| openstackgerrit | Michael Johnson proposed openstack/octavia master: Adds v2 load balancer API section https://review.openstack.org/438757 | 00:05 |
|---|---|---|
| *** armax has quit IRC | 00:26 | |
| *** amotoki has joined #openstack-lbaas | 00:26 | |
| *** nrado has quit IRC | 00:31 | |
| *** catintheroof has joined #openstack-lbaas | 00:35 | |
| *** gongysh has quit IRC | 00:42 | |
| *** reedip has quit IRC | 00:59 | |
| *** catintheroof has quit IRC | 01:08 | |
| *** yamamoto has joined #openstack-lbaas | 01:12 | |
| *** ducnc has quit IRC | 01:35 | |
| *** gongysh has joined #openstack-lbaas | 01:50 | |
| openstackgerrit | Zhi Chang proposed openstack/neutron-lbaas master: Add Loadbalancer migration https://review.openstack.org/447177 | 02:15 |
| *** ipsecguy_ has quit IRC | 02:22 | |
| *** ipsecguy has joined #openstack-lbaas | 02:22 | |
| *** ducnc has joined #openstack-lbaas | 02:25 | |
| *** reedip has joined #openstack-lbaas | 02:27 | |
| *** yamamoto has quit IRC | 02:41 | |
| *** yamamoto has joined #openstack-lbaas | 02:48 | |
| *** aojea has joined #openstack-lbaas | 02:48 | |
| *** sanfern has joined #openstack-lbaas | 02:52 | |
| *** aojea has quit IRC | 02:53 | |
| reedip | johnsom : there is no project in Launchpad for Octaviaclient | 03:00 |
| johnsom | reedip it is octavia in launchpad | 03:00 |
| johnsom | I don't think there is enough to want a separate launchpad just for the client | 03:01 |
| reedip | but should bugs of octaviaclient logged in octavia ? | 03:01 |
| johnsom | Yes | 03:02 |
| reedip | johnsom : thats why I mentioned OSC :D | 03:02 |
| reedip | but anyways, got your point for now | 03:02 |
| johnsom | Yeah, I understand, it is just openstackclient is not the right place. I put octavia launchpad in for the bugs/launchpad in infra. It is just hidden in got | 03:03 |
| johnsom | Git | 03:03 |
| reedip | lol | 03:04 |
| openstackgerrit | Zhi Chang proposed openstack/neutron-lbaas master: Add Loadbalancer migration https://review.openstack.org/447177 | 03:07 |
| openstackgerrit | Zhi Chang proposed openstack/neutron-lbaas master: Add Loadbalancer migration https://review.openstack.org/447177 | 03:13 |
| *** armax has joined #openstack-lbaas | 03:24 | |
| *** yamamoto has quit IRC | 03:25 | |
| openstackgerrit | Zhi Chang proposed openstack/neutron-lbaas master: Add Loadbalancer migration https://review.openstack.org/447177 | 03:26 |
| *** links has joined #openstack-lbaas | 03:44 | |
| *** MarkMielke has joined #openstack-lbaas | 03:49 | |
| *** yamamoto has joined #openstack-lbaas | 04:09 | |
| *** sticker_ has quit IRC | 04:19 | |
| *** sticker_ has joined #openstack-lbaas | 04:19 | |
| *** sticker_ has quit IRC | 04:31 | |
| *** sticker_ has joined #openstack-lbaas | 04:32 | |
| *** rcernin has joined #openstack-lbaas | 05:43 | |
| *** rcernin has quit IRC | 05:57 | |
| openstackgerrit | Zhi Chang proposed openstack/neutron-lbaas master: Add Loadbalancer migration https://review.openstack.org/447177 | 06:18 |
| openstackgerrit | huangshan proposed openstack/octavia master: Remove log translations from octavia https://review.openstack.org/447872 | 06:26 |
| *** gcheresh_ has joined #openstack-lbaas | 06:28 | |
| *** aojea has joined #openstack-lbaas | 06:30 | |
| *** aojea has quit IRC | 06:54 | |
| *** amotoki_ has joined #openstack-lbaas | 07:00 | |
| *** amotoki has quit IRC | 07:03 | |
| *** sticker_ is now known as sticker | 07:05 | |
| *** krypto has joined #openstack-lbaas | 07:08 | |
| *** krypto has joined #openstack-lbaas | 07:08 | |
| *** amotoki has joined #openstack-lbaas | 07:18 | |
| *** krypto has quit IRC | 07:21 | |
| *** amotoki_ has quit IRC | 07:21 | |
| *** krypto has joined #openstack-lbaas | 07:21 | |
| openstackgerrit | Reedip proposed openstack/octavia master: Introduce Octavia v2 API for Health Monitor https://review.openstack.org/406256 | 07:26 |
| openstackgerrit | Reedip proposed openstack/octavia master: Introduce Octavia v2 API for Health Monitor https://review.openstack.org/406256 | 07:27 |
| *** nrado has joined #openstack-lbaas | 07:28 | |
| *** eezhova has joined #openstack-lbaas | 07:31 | |
| *** tesseract has joined #openstack-lbaas | 07:35 | |
| openstackgerrit | huangshan proposed openstack/neutron-lbaas master: Remove log translations from neutron-lbaas https://review.openstack.org/447921 | 07:36 |
| *** ptoohill- has quit IRC | 07:52 | |
| *** ptoohill has quit IRC | 07:53 | |
| *** pcaruana has joined #openstack-lbaas | 08:11 | |
| *** aojea has joined #openstack-lbaas | 08:21 | |
| *** krypto has quit IRC | 08:25 | |
| *** krypto has joined #openstack-lbaas | 08:26 | |
| openstackgerrit | Reedip proposed openstack/octavia master: Introduce Octavia v2 API for Health Monitor https://review.openstack.org/406256 | 08:26 |
| -openstackstatus- NOTICE: Wiki is broken with database problems, we are working to resolve it | 08:31 | |
| *** ChanServ changes topic to "Wiki is broken with database problems, we are working to resolve it" | 08:31 | |
| *** eezhova has quit IRC | 08:31 | |
| *** ChanServ changes topic to "Welcome to LBaaS / Octavia" | 08:40 | |
| -openstackstatus- NOTICE: Wiki problems have been fixed, it's up and running | 08:40 | |
| openstackgerrit | Zhi Chang proposed openstack/neutron-lbaas master: Add Loadbalancer migration https://review.openstack.org/447177 | 08:44 |
| *** reedip has quit IRC | 08:56 | |
| *** eezhova has joined #openstack-lbaas | 09:02 | |
| *** openstackgerrit has quit IRC | 09:03 | |
| *** sanfern has quit IRC | 09:04 | |
| *** sanfern has joined #openstack-lbaas | 09:04 | |
| *** openstackgerrit has joined #openstack-lbaas | 09:06 | |
| openstackgerrit | huangshan proposed openstack/octavia master: Remove log translations from octavia https://review.openstack.org/447872 | 09:06 |
| *** reedip has joined #openstack-lbaas | 09:09 | |
| *** sanfern has quit IRC | 09:35 | |
| *** sanfern has joined #openstack-lbaas | 09:37 | |
| *** sanfern has quit IRC | 09:58 | |
| *** sanfern has joined #openstack-lbaas | 10:01 | |
| *** nrado has quit IRC | 10:03 | |
| openstackgerrit | Zhi Chang proposed openstack/neutron-lbaas master: Add Loadbalancer migration https://review.openstack.org/447177 | 10:07 |
| openstackgerrit | Ji Chengke proposed openstack/octavia master: Use HostAddressOpt for opts that accept IP and hostnames https://review.openstack.org/447989 | 10:11 |
| *** amotoki has quit IRC | 10:17 | |
| rm_work | hmm ok surprisingly that did not fix my dib problem... | 10:18 |
| reedip | rm_work : which one? | 10:18 |
| rm_work | I thought this would fix it: https://review.openstack.org/#/c/447701/1 | 10:18 |
| rm_work | actually I should probably remove my +A | 10:19 |
| *** reedip has quit IRC | 10:30 | |
| *** kobis1 has joined #openstack-lbaas | 10:30 | |
| *** kobis has quit IRC | 10:31 | |
| *** krypto has quit IRC | 10:41 | |
| *** nrado has joined #openstack-lbaas | 10:49 | |
| *** krypto has joined #openstack-lbaas | 10:51 | |
| *** sanfern has quit IRC | 10:51 | |
| *** faizy has joined #openstack-lbaas | 10:56 | |
| *** krypto has quit IRC | 10:57 | |
| *** faizy has quit IRC | 10:59 | |
| *** faizy has joined #openstack-lbaas | 11:00 | |
| *** aojea_ has joined #openstack-lbaas | 11:01 | |
| *** aojea has quit IRC | 11:03 | |
| *** krypto has joined #openstack-lbaas | 11:10 | |
| *** krypto has quit IRC | 11:10 | |
| *** krypto has joined #openstack-lbaas | 11:10 | |
| *** faizy_ has joined #openstack-lbaas | 11:24 | |
| *** faizy has quit IRC | 11:25 | |
| *** reedip has joined #openstack-lbaas | 11:33 | |
| *** gongysh has quit IRC | 11:33 | |
| *** amotoki has joined #openstack-lbaas | 11:36 | |
| *** sanfern has joined #openstack-lbaas | 11:40 | |
| openstackgerrit | huangshan proposed openstack/octavia master: Remove log translations from octavia https://review.openstack.org/447872 | 11:41 |
| *** faizy_ has quit IRC | 11:48 | |
| *** krypto has quit IRC | 12:00 | |
| *** krypto has joined #openstack-lbaas | 12:00 | |
| *** krypto has quit IRC | 12:00 | |
| *** krypto has joined #openstack-lbaas | 12:00 | |
| *** krypto has quit IRC | 12:28 | |
| *** krypto has joined #openstack-lbaas | 12:30 | |
| *** gcheresh has joined #openstack-lbaas | 12:36 | |
| *** gcheresh_ has quit IRC | 12:36 | |
| openstackgerrit | Nir Magnezi proposed openstack/octavia master: Auto detect haproxy user_group https://review.openstack.org/429398 | 12:41 |
| openstackgerrit | Merged openstack/octavia master: Remove local DIB elements dir https://review.openstack.org/447701 | 12:46 |
| openstackgerrit | huangshan proposed openstack/neutron-lbaas master: Remove log translations from neutron-lbaas https://review.openstack.org/447921 | 12:48 |
| *** gongysh has joined #openstack-lbaas | 13:02 | |
| *** gongysh has quit IRC | 13:07 | |
| *** gcheresh_ has joined #openstack-lbaas | 13:10 | |
| *** gcheresh has quit IRC | 13:10 | |
| *** gongysh has joined #openstack-lbaas | 13:13 | |
| *** gongysh has quit IRC | 13:15 | |
| *** links has quit IRC | 13:24 | |
| *** faizy_ has joined #openstack-lbaas | 13:31 | |
| *** faizy__ has joined #openstack-lbaas | 13:33 | |
| *** blogan has quit IRC | 13:36 | |
| *** faizy_ has quit IRC | 13:36 | |
| *** blogan has joined #openstack-lbaas | 13:36 | |
| *** faizy__ has quit IRC | 13:51 | |
| *** catintheroof has joined #openstack-lbaas | 13:57 | |
| *** chlong has joined #openstack-lbaas | 14:18 | |
| *** chlong has quit IRC | 14:24 | |
| rm_work | grrrr wtf | 14:31 |
| rm_work | curl -X POST http://127.0.0.1:9876/v2.0/lbaas/loadbalancers.json -H "X-Auth-Token: $MYTOKEN" -d '{"loadbalancer": {"vip_subnet_id": "'$DEFAULT_NETWORK'", "name": "lb1", "admin_state_up": true}}' | 14:31 |
| rm_work | is that wrong? | 14:31 |
| rm_work | getting: | 14:32 |
| rm_work | 2017-03-21 23:28:28.655 16542 DEBUG wsme.api [req-aca54109-b5ef-4a8b-9904-a3f651757de1 - 85e70746bc5d44888d3a49f427d4299d - default default] Client-side error: Unknown argument: "{"loadbalancer": {"vip_subnet_id": "183f7734-45ba-44d0-9bed-acd01e3e13d3", "name": "lb1", "admin_state_up": true}}" format_exception /usr/local/lib/python2.7/dist-packages/wsme/api.py:222 | 14:32 |
| rm_work | This did work before, I thought | 14:33 |
| rm_work | that's really weird wording, too | 14:36 |
| rm_work | "unknown argument: <the whole request>" | 14:36 |
| rm_work | it's like it isn't unpacking the json | 14:36 |
| reedip | thats weird | 14:36 |
| *** diltram_ has joined #openstack-lbaas | 14:37 | |
| *** diltram has quit IRC | 14:37 | |
| reedip | rm_work : which patch ? | 14:37 |
| rm_work | ah got it | 14:38 |
| rm_work | apparently with curl I do still have to send content-type json | 14:38 |
| rm_work | i thought it'd assume it | 14:38 |
| reedip | no... I dont think so rm_work | 14:39 |
| rm_work | well, obviously it does not :P | 14:39 |
| reedip | no I mean, generally it picks up json itself.. | 14:39 |
| *** faizy has joined #openstack-lbaas | 14:57 | |
| *** armax has quit IRC | 14:57 | |
| rm_work | grrr parallels snapshots totally destroy networking | 14:58 |
| rm_work | so i can't stack and then snapshot and revert for testing purposes :( | 14:59 |
| *** pck is now known as pckizer | 14:59 | |
| *** faizy has quit IRC | 15:10 | |
| *** faizy has joined #openstack-lbaas | 15:11 | |
| *** faizy_ has joined #openstack-lbaas | 15:14 | |
| *** faizy__ has joined #openstack-lbaas | 15:17 | |
| *** faizy has quit IRC | 15:18 | |
| johnsom | Oye, log translation stuff.... | 15:20 |
| johnsom | That was not well managed.... | 15:20 |
| *** faizy_ has quit IRC | 15:20 | |
| *** faizy__ has quit IRC | 15:21 | |
| *** faizy__ has joined #openstack-lbaas | 15:21 | |
| reedip | https://review.openstack.org/#/c/447921/ ?? | 15:22 |
| johnsom | Yeah, and octavia | 15:22 |
| johnsom | Just the process of deciding to remove them, etc. | 15:22 |
| *** ptoohill has joined #openstack-lbaas | 15:22 | |
| *** ptoohill1 has joined #openstack-lbaas | 15:23 | |
| *** faizy__ has quit IRC | 15:24 | |
| reedip | I stated the same to kevinbenton earlier today.... this has to be done logically | 15:24 |
| reedip | not bruteforce | 15:24 |
| *** faizy__ has joined #openstack-lbaas | 15:24 | |
| johnsom | Yeah, i.e. start with the hacking rules which everyone is either ignoring or just deleting instead of enforcing the "don't" | 15:25 |
| nmagnezi | this patch actually surprised me. I was waiting to see what you will say about it before I review :) | 15:25 |
| johnsom | Yeah, me too. I just dug through the e-mail chains and found it really is a decision. http://lists.openstack.org/pipermail/openstack-dev/2017-March/114191.html | 15:26 |
| reedip | Yup, thats the email ! | 15:26 |
| *** faizy__ has quit IRC | 15:26 | |
| reedip | I had to verify code change in OSC | 15:27 |
| reedip | and in Neutronclient | 15:27 |
| johnsom | I kind of hate to do it. It was a lot of work and frankly I like to have things translated for folks, but I guess if it doesn't add value in today's world.... | 15:27 |
| *** faizy__ has joined #openstack-lbaas | 15:27 | |
| reedip | now LBaaS, FWaaS, octavia ... this is gonna be a big change , with a lot of failures bcz of the brute force nature for patch counts | 15:27 |
| reedip | johnsom : it does, they are going to discuss it once more in Boston , atleast for the Log translations | 15:28 |
| johnsom | Yeah, but it doesn't look like projects are waiting. From what I can see nova has already deleted most of theirs | 15:28 |
| reedip | nova was among the first ones, sean started the mail chain bcz of this | 15:29 |
| *** faizy__ has quit IRC | 15:29 | |
| reedip | but this decision was taken in Barcelona, it just manifested now | 15:29 |
| *** faizy__ has joined #openstack-lbaas | 15:30 | |
| *** krypto has quit IRC | 15:33 | |
| *** krypto has joined #openstack-lbaas | 15:34 | |
| *** krypto has quit IRC | 15:34 | |
| *** krypto has joined #openstack-lbaas | 15:34 | |
| nmagnezi | honestly i don't understand the value of removing those translations. I can understand a decision not to translate logs in the first place, but once it is there, actively removing it seems strange to me. | 15:36 |
| nmagnezi | but maybe I just need to read that mail :-) | 15:36 |
| reedip | nmagnezi :D | 15:36 |
| johnsom | The argument is they don't have the translation resources they used to and that translated logs makes it harder for folks to google the log entries | 15:38 |
| *** kobis has joined #openstack-lbaas | 15:40 | |
| *** kobis1 has quit IRC | 15:41 | |
| *** armax has joined #openstack-lbaas | 15:49 | |
| rm_work | ok, got some comments up for the listener and pool patches | 15:51 |
| rm_work | if we can get movement on those things... soon... that'd be sweet | 15:51 |
| rm_work | i'm like 60 seconds from just *doing it* again | 15:52 |
| rm_work | because I am incredibly impatient :P | 15:52 |
| *** gcheresh_ has quit IRC | 15:52 | |
| reedip | rm_work : the HM patches await your impatience for review :) | 15:53 |
| *** eezhova has quit IRC | 15:53 | |
| rm_work | i'll get to that after i get all of the ones that are necessary for basic functionality | 15:53 |
| rm_work | IE, listeners/pools/members | 15:53 |
| rm_work | hmm, listeners may not be doing quotas? :/ | 15:54 |
| * rm_work looks closer | 15:54 | |
| reedip | sindhu : I think we need to add quotas | 15:54 |
| rm_work | yeah looks like it's missing | 15:56 |
| rm_work | soooo | 15:56 |
| johnsom | I had planned to come back and do dependent patches for quota like we did on lb for project_id and vip_network_id. | 16:01 |
| johnsom | Don't let me stop you from doing so.... Grin | 16:01 |
| rm_work | erg | 16:02 |
| rm_work | well | 16:02 |
| rm_work | i will give people like | 16:02 |
| rm_work | until morning | 16:02 |
| rm_work | if it's not fixed by then i will just do it | 16:02 |
| rm_work | I dunno if that's rude though... | 16:03 |
| rm_work | but i've got nothing to do right now besides focus on this, and it's my #1 priority | 16:03 |
| rm_work | and i would like this stuff to merge soon, so the faster we can get iterations going | 16:03 |
| rm_work | the better | 16:03 |
| johnsom | Yeah, don't be idle.... | 16:03 |
| rm_work | sindhu | 16:03 |
| johnsom | Ping folks and if they are working on other tasks, ask if you can finish up the patches. | 16:04 |
| rm_work | I assume sindhu is working on it as he's put -W on the patch | 16:04 |
| johnsom | I think she is.... | 16:04 |
| johnsom | I will bring this up in our sync meeting today. | 16:05 |
| rm_work | ah | 16:05 |
| rm_work | excellent | 16:05 |
| sindhu | rm_work: yup working on listeners and health monitor today | 16:06 |
| rm_work | sweet | 16:07 |
| rm_work | hoping you can get the quotas/project_id stuff added to listeners, along with the loadbalancer_id name fix | 16:07 |
| *** krypto has quit IRC | 16:12 | |
| *** amotoki has quit IRC | 16:13 | |
| rm_work | or if not quotas yet, then at least the project_id code | 16:16 |
| rm_work | so it'll work with auth enabled | 16:16 |
| sindhu | yup on it :) | 16:19 |
| *** krypto has joined #openstack-lbaas | 16:25 | |
| *** krypto has quit IRC | 16:32 | |
| *** eezhova has joined #openstack-lbaas | 16:37 | |
| *** faizy__ has quit IRC | 16:40 | |
| *** krypto has joined #openstack-lbaas | 16:49 | |
| *** eezhova has quit IRC | 16:55 | |
| *** eezhova has joined #openstack-lbaas | 16:55 | |
| *** aojea_ has quit IRC | 16:59 | |
| *** krypto has quit IRC | 17:18 | |
| *** krypto has joined #openstack-lbaas | 17:19 | |
| *** krypto has quit IRC | 17:24 | |
| *** krypto has joined #openstack-lbaas | 17:25 | |
| *** nrado has quit IRC | 17:28 | |
| *** catintheroof has quit IRC | 17:30 | |
| *** catintheroof has joined #openstack-lbaas | 17:30 | |
| *** beagles is now known as beagles_lunch | 17:48 | |
| *** eezhova has quit IRC | 18:01 | |
| *** markd_ has quit IRC | 18:04 | |
| johnsom | reedip Are you still around? | 18:05 |
| *** eezhova has joined #openstack-lbaas | 18:08 | |
| *** aojea has joined #openstack-lbaas | 18:17 | |
| rm_work | out for a bit | 18:18 |
| *** krypto has quit IRC | 18:19 | |
| *** blogan_ has joined #openstack-lbaas | 18:20 | |
| *** gcheresh_ has joined #openstack-lbaas | 18:21 | |
| *** blogan has quit IRC | 18:23 | |
| ankur-gupta-f4 | johnsom: | 18:26 |
| ankur-gupta-f4 | hes in india time zone fyi. | 18:26 |
| johnsom | Ok, I was just checking. I thought he was somewhere in asia | 18:28 |
| johnsom | well, like closer to japan time | 18:29 |
| ankur-gupta-f4 | yea something like that | 18:29 |
| *** blogan_ is now known as blogan | 18:30 | |
| *** tesseract has quit IRC | 18:31 | |
| *** beagles_lunch is now known as beagles | 18:33 | |
| *** pcaruana has quit IRC | 18:39 | |
| *** aojea has quit IRC | 18:40 | |
| *** aojea has joined #openstack-lbaas | 18:46 | |
| rm_work | wtf, do listeners have ... ONE load balancer? or many? | 18:46 |
| rm_work | I thought listeners couldn't possibly be shared | 18:46 |
| rm_work | but the return in neutron-lbaas for a listener | 18:46 |
| rm_work | has an ARRAY of "loadbalancers" which are objects with 'id': uuid | 18:47 |
| rm_work | https://developer.openstack.org/api-ref/networking/v2/?expanded=list-listeners-detail#id657 | 18:48 |
| rm_work | how is that even possible | 18:48 |
| rm_work | shouldn't it be one loadbalancer_id per listener? | 18:48 |
| *** armax has quit IRC | 18:49 | |
| johnsom | load balancers have 0 or more listeners. listeners cannot be shared between lbs, but do get instantiated on more than one amp. | 18:51 |
| johnsom | Also, rm_work, I would not trust those docs at all. They do not match reality..... | 18:54 |
| rm_work | hmmmm k | 18:54 |
| johnsom | When they mixed the v1 api and the v2 api, the seperation back out was not clean.... | 18:54 |
| sindhu | rm_work: ping? | 18:54 |
| rm_work | yeah k | 18:55 |
| rm_work | sindhu: sup | 18:55 |
| sindhu | rm_work: hey, so if I make the change 'loadbalancer_id' and make corresponding change in the controller to send the request to db in the format that it needs, when the db sends it response it'll hold 'load_balancer_id' but the types will have 'loadbalancer_id'. won't there be a mis match? | 18:57 |
| rm_work | you can resolve that | 18:57 |
| rm_work | one sec | 18:57 |
| johnsom | rm_work However, in this case, it does appear that neutron-lbaas is returning an array.... | 18:59 |
| johnsom | https://www.irccloud.com/pastebin/X8vWaTMT/ | 18:59 |
| johnsom | So, maybe something intended for the future? | 18:59 |
| sindhu | rm_work: will push up the changes I have made, can you take a look | 19:00 |
| rm_work | yeah | 19:00 |
| openstackgerrit | Sindhu Devale proposed openstack/octavia master: Octavia v2 API for listeners https://review.openstack.org/424744 | 19:01 |
| sindhu | will have lunch and be back | 19:01 |
| ndahiwade | johnsom,: quick question | 19:09 |
| *** chlong has joined #openstack-lbaas | 19:13 | |
| *** aojea has quit IRC | 19:17 | |
| *** nrado has joined #openstack-lbaas | 19:17 | |
| *** aojea has joined #openstack-lbaas | 19:18 | |
| rm_work | sindhu: commented... hopefully that makes sense, i'm really tired | 19:22 |
| *** aojea has quit IRC | 19:22 | |
| rm_work | disappearing for now, but I can poke at stuff tomorrow | 19:22 |
| sindhu | rm_work: thank you :) will fix it | 19:27 |
| *** gcheresh_ has quit IRC | 19:36 | |
| *** gcheresh_ has joined #openstack-lbaas | 19:49 | |
| *** armax has joined #openstack-lbaas | 20:02 | |
| *** chlong has quit IRC | 20:21 | |
| johnsom | ndahiwade What is your question? | 20:24 |
| ndahiwade | johnsom, https://review.openstack.org/#/c/431207/3/octavia/api/v1/controllers/l7policy.py | 20:29 |
| ndahiwade | we don't need to get project_id for L7Policy and Rules? | 20:30 |
| m-greene- | Question to the team: neutron_lbaas/tests/tempest/v2/api/test_members_non_admin.py fails in several tests for us (F5 service provider) becuase the tests are configured to use 127.0.0.1. This only makes sense for devstack (per the comment), as any external device is going to reject packets to a loopback address. Does neutron_lbass or octavia have an official stance on these tests? We modified the tests to use 128 and a par | 20:32 |
| m-greene- | is rejecting this change because the tests no longer reflect upstream. | 20:32 |
| johnsom | No, it will inherit from the load balancer | 20:32 |
| ndahiwade | johnsom, got it, thanks:) | 20:33 |
| johnsom | m-greene- The last part "We modified..." got cut off | 20:37 |
| m-greene- | We modified the tests to use 128 and a partner is rejecting this change because the tests no longer reflect upstream. | 20:37 |
| johnsom | Ah | 20:37 |
| johnsom | m-grerene- I would not use 128 as that is a valid network. I would use RFC1918 private address space. | 20:39 |
| johnsom | m-greene- Opps, typo above for your nick. | 20:40 |
| m-greene- | johnsom: That is fine- their argument is more simply against any change to the tests. | 20:40 |
| johnsom | m-greene- Feel free to post a patch to fix the tests | 20:41 |
| m-greene- | this is back on mitaka :) | 20:41 |
| *** aojea has joined #openstack-lbaas | 20:41 | |
| johnsom | m-greene- As you know, these tests need work. I have no problem fixing them to be flexible for the vendors. | 20:41 |
| m-greene- | johnsom: great- I’d expect us to submit a patch and review, but was curious if these tests are gold and we should create another set for vendors. You’ve answered my question- we can change, but need to ensure the ref driver still passes. | 20:43 |
| johnsom | m-greene- Ah, mitaka is a problem. It goes EOL in three weeks and per the stable release policy we can only backport security fixes to mitaka at this time. https://releases.openstack.org/ and https://docs.openstack.org/project-team-guide/stable-branches.html#support-phases | 20:43 |
| m-greene- | johnsom: Understood. F5 (and partner) will need to keep our forks alive for much longer. We have customers at each end of the spectrum. Some pushing us to newton/ocata, and others just moving up from juno/kilo (to liberty/mitaka). | 20:46 |
| *** chlong has joined #openstack-lbaas | 20:48 | |
| openstackgerrit | Sindhu Devale proposed openstack/octavia master: Octavia v2 API for listeners https://review.openstack.org/424744 | 20:48 |
| johnsom | m-greene- Yeah, it is an odd situation for sure. I just need to follow the OpenStack stable policy. We can get a fix in master and I can argue for backports as far as newton for that change. | 20:48 |
| xgerman | yeah, those forks are an elephant in the room | 20:51 |
| *** cody-somerville has joined #openstack-lbaas | 21:15 | |
| reedip | johnsom : just woke up | 21:20 |
| reedip | ankur-gupta-f4 : Philippiness time zone :P | 21:20 |
| ankur-gupta-f4 | Ah yea. Forgot | 21:21 |
| ankur-gupta-f4 | Im in PST now and its throwing me off | 21:21 |
| johnsom | reedip Sorry about that. Yeah, I was close. Grab a cup of whatever your morning beverage is then ping me to talk about the DNS patch. | 21:21 |
| johnsom | ankur-gupta-f4 PST rocks.... | 21:21 |
| * johnsom waves from probably south of you.... | 21:22 | |
| reedip | johnsom : fire away | 21:22 |
| johnsom | reedip So I loaded it up and kicked the tires. I made a few observations and commented on the patch. | 21:23 |
| johnsom | The first issue I hit is that not all of the drivers create the VIP there. Octavia for example creates it's own. | 21:23 |
| johnsom | I also wondered is dns-<lb-id> that much more useful than the default host-<ip> neutron pops on there | 21:24 |
| johnsom | There is also the failover issue and why we explicitly delete the neutron dns name record... | 21:25 |
| reedip | ok | 21:25 |
| johnsom | I was going to try to override the vip creation for the octavia driver and try the failover again, just haven't got there yet. | 21:27 |
| reedip | so , I can probably use the host-<ip> format from neutron, which would have consistency between neutron and octavia | 21:27 |
| johnsom | Yeah, so then you don't need this patch at all, it comes in automagically when we create the port. | 21:27 |
| openstackgerrit | Shashank Kumar Shankar proposed openstack/octavia master: Move API v1, v2 hooks to API Common https://review.openstack.org/448288 | 21:27 |
| johnsom | https://www.irccloud.com/pastebin/kcRMHJJg/ | 21:28 |
| johnsom | And the VIP port: | 21:28 |
| johnsom | https://www.irccloud.com/pastebin/hTGjtkMC/ | 21:28 |
| reedip | johnsom : so basically we need to copy the name to the dns-name | 21:29 |
| reedip | IIUC | 21:29 |
| johnsom | Oh, interesting, I didn't notice that dns_name didn't get filled in. | 21:29 |
| reedip | and that was what the patch was :D | 21:30 |
| johnsom | Do you know where the docs are for this dns extension? It's not in the main neutron API ref.... | 21:30 |
| johnsom | Yeah, but dns_assignment has it, so what is the difference? | 21:30 |
| reedip | johnsom : you mean https://docs.openstack.org/ocata/networking-guide/config-dns-int.html ? | 21:32 |
| reedip | johnsom : the bug was that itself, though we have a DNS Assignment, we do not have a DNS Name, as mentioned in the Link ^^ populated by LBaaS when we create a Loadbalancer | 21:34 |
| reedip | designate may require this | 21:34 |
| johnsom | reedip It is this paragraph that is probably what caused the failures with failover: | 21:36 |
| johnsom | Instead of having the Compute service create the port for the instance, the user might have created it and assigned a value to its dns_name attribute. In this case, the value assigned to the dns_name attribute must be equal to the value that Compute service will assign to the instance’s hostname, in this example my-vm. Otherwise, the instance boot will | 21:36 |
| johnsom | fail. | 21:36 |
| johnsom | And also this paragraph: | 21:37 |
| johnsom | If the port is going to be used for an instance boot, the value assigned to dns_name must be equal to the hostname that the Compute service will assign to the instance. Otherwise, the boot will fail. | 21:37 |
| reedip | johnsom : but in loadbalancer case, would the user be creating a port ? | 21:37 |
| johnsom | Well in the neutron-lbaas case it is the neutron service account creating the port. | 21:38 |
| reedip | and would the loadbalancer port be used for an Instance boot ? | 21:39 |
| johnsom | It might in a failover scenario, but it fails just attaching the port to an instance.... | 21:40 |
| reedip | johnsom : but in that case, if we do not assign a DNS Name, wont it still fail ? | 21:41 |
| reedip | I mean the port doesnt have a DNS Name and the COmpute instance checks for it | 21:41 |
| johnsom | No, it does not fail. No DNS name is normal | 21:42 |
| reedip | hmm | 21:42 |
| johnsom | It's only when you populate that dns name that it becomes fragile and restricts what you can do with a port. | 21:42 |
| johnsom | Personally I think it is implemented incorrectly. | 21:42 |
| *** gcheresh_ has quit IRC | 21:42 | |
| johnsom | I don't think the dns_name should have anything to do with the nova instance the port is attached to.. I should be able to call it "lbaas-is-great" and be able to plug it into any instance.... | 21:43 |
| johnsom | IMO | 21:43 |
| reedip | johnsom : that you may be right... but they would have some reason for this restriction | 21:44 |
| reedip | lemme ask the designate team member as well | 21:46 |
| johnsom | Tell them Hi.... | 21:48 |
| johnsom | (I have mentioned this issue to them at the Barcelona summit) | 21:48 |
| reedip | :) | 21:50 |
| reedip | johnsom : okay, we will pick it up later ... :) | 21:50 |
| reedip | thanks though for the assist | 21:50 |
| reedip | next would be the timeout patch, I will ping you tomorrow for that :) | 21:51 |
| johnsom | Fair enough | 21:53 |
| reedip | ok, going back to sleep now ! | 21:53 |
| *** catintheroof has quit IRC | 21:57 | |
| *** catintheroof has joined #openstack-lbaas | 21:58 | |
| *** catintheroof has quit IRC | 21:58 | |
| openstackgerrit | Reedip proposed openstack/octavia master: Introduce Octavia v2 API for Health Monitor https://review.openstack.org/406256 | 22:14 |
| *** chlong has quit IRC | 22:27 | |
| *** eezhova has quit IRC | 22:27 | |
| openstackgerrit | Shashank Kumar Shankar proposed openstack/octavia master: Introduce Octavia v2 API for pools https://review.openstack.org/405922 | 22:35 |
| *** aojea has quit IRC | 22:53 | |
| *** harlowja has quit IRC | 23:06 | |
| openstackgerrit | Sindhu Devale proposed openstack/octavia master: Octavia v2 API for listeners https://review.openstack.org/424744 | 23:13 |
| openstackgerrit | Ankur proposed openstack/octavia master: Change v2 API LB create return status code https://review.openstack.org/448307 | 23:28 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!