12 | |
| *** fnaval has joined #openstack-lbaas | 17:19 | |
| *** rpittau is now known as rpittau|afk | 17:31 | |
| *** gthiemonge has quit IRC | 18:05 | |
| johnsom | cgoncalves Are you still looking at that verification patch? | 18:08 |
| johnsom | octavia.tests.functional.api.v2.test_l7policy.TestL7Policy.test_listener_pool_protocol_map_put 1068.653 | 18:08 |
| johnsom | This is part of the problem..... | 18:09 |
| cgoncalves | johnsom, no | 18:09 |
| johnsom | There are a few others that are in triple digit seconds per test as well. | 18:09 |
| johnsom | Ok, on my list | 18:09 |
| johnsom | I don't think we need a functional test that takes 17 minutes. | 18:10 |
| johnsom | Excellent catch BTW. Kudos | 18:10 |
| cgoncalves | my laptop fan was the whistle-blower | 18:10 |
| johnsom | lol, yeah I have a core graph on a small screen that is also my "something is horribly wrong" warning. | 18:11 |
| *** salmankhan has quit IRC | 18:11 | |
| johnsom | https://usercontent.irccloud-cdn.com/file/7yLPgIHB/1552068755.JPG | 18:12 |
| cgoncalves | windows?? | 18:13 |
| johnsom | Yeah | 18:14 |
| johnsom | Previously necessary to make some corporate tools run smoothly.... | 18:14 |
| johnsom | Xming works nicely though. | 18:15 |
| cgoncalves | "previously", heh | 18:15 |
| *** ccamposr has quit IRC | 18:32 | |
| *** henriqueof has joined #openstack-lbaas | 19:02 | |
| *** irclogbot_1 has quit IRC | 20:13 | |
| *** abaindur has joined #openstack-lbaas | 20:15 | |
| cgoncalves | octavia.tests.functional.api.v2.test_l7policy.TestL7Policy.test_listener_pool_protocol_map_put 368.900s | 20:16 |
| cgoncalves | getting better | 20:16 |
| *** abaindur has quit IRC | 20:31 | |
| *** abaindur has joined #openstack-lbaas | 20:32 | |
| *** abaindur has quit IRC | 20:32 | |
| *** abaindur has joined #openstack-lbaas | 20:32 | |
| *** yamamoto has joined #openstack-lbaas | 20:49 | |
| *** yamamoto has quit IRC | 20:54 | |
| openstackgerrit | OpenStack Release Bot proposed openstack/octavia-lib stable/stein: Update .gitreview for stable/stein https://review.openstack.org/642154 | 21:19 |
| openstackgerrit | OpenStack Release Bot proposed openstack/octavia-lib stable/stein: Update UPPER_CONSTRAINTS_FILE for stable/stein https://review.openstack.org/642155 | 21:19 |
| openstackgerrit | OpenStack Release Bot proposed openstack/octavia-lib master: Update master for stable/stein https://review.openstack.org/642156 | 21:19 |
| johnsom | Continuing with my PTL paperwork duties.... octavia-lib now has a stable/stein branch. | 21:21 |
| johnsom | Also, these will be our "marketing highlights" for Stein: | 21:21 |
| johnsom | https://www.irccloud.com/pastebin/CW0rjf9B/ | 21:21 |
| johnsom | Let me know if you have feedback on those.... | 21:21 |
| johnsom | It's a bulleted list when rendered in the release docs. | 21:23 |
| roukoswarf | whats the metadata tags? | 21:24 |
| johnsom | From the API guide: "A list of simple strings assigned to the resource." | 21:25 |
| johnsom | People use them to track resources when they use automation tools like terraform, etc. | 21:25 |
| roukoswarf | oh, so not a traffic tagging at all | 21:25 |
| johnsom | Nope, It's an API only feature that most of the OpenStack services support. | 21:26 |
| johnsom | This is another reference: https://specs.openstack.org/openstack/api-wg/guidelines/tags.html | 21:26 |
| johnsom | We also have this in the API guide: https://developer.openstack.org/api-ref/load-balancer/v2/index.html?expanded=create-a-load-balancer-detail#filtering-by-tags | 21:28 |
| roukoswarf | im just lurking waiting for something to replace f5 irules so i can make the argument to move away from f5/a10, which is keeping us on lbaas | 21:28 |
| johnsom | Ah, good to know. Do you have some sample use cases you could put in a story for us? | 21:28 |
| johnsom | Not that those aren't fine devices, but if it's a feature you are looking for out of the Octavia API it would be good to know. | 21:29 |
| roukoswarf | we are still early stages migrating to openstack, the devs arent 100% sure whats missing yet, but we have a lot of irule crazy lying around, so i will let you know once we start porting more. | 21:32 |
| roukoswarf | we have a10/f5 via lbaas for now, as a fallback | 21:32 |
| johnsom | Ok. With the new flavors, the F5 driver could support the operator injecting irules (it was discussed during the flavors design work). | 21:32 |
| roukoswarf | i would like to be free of the lbaas dep | 21:32 |
| roukoswarf | any idea when an f5/a10 driver for octavia will be here? instead of relying on legacy lbaas? | 21:33 |
| johnsom | I don't know on those two. It's up to the vendors as we have no way to test, etc. | 21:33 |
| roukoswarf | seems like they arent in much of a hurry to move, just to get a10 working how we needed i had to add a bunch of code to their driver. | 21:34 |
| roukoswarf | and that just getting their stuff to do what we need with legacy lbaas. | 21:34 |
| johnsom | Yeah, I think A10 had some development team changes..... | 21:34 |
| johnsom | We are here to help them develop their drivers though. | 21:35 |
| roukoswarf | their plumbing code must have been written by a college student with a point to prove and way too much time cause it was very strangely made. | 21:35 |
| johnsom | The Octavia driver specification was written by F5 | 21:35 |
| *** luksky has joined #openstack-lbaas | 21:35 | |
| johnsom | Unfortunately they also had some internal changes. I hope they can get back to it soon if they haven't already. | 21:36 |
| roukoswarf | whats stopping octavia devs from writing it? time, money, hardware, or all of the above? | 21:37 |
| roukoswarf | (the octavia drivers for f5/a10) | 21:38 |
| johnsom | Yes. But mostly, we don't have access to the hardware to test against. Also we would need the approval from the company, etc. | 21:38 |
| roukoswarf | approval from the company to work on their customer facing api? | 21:39 |
| johnsom | Well, I think it would be a good idea if we represent is as "the F5 driver" or such. | 21:40 |
| roukoswarf | sure, but if they arent competing currently, if you built one in-tree they would be better off just commiting to that once they get the bandwidth | 21:41 |
| johnsom | Most of us are sponsored (or were until recently) by employers that have their own needs/agenda. So developing for another company isn't always on that list. | 21:41 |
| roukoswarf | ah | 21:41 |
| johnsom | With Octavia we made the decision to have all but the reference driver out of the tree, simply because it's overhead for the driver developers to get us to review, etc. With the new plugin model for the drivers, this makes it pretty easy to build an out-of-tree driver. | 21:42 |
| roukoswarf | i wouldnt mind porting the a10 driver if it wasnt written insanely | 21:43 |
| roukoswarf | but its written insanely. | 21:43 |
| johnsom | I wrote a decent guide for creating one if you get the itch: https://docs.openstack.org/octavia/latest/contributor/guides/providers.html | 21:43 |
| roukoswarf | pretty docs, ill need to read them | 21:45 |
| roukoswarf | for now my "sponsor" is busy getting bootstrapped in openstack, might have some bandwidth eventually. | 21:46 |
| johnsom | lol, sounds good | 21:47 |
| openstackgerrit | Merged openstack/octavia-lib stable/stein: Update .gitreview for stable/stein https://review.openstack.org/642154 | 21:47 |
| openstackgerrit | Merged openstack/octavia-lib master: Update master for stable/stein https://review.openstack.org/642156 | 21:47 |
| *** roukoswarf has quit IRC | 21:49 | |
| *** Swami has joined #openstack-lbaas | 22:28 | |
| rm_work | johnsom: this needs to be rebased? https://review.openstack.org/#/c/617015/3 | 22:53 |
| rm_work | do we need these soon? | 22:53 |
| rm_work | like, *today*, or soon | 22:53 |
| johnsom | No, we can boot that to train | 22:53 |
| rm_work | k | 22:53 |
| johnsom | It's non-critical cleanup | 22:54 |
| rm_work | kk | 22:54 |
| rm_work | anything of yours i need to look at today? | 22:54 |
| johnsom | Nope, we are in feature freeze now IMO | 22:54 |
| johnsom | We got the last hot item in this morning | 22:54 |
| johnsom | I'm just working on paperwork stuffs today (release note cleanup, marketing highlights, stable/stein for the octavia-lib, etc.) | 22:55 |
| rm_work | k | 22:56 |
| johnsom | Well, among other distractions. | 22:58 |
| cgoncalves | we can start purging API v1, event queue and tempest+octavia-v1-dsvm jobs! | 22:59 |
| johnsom | cgoncalves It's September or "U" right? | 23:01 |
| johnsom | We plan to retire neutron-lbaas and neutron-lbaas-dashboard in September 2019 or the "U" OpenStack release cycle, whichever comes first. This means the code will be be removed and will not be released as part of the "U" OpenStack release. | 23:01 |
| johnsom | Not that it wouldn't so refreshing to that... | 23:01 |
| *** irclogbot_1 has joined #openstack-lbaas | 23:01 | |
| cgoncalves | https://releases.openstack.org/train/schedule.html | 23:03 |
| johnsom | Hmmm | 23:04 |
| johnsom | Glad to see the Train schedule posted. | 23:04 |
| johnsom | We should probably vote on that next week and update the statement to reflect the "T" cycle given the change in release cadence. | 23:06 |
| cgoncalves | sure. not that it would change anything anyway | 23:11 |
| cgoncalves | September 2019 comes first than T (and U) | 23:11 |
| johnsom | +1 | 23:11 |
| johnsom | Geez, we might all fight over which parts of the code we want to rip out! grin | 23:12 |
| cgoncalves | lol! | 23:12 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Add a prelude to the Stein release notes https://review.openstack.org/642177 | 23:24 |
| openstackgerrit | Ken'ichi Ohmichi proposed openstack/octavia master: Remove unnecessary `openstack flavor show` https://review.openstack.org/642178 | 23:25 |
| *** fnaval has quit IRC | 23:26 | |
| cgoncalves | should we be calling "octavia flavors" or "load balancer flavors"? | 23:30 |
| johnsom | I have been saying octavia because they are different than the crappy/barely implemented flavors that neutron-lbaas had. | 23:30 |
| johnsom | But, feel free to vote/comment | 23:31 |
| cgoncalves | it's a minor thing, nothing special | 23:31 |
| *** fnaval has joined #openstack-lbaas | 23:34 | |
| cgoncalves | in functional, creating 199 listeners took 499 seconds :/ | 23:39 |
| cgoncalves | each creation started being subsecond and ended taking ~6 seconds | 23:40 |
| johnsom | Hmmm, interesting | 23:40 |
| rm_work | cgoncalves: my functional tests locally do not take nearly that long | 23:41 |
| rm_work | what are you running? | 23:41 |
| rm_work | just octavia `tox -e functional`? | 23:41 |
| johnsom | He is looking at the new validation patch that takes like 17 minutes to run functional on | 23:41 |
| cgoncalves | https://gist.github.com/cgoncalves/db1807bfd452de0a09262062c1888b02 | 23:41 |
| rm_work | i got on a "long test" kick a while back and there was only one thing that took any time | 23:41 |
| rm_work | hmmmmmm | 23:41 |
| rm_work | which patch is that | 23:42 |
| johnsom | octavia.tests.functional.api.v2.test_l7policy.TestL7Policy.test_listener_pool_protocol_map_put 1068.653 | 23:42 |
| rm_work | i mean which patch | 23:42 |
| rm_work | not which test | 23:42 |
| johnsom | https://review.openstack.org/594040 | 23:42 |
| cgoncalves | https://review.openstack.org/#/c/594040/ | 23:42 |
| johnsom | Yeah, I know, just pointing out the issue, though I think what Carlos found is a different, yet related issue. | 23:42 |
| rm_work | jinx | 23:42 |
| cgoncalves | there's a slight improvement we can make: https://gist.github.com/cgoncalves/b3e0f87132d69a6e8910c416af61410d | 23:44 |
| rm_work | wow, -2 johnsom ? :D | 23:44 |
| rm_work | brutal | 23:44 |
| cgoncalves | savage! | 23:45 |
| johnsom | rm_work You want a 25 minute functional test to merge? It was already +W.... | 23:45 |
| johnsom | Oh and.... Fire me! grin | 23:46 |
| rm_work | lol | 23:46 |
| rm_work | :D | 23:46 |
| rm_work | I mean I thought -2 was for "this will never merge, ever, because it is conceptually wrong or not what we want" | 23:46 |
| cgoncalves | hmm, about that. today is last day for PTL nominations | 23:46 |
| rm_work | oh what already? | 23:47 |
| rm_work | seriously? | 23:47 |
| johnsom | No, they extended it to the 14th | 23:47 |
| rm_work | i thought we had one more week | 23:47 |
| rm_work | ok good | 23:47 |
| rm_work | johnsom: get a job already | 23:47 |
| rm_work | geeze | 23:47 |
| johnsom | Working on it | 23:47 |
| rm_work | <3 | 23:47 |
| rm_work | it's been like, a *whole week* | 23:47 |
| johnsom | But, -2 is the only way I could pull it out of the gates/stop the +W from going through with a recheck | 23:48 |
| rm_work | wait, did people actually +W? | 23:48 |
| johnsom | Yes | 23:48 |
| rm_work | wutlol | 23:48 |
| rm_work | k | 23:48 |
| rm_work | now i am also curious. but, i'll wait for cgoncalves to post his findings so we don't duplicate effort | 23:49 |
| cgoncalves | I'm super slow on this. you folks have way more experience so don't make me hold you | 23:49 |
| johnsom | I am fighting with sphinx that doesn't like my prelude... | 23:50 |
| cgoncalves | that's because you missed tags | 23:50 |
| johnsom | It has a super helpful error: "<reno.sphinxext current branch>:16:Unexpected indentation." | 23:50 |
| johnsom | Doesn't even say which file, but I have figured that part out. | 23:50 |
| rm_work | lol | 23:51 |
| johnsom | Ah, yeah, should add tags too | 23:51 |
| rm_work | such full of help | 23:51 |
| rm_work | ah also now it's merge-conflicting | 23:53 |
| rm_work | lol | 23:53 |
| rm_work | dear god well | 23:56 |
| rm_work | i have a fix i want to push anyway just for formatting | 23:57 |
| rm_work | and merge-conflict | 23:57 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!
