| *** abaindur has quit IRC | 00:05 | |
| *** abaindur has joined #openstack-lbaas | 00:14 | |
| *** abaindur has quit IRC | 00:14 | |
| *** abaindur has joined #openstack-lbaas | 00:14 | |
| *** goldyfruit has joined #openstack-lbaas | 00:39 | |
| *** abaindur has quit IRC | 00:43 | |
| *** abaindur has joined #openstack-lbaas | 00:44 | |
| *** sapd1 has joined #openstack-lbaas | 00:54 | |
| *** redrobot has quit IRC | 01:07 | |
| openstackgerrit | Brian Haley proposed openstack/octavia master: Make octavia-grenade job use python3 https://review.opendev.org/693486 | 01:13 |
|---|---|---|
| *** haleyb has joined #openstack-lbaas | 01:16 | |
| *** goldyfruit has quit IRC | 01:43 | |
| rm_work | Then maybe we just need to make that job non-voting in your change, and then add a follow-up that makes it voting agaib | 01:54 |
| *** ricolin has joined #openstack-lbaas | 02:04 | |
| *** chungpht has joined #openstack-lbaas | 02:27 | |
| *** goldyfruit has joined #openstack-lbaas | 02:41 | |
| openstackgerrit | Sam Morrison proposed openstack/octavia-lib master: Add availability_zone to the LoadBalancer model https://review.opendev.org/694707 | 02:58 |
| rm_work | johnsom: thoughts? | 02:59 |
| rm_work | sorrison: ah making a minor fix too | 03:01 |
| rm_work | quickly to avoid wasting a ton more time | 03:01 |
| openstackgerrit | Sam Morrison proposed openstack/octavia-lib master: Add availability_zone to the LoadBalancer model https://review.opendev.org/694707 | 03:02 |
| sorrison | And another one | 03:02 |
| sorrison | Yeah I have not been able to figure out the circular dependency on the 2 patches | 03:02 |
| rm_work | ahh you fixed it | 03:02 |
| rm_work | haha | 03:02 |
| rm_work | i was clicking "publish" on a quick edit to fix the spelling | 03:02 |
| rm_work | but you beat me to it | 03:03 |
| sorrison | gotta be quick! | 03:03 |
| rm_work | so yeah i'm thinking we do this: | 03:03 |
| *** abaindur has quit IRC | 03:05 | |
| *** abaindur has joined #openstack-lbaas | 03:06 | |
| openstackgerrit | Adam Harwell proposed openstack/octavia-lib master: Add availability_zone to the LoadBalancer model https://review.opendev.org/694707 | 03:07 |
| openstackgerrit | Adam Harwell proposed openstack/octavia-lib master: Return tips jobs to voting https://review.opendev.org/694895 | 03:07 |
| rm_work | thoughts? johnsom sorrison | 03:07 |
| sorrison | once the octavia-lib patch goes in then octavia will start failing until https://review.opendev.org/#/c/693762/9. gets merged | 03:08 |
| sorrison | so basically they need to go in together to avoid breaking other stuff | 03:09 |
| rm_work | yes | 03:09 |
| rm_work | we need to coordinate | 03:09 |
| rm_work | so first we need to merge johnsom's patch that is before my API patch, and also merge my API patch | 03:10 |
| rm_work | then we merge the tips job, then merge yours, then the tips fix | 03:10 |
| sorrison | Yip, and I'm pretty happy with my patch so welcome any feedback | 03:10 |
| sorrison | think I got all the test coverage I need | 03:10 |
| rm_work | err sec gotta fix that i messed something up | 03:10 |
| rm_work | wait where is the tips run coming from O_o | 03:11 |
| openstackgerrit | Adam Harwell proposed openstack/octavia-lib master: Add availability_zone to the LoadBalancer model https://review.opendev.org/694707 | 03:13 |
| rm_work | wtf | 03:14 |
| rm_work | i am bad | 03:14 |
| openstackgerrit | Adam Harwell proposed openstack/octavia-lib master: Add availability_zone to the LoadBalancer model https://review.opendev.org/694707 | 03:15 |
| rm_work | OK | 03:16 |
| rm_work | that worked, it's running as non-voting now | 03:16 |
| rm_work | and it *doesn't* run in Gate, only Check | 03:16 |
| rm_work | added depends-on too so it will be easier to merge in the right order | 03:17 |
| rm_work | eugh my lib change needs to merge too | 03:18 |
| rm_work | but i need to fix the pep8 | 03:18 |
| rm_work | dumb whitespace errors | 03:19 |
| openstackgerrit | Adam Harwell proposed openstack/octavia-lib master: Availability zone / metadata validation https://review.opendev.org/694057 | 03:20 |
| rm_work | cgoncalves: re-review https://review.opendev.org/#/c/692427/ ? | 03:20 |
| rm_work | would love to get that merged | 03:20 |
| rm_work | do you think the SQL query thing is a *blocker*? | 03:21 |
| rm_work | I agree with all of michael's points here, personally | 03:21 |
| rm_work | you and ann aren't *wrong*, it's just, not really a big enough issue to be worth thinking about there IMO | 03:22 |
| openstackgerrit | Adam Harwell proposed openstack/octavia-lib master: Add availability_zone to the LoadBalancer model https://review.opendev.org/694707 | 03:23 |
| rm_work | OK SO, if we can manage to *not touch* any of these CRs for a moment... the checks can all run and hopefully mostly pass | 03:24 |
| openstackgerrit | Adam Harwell proposed openstack/octavia-lib master: Return tips jobs to voting https://review.opendev.org/694895 | 03:24 |
| *** abaindur has quit IRC | 03:24 | |
| rm_work | otherwise we basically have to recheck every single one lol | 03:24 |
| rm_work | they run quick though in octavia-lib, no dsvm jobs | 03:25 |
| *** goldyfruit has quit IRC | 04:00 | |
| rm_work | sorrison: remember that in octavia, the first review field is "Backport-Candidate" :) | 04:08 |
| sorrison | yeah that is tripping me up all the time! | 04:09 |
| rm_work | haha yeah | 04:09 |
| sorrison | fixed :-) | 04:10 |
| rm_work | so ok, maybe you should provide feedback on my troll commit here: https://review.opendev.org/#/c/683028/ XD | 04:10 |
| sorrison | lol | 04:10 |
| sorrison | I am real | 04:11 |
| *** abaindur has joined #openstack-lbaas | 04:32 | |
| rm_work | added all of the patches in-order to the priority review etherpad, in their own list | 04:57 |
| *** abaindur has quit IRC | 05:22 | |
| *** abaindur has joined #openstack-lbaas | 05:23 | |
| *** pcaruana has joined #openstack-lbaas | 05:43 | |
| *** openstackstatus has joined #openstack-lbaas | 06:09 | |
| *** ChanServ sets mode: +v openstackstatus | 06:09 | |
| openstackgerrit | Merged openstack/octavia master: Update flavor guide to be cut/paste friendly https://review.opendev.org/693751 | 06:27 |
| *** AlexStaf has joined #openstack-lbaas | 06:32 | |
| *** abaindur has quit IRC | 06:44 | |
| *** abaindur has joined #openstack-lbaas | 06:44 | |
| *** pcaruana has quit IRC | 06:56 | |
| *** rcernin has quit IRC | 06:58 | |
| openstackgerrit | Ajay Kumar proposed openstack/octavia-lib master: OpenStack is dropping the py2.7 support in ussuri cycle. https://review.opendev.org/692077 | 07:26 |
| *** gcheresh has joined #openstack-lbaas | 08:09 | |
| *** ivve has joined #openstack-lbaas | 08:17 | |
| *** gcheresh_ has joined #openstack-lbaas | 08:18 | |
| *** gcheresh has quit IRC | 08:18 | |
| *** tesseract has joined #openstack-lbaas | 08:18 | |
| *** pcaruana has joined #openstack-lbaas | 08:36 | |
| *** abaindur has quit IRC | 08:38 | |
| *** luksky has joined #openstack-lbaas | 08:42 | |
| *** tkajinam has quit IRC | 08:56 | |
| *** rpittau|afk is now known as rpittau | 09:08 | |
| *** pzy1 has joined #openstack-lbaas | 10:34 | |
| pzy1 | Hi, In LBaaS v1 and v2 version, There is no high availiblity function. Why does not have this function? | 10:36 |
| cgoncalves | pzy1, LBaaS v1 is unsupported. neutron-lbaas with the haproxy driver has some level of HA if I remember it right | 10:57 |
| cgoncalves | pzy1, Octavia (which implements a superset of the v2 API) offers high availability | 10:57 |
| cgoncalves | active-standby load balancers | 10:57 |
| *** pzy1 has quit IRC | 11:32 | |
| *** pzy1 has joined #openstack-lbaas | 11:40 | |
| *** pzy151 has joined #openstack-lbaas | 11:46 | |
| *** pzy151 has quit IRC | 11:47 | |
| *** pzy1 has quit IRC | 11:47 | |
| *** pzy1 has joined #openstack-lbaas | 11:47 | |
| pzy1 | Why doesn't provide HA function in LBaaS v2? What considerations does the community have? For example, It's up to users to use keepalived to achieve high availability? | 11:52 |
| *** pzy1 has quit IRC | 12:17 | |
| *** goldyfruit has joined #openstack-lbaas | 12:18 | |
| *** icey has quit IRC | 12:34 | |
| *** goldyfruit has quit IRC | 13:08 | |
| *** tkajinam has joined #openstack-lbaas | 13:56 | |
| *** tkajinam has quit IRC | 13:56 | |
| *** tkajinam has joined #openstack-lbaas | 13:57 | |
| *** luksky has quit IRC | 14:02 | |
| rm_work | LbaasV2 is an API spec. There are multiple implementations with multiple providers | 14:11 |
| rm_work | pzy1: neutron lbaas had a default provider that did not support HA | 14:12 |
| rm_work | Ah they're gone | 14:12 |
| *** goldyfruit has joined #openstack-lbaas | 14:19 | |
| *** goldyfruit_ has joined #openstack-lbaas | 14:28 | |
| *** goldyfruit has quit IRC | 14:30 | |
| *** tkajinam has quit IRC | 15:00 | |
| *** ataraday_ has joined #openstack-lbaas | 15:29 | |
| johnsom | Yeah, neutron-lbaas never was HA without third party drivers or running Octavia | 15:30 |
| *** ivve has quit IRC | 15:37 | |
| *** gcheresh_ has quit IRC | 15:43 | |
| *** pcaruana has quit IRC | 15:50 | |
| *** pcaruana has joined #openstack-lbaas | 15:51 | |
| *** armax has quit IRC | 15:54 | |
| *** armax has joined #openstack-lbaas | 15:56 | |
| rm_work | cgoncalves: https://review.opendev.org/#/c/692427/ | 15:58 |
| rm_work | xgerman: https://review.opendev.org/#/c/692427/ :D | 15:59 |
| openstackgerrit | Ann Taraday proposed openstack/octavia master: Add option to set default ssl ciphers in haproxy https://review.opendev.org/685337 | 16:29 |
| *** abaindur has joined #openstack-lbaas | 16:39 | |
| *** tesseract has quit IRC | 16:51 | |
| *** rpittau is now known as rpittau|afk | 17:06 | |
| cgoncalves | the in-use flavor patch is a well engineered fix | 17:09 |
| johnsom | Thank you | 17:11 |
| *** ivve has joined #openstack-lbaas | 17:21 | |
| ataraday_ | rm_work, Do you have time for quick review of https://review.opendev.org/#/c/693144/ (octaviaclient fix) | 17:27 |
| openstackgerrit | Brian Haley proposed openstack/octavia master: Stop testing python 2 https://review.opendev.org/687370 | 17:37 |
| rm_work | ataraday_: testing it in my env now | 17:41 |
| rm_work | and using some UUID named LBs :D | 17:41 |
| rm_work | ataraday_: so, this only addresses the "show" case -- which means that delete/update now technically follow a different code-path for lookups than a show | 17:44 |
| rm_work | that's a *little* concerning give that something could feasibly then cause a "show" to work fine, and a subsequent update/delete to fail or behave differently -- frustrating for a user, and not ideal | 17:44 |
| rm_work | i don't know exactly what case would cause that at the moment, but different code path == possible | 17:44 |
| rm_work | commenting as much | 17:45 |
| rm_work | i kinda want to hack up our coverage gate to check out master and run coverage first, THEN check out the working-branch again and run again, and compare coverage rates to ensure it doesn't ever drop. | 17:48 |
| rm_work | let's see though... in my env, with your patch: show-by-id ~= 5.2s, show-by-name ~= 6.2s, show-by-uuid-name ~= 5.9s (wtf? should be higher than both...) | 17:57 |
| rm_work | without patch: show-by-id: ~5.9s, show-by-name ~= 6.2s, show-by-uuid-name ~= 6.0s | 17:59 |
| rm_work | basically seems like everything is within margin of error practically, other than about a 20% decrease for the show-by-id case | 18:00 |
| rm_work | which, yeah, that's not too bad :) | 18:00 |
| rm_work | though honestly there's so much overhead in the client that it's really painful to do this testing using the CLI | 18:01 |
| *** ricolin has quit IRC | 18:04 | |
| ataraday_ | rm_work, may be other operation can be done in follow up changes? | 18:07 |
| ataraday_ | It is perfect variant to refactor all client code, but not sure I have time for this right now | 18:10 |
| *** armax has quit IRC | 18:11 | |
| rm_work | hah yeah, understood | 18:11 |
| rm_work | hmm | 18:12 |
| rm_work | let me think | 18:12 |
| rm_work | i'm just concerned that we could say "ok yeah, followup for the rest" and it will literally never happen until something breaks T_T | 18:13 |
| ataraday_ | I understand that, but at same time we could have something improved now - and have an open story for improvement the rest. | 18:15 |
| rm_work | yeah | 18:16 |
| rm_work | i said i'd think on it :D | 18:16 |
| rm_work | that may also mean i try to find time to do it myself | 18:16 |
| ataraday_ | it will be hard to find time to do everything :) | 18:30 |
| rm_work | alright, i guess i'm ok with this in the meantime if johnsom is ok with it too (given the comment I made) | 18:31 |
| rm_work | i know he already +2'd it | 18:31 |
| johnsom | looking | 18:32 |
| johnsom | Yeah, technically update and show would follow the slower path still. | 18:34 |
| johnsom | I personally am ok with incremental fixes. | 18:34 |
| rm_work | it's less that it's still "slower" but that it's "different" | 18:34 |
| rm_work | alright | 18:34 |
| rm_work | i'll +2 | 18:34 |
| *** abaindur has quit IRC | 18:46 | |
| openstackgerrit | Merged openstack/python-octaviaclient master: Do not get all resources if ID is passed https://review.opendev.org/693144 | 18:46 |
| rm_work | o/ | 18:47 |
| rm_work | now you: https://review.opendev.org/#/c/691135/ :D | 18:47 |
| *** armax has joined #openstack-lbaas | 18:51 | |
| openstackgerrit | Merged openstack/octavia master: Stop allowing the deletion of an in-use flavor https://review.opendev.org/692427 | 18:56 |
| rm_work | sorrison: so for tempest... Should I start a patch that adds the clients and stuff for the AZs? Do you know enough about nova to be able to tweak a gate to actually have a second AZ? | 18:58 |
| colin- | i'm seeing an odd behavior where operating status for a pool and member-less VIP stays in an OFFLINE state in one region but in another it transitions to ONLINE when configured identically | 19:09 |
| colin- | what absent connectivity or misconfiguration might account for this, any tips? | 19:09 |
| colin- | pool-less and member-less, the idea is just to validate that the HAProxy listener is receiving TCP conns, nothing else | 19:09 |
| *** TrevorV has joined #openstack-lbaas | 19:11 | |
| johnsom | rm_work only 20 more lines of coverage for client and we could bump the minimum to 95% | 19:12 |
| rm_work | Yeah I was looking at that | 19:12 |
| rm_work | Really just need to make a "test_utils.py" | 19:12 |
| johnsom | colin- It should be offline. Is the health manager in the region with the "ONLINE" getting it's heartbeat packets? | 19:13 |
| johnsom | Are they both TCP listeners? | 19:13 |
| eandersson | johnsom does the lb start as online until the heartbeat is received? | 19:13 |
| rm_work | Does one of them have a healthmonitor and one not? | 19:14 |
| colin- | they are both TCP mons | 19:14 |
| rm_work | Hmm k | 19:14 |
| colin- | sorry TCP listeners | 19:14 |
| johnsom | I don't think so, I think it starts offline, but haven't looked at that code in a while | 19:14 |
| colin- | no monitors configured on either | 19:14 |
| colin- | it does start offline | 19:14 |
| colin- | i validated that | 19:14 |
| rm_work | Yeah I'd think offline would be the expected status | 19:14 |
| colin- | maybe the heartbeats aren't making it on the OFFLINE ones | 19:14 |
| colin- | the UDP5000 | 19:14 |
| colin- | i can check for that | 19:14 |
| johnsom | It starts offline: https://github.com/openstack/octavia/blob/master/octavia/db/prepare.py#L69 | 19:14 |
| johnsom | UDP 5555 | 19:15 |
| johnsom | is the default | 19:15 |
| colin- | meant that | 19:16 |
| *** abaindur has joined #openstack-lbaas | 19:23 | |
| *** abaindur has quit IRC | 19:28 | |
| *** abaindur has joined #openstack-lbaas | 19:28 | |
| *** abaindur has quit IRC | 19:29 | |
| *** abaindur has joined #openstack-lbaas | 19:30 | |
| colin- | nope, that was already permitted | 19:34 |
| colin- | that would have been easy :) | 19:35 |
| openstackgerrit | Brian Haley proposed openstack/octavia master: Make octavia-grenade job use python3 https://review.opendev.org/693486 | 19:36 |
| *** goldyfruit___ has joined #openstack-lbaas | 19:55 | |
| *** gmann is now known as gmann_afk | 19:57 | |
| *** goldyfruit_ has quit IRC | 19:57 | |
| johnsom | colin- Did you figure out what was up? | 20:30 |
| johnsom | Some transactions will reset the operating status, but for the most part it is driven by the observed status of the engine, so would update in the next heartbeat interval | 20:31 |
| *** openstackgerrit has quit IRC | 20:35 | |
| *** gcheresh_ has joined #openstack-lbaas | 20:37 | |
| *** goldyfruit_ has joined #openstack-lbaas | 20:58 | |
| *** goldyfruit___ has quit IRC | 21:00 | |
| *** gmann_afk is now known as gmann | 21:03 | |
| eandersson | johnsom think it was a misconfiguration on the hm | 21:10 |
| johnsom | Ok, cool | 21:11 |
| colin- | yes | 21:35 |
| colin- | oh i forgot erik idles here | 21:35 |
| *** gcheresh_ has quit IRC | 21:36 | |
| eandersson | I don't idle, I lurk. | 21:39 |
| johnsom | He just wishes he could be cool enough to hang here | 21:42 |
| *** openstackgerrit has joined #openstack-lbaas | 21:55 | |
| openstackgerrit | Brian Haley proposed openstack/octavia master: Stop testing python 2 https://review.opendev.org/687370 | 21:55 |
| *** rcernin has joined #openstack-lbaas | 21:59 | |
| *** rcernin has quit IRC | 22:01 | |
| *** rcernin has joined #openstack-lbaas | 22:01 | |
| *** rcernin has quit IRC | 22:01 | |
| *** rcernin has joined #openstack-lbaas | 22:02 | |
| colin- | like a mutalisk | 22:07 |
| colin- | oops wrong sc unit | 22:08 |
| colin- | how embarrassing | 22:08 |
| colin- | they were just named lurkers after all | 22:08 |
| eandersson | mutalisks are a great counter to lurkers | 22:08 |
| johnsom | Those designate people trolling for more code to borrow.... | 22:09 |
| eandersson | Can't help it - always need more pylons | 22:18 |
| johnsom | Or more laughs.... | 22:21 |
| eandersson | =] | 22:27 |
| *** TrevorV has quit IRC | 22:29 | |
| *** pcaruana has quit IRC | 22:33 | |
| *** ataraday_ has quit IRC | 23:02 | |
| *** tkajinam has joined #openstack-lbaas | 23:07 | |
| *** abaindur has quit IRC | 23:07 | |
| *** abaindur has joined #openstack-lbaas | 23:09 | |
| *** armax has quit IRC | 23:13 | |
| *** armax has joined #openstack-lbaas | 23:14 | |
| *** armax has quit IRC | 23:16 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Fix filtering with API tags https://review.opendev.org/695112 | 23:22 |
| *** abaindur has quit IRC | 23:34 | |
| *** abaindur has joined #openstack-lbaas | 23:36 | |
| *** ivve has quit IRC | 23:54 | |
| *** goldyfruit_ has quit IRC | 23:59 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!