| *** sshank has quit IRC | 00:01 | |
| *** atoth has quit IRC | 00:16 | |
| rm_work | johnsom: crap | 00:36 |
|---|---|---|
| rm_work | johnsom: well, it works, but it doesn't solve the slowness issue | 00:36 |
| johnsom | Hahahahaha | 00:36 |
| rm_work | still 4s requests | 00:36 |
| johnsom | I put a debug in just to make sure it still isn't reloading, you can check for that | 00:37 |
| rm_work | yeah i see nothing about that | 00:38 |
| rm_work | well, SOMETIMES it seems to reload randomly | 00:38 |
| rm_work | but like | 00:38 |
| rm_work | 1/40 requests | 00:38 |
| rm_work | ugh why is this slow | 00:39 |
| rm_work | i don't actually have much experience with profiling | 00:39 |
| johnsom | I do | 00:40 |
| johnsom | Maybe I can do some local poking at it tomorrow. | 00:40 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Add v2 Quota API section https://review.openstack.org/478327 | 00:43 |
| johnsom | Wahooo! | 00:43 |
| johnsom | Ok, after those merge I will link us from the official OpenStack APIs page. | 00:43 |
| johnsom | So I'm going to disappear for the evening soon-ish. Since I will get a start before you tomorrow, paste any pointers you have to what actions you are seeing slow and any other conditions that might be impacting it (# of requests, etc.). | 00:46 |
| johnsom | Things like if it's GETS or PUTs, specific objects, etc. | 00:47 |
| johnsom | How many projects and how many LBs you have | 00:47 |
| johnsom | etc. | 00:47 |
| rm_work | ok | 00:48 |
| rm_work | basically i'm just doing a LB list | 00:48 |
| rm_work | and i'm doing it with 4 clints | 00:48 |
| rm_work | *clients | 00:48 |
| rm_work | in a while loop | 00:48 |
| rm_work | with 1 client i see ~1.5s responses | 00:48 |
| johnsom | That should be pretty fast, just really up to the DB... | 00:48 |
| rm_work | and it scales up to about 4s with 4 clients | 00:48 |
| rm_work | i haven't tested with more | 00:49 |
| johnsom | About how man LBs in the list? | 00:49 |
| rm_work | with the old pre-rbac code, i was seeing 0.3s even with 4 clients | 00:49 |
| rm_work | uhh | 00:49 |
| *** diltram has quit IRC | 00:49 | |
| johnsom | I have a script somewhere that I used to create like 2000 LBs in the DB, so I should be able to use that. | 00:49 |
| rm_work | 11 | 00:50 |
| rm_work | and i'm admin | 00:50 |
| johnsom | Ha, ok | 00:50 |
| rm_work | so that's... all of them | 00:50 |
| rm_work | seeing about 1s response time per request actually it seems | 00:50 |
| rm_work | plus some tiny overhead | 00:50 |
| rm_work | took it to 3 and i see 3s | 00:51 |
| rm_work | 2 is just around 2s | 00:51 |
| rm_work | basically it is probably running single-threaded so | 00:51 |
| rm_work | it's just a factor of them queueing up, but taking ~1.1s each | 00:51 |
| rm_work | i need to figure out how to make gunicorn run this | 00:52 |
| johnsom | Yeah, I was thinking tomorrow I might jump into that darn uwsgi thing | 00:52 |
| johnsom | You running behind apache now? | 00:52 |
| *** diltram has joined #openstack-lbaas | 00:52 | |
| rm_work | i'm running "octavia-api" now | 00:53 |
| rm_work | >_< | 00:53 |
| johnsom | Oh | 00:53 |
| rm_work | also: were you going to +2 https://review.openstack.org/#/c/476690/ ? | 00:53 |
| rm_work | xgerman_ still needs to as well, but you didn't do it yet | 00:53 |
| rm_work | i'm trying to figure out how gunicorn and pecan interact | 00:54 |
| rm_work | i thought wsgi was wsgi but | 00:54 |
| rm_work | it seems like that's not true | 00:54 |
| rm_work | even a little bit | 00:54 |
| rm_work | does it HAVE to be uwsgi BTW? because we already require gunicorn sooooo | 00:55 |
| rm_work | it'd make sense to just use that IMO | 00:55 |
| rm_work | but maybe they are picky? | 00:55 |
| johnsom | https://www.irccloud.com/pastebin/PQJhgN07/ | 00:56 |
| johnsom | They were kind enough in the re-write to specify uwsgi | 00:57 |
| rm_work | >_> | 00:57 |
| *** blogan has quit IRC | 00:57 | |
| johnsom | I'm just bitter and grumpy about the whole thing. It shouldn't have changed mid-cycle | 00:57 |
| rm_work | yes same | 00:57 |
| rm_work | well, as part of doing that, it'll mean gunicorn should "just work" too right? | 00:58 |
| johnsom | In theory | 00:58 |
| rm_work | right not gunicorn explodes | 00:58 |
| rm_work | *right now | 00:58 |
| johnsom | Yeah, some multiprocessing thing? | 00:58 |
| rm_work | err | 01:06 |
| rm_work | i think it's related to loading stuff | 01:06 |
| rm_work | one sec i'll repro | 01:06 |
| rm_work | oh maybe it works now | 01:08 |
| rm_work | i think the bug was policy related lol | 01:08 |
| rm_work | testing performance on gunicorn momentarily | 01:10 |
| rm_work | oh nm it's on access | 01:11 |
| rm_work | http://paste.openstack.org/show/JdV3RGM0qcdjephpxmGK/ | 01:12 |
| rm_work | johnsom: ^^ | 01:12 |
| rm_work | am I missing pieces of required gunicorn config? | 01:12 |
| rm_work | or probably I have to make a pecan config? | 01:17 |
| rm_work | and use gunicorn_pecan as a runner? | 01:17 |
| *** csomerville has quit IRC | 01:22 | |
| *** cody-somerville has joined #openstack-lbaas | 01:27 | |
| *** diltram has quit IRC | 01:31 | |
| *** diltram has joined #openstack-lbaas | 01:36 | |
| *** JudeC has quit IRC | 01:55 | |
| openstackgerrit | Merged openstack/octavia master: Allow operators to disable v1 or v2.0 api endpoints https://review.openstack.org/476690 | 02:14 |
| openstackgerrit | Merged openstack/octavia master: Allow operators to disallow creation of TLS Termination listeners https://review.openstack.org/476701 | 02:15 |
| *** cody-somerville has quit IRC | 02:16 | |
| *** cody-somerville has joined #openstack-lbaas | 02:22 | |
| *** cody-somerville has quit IRC | 02:22 | |
| *** cody-somerville has joined #openstack-lbaas | 02:22 | |
| *** raginbajin has quit IRC | 02:34 | |
| *** raginbajin has joined #openstack-lbaas | 02:40 | |
| *** csomerville has joined #openstack-lbaas | 02:53 | |
| *** cody-somerville has quit IRC | 02:58 | |
| openstackgerrit | tuninger proposed openstack/neutron-lbaas master: Add the code for updating the servers statistics based on haproxy stats data. Fixes Bug1699945. https://review.openstack.org/476848 | 03:19 |
| reedip | johnsom , rm_work : Hi, any reviews on https://review.openstack.org/458308 ? | 03:26 |
| rm_work | yeah i need to look at it | 03:27 |
| rm_work | but i don't know much about that | 03:27 |
| reedip | rm_work : Its about putting QoS policies on VIP Ports so that if we put bandwidth limits on the VIP ports, those values are honoured | 03:27 |
| reedip | bzhao : do you need to add anything to this ?? ^^ | 03:30 |
| openstackgerrit | Ji Chengke proposed openstack/octavia master: Change default vagrant box to xenial64(16.04) https://review.openstack.org/478360 | 03:36 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Don't reload the policy for every API request. https://review.openstack.org/478280 | 03:42 |
| rm_work | ^^ rebased | 03:42 |
| *** sanfern has joined #openstack-lbaas | 03:51 | |
| *** gans has joined #openstack-lbaas | 03:59 | |
| *** links has joined #openstack-lbaas | 04:00 | |
| *** armax has joined #openstack-lbaas | 04:00 | |
| *** rstarmer has joined #openstack-lbaas | 04:22 | |
| openstackgerrit | Santhosh Fernandes proposed openstack/octavia master: This patch will emit LB provisioning status to be sync with neutron db https://review.openstack.org/478385 | 05:11 |
| openstackgerrit | Santhosh Fernandes proposed openstack/octavia master: [WIP] This patch will emit LB provisioning status to be sync with neutron db https://review.openstack.org/478385 | 05:14 |
| *** yamamoto has joined #openstack-lbaas | 05:34 | |
| *** armax has quit IRC | 05:40 | |
| *** kobis has joined #openstack-lbaas | 05:54 | |
| *** rcernin has joined #openstack-lbaas | 06:02 | |
| openstackgerrit | Rajat Sharma proposed openstack/octavia master: Replace the usage of 'manager' with 'os_primary' https://review.openstack.org/478399 | 06:22 |
| *** dayou has quit IRC | 06:39 | |
| *** med_ has quit IRC | 06:53 | |
| *** yamamoto has quit IRC | 06:55 | |
| *** Guest39045 has joined #openstack-lbaas | 06:59 | |
| *** yamamoto has joined #openstack-lbaas | 07:04 | |
| *** tesseract has joined #openstack-lbaas | 07:08 | |
| openstackgerrit | Ji Chengke proposed openstack/octavia master: Fix nova secgroup-add-rule which is unavailable https://review.openstack.org/478408 | 07:21 |
| *** aojea has joined #openstack-lbaas | 07:23 | |
| *** gans has quit IRC | 07:38 | |
| *** csomerville has quit IRC | 07:57 | |
| openstackgerrit | tuninger proposed openstack/neutron-lbaas master: Add the code for updating the servers statistics based on haproxy stats data. Fixes Bug1699945. https://review.openstack.org/476848 | 08:01 |
| *** aojea has quit IRC | 08:12 | |
| *** aojea has joined #openstack-lbaas | 08:13 | |
| openstackgerrit | Santhosh Fernandes proposed openstack/octavia master: This patch will emit LB provisioning status to be sync with neutron db https://review.openstack.org/478385 | 08:28 |
| *** dayou has joined #openstack-lbaas | 08:35 | |
| *** diltram has quit IRC | 08:36 | |
| *** diltram has joined #openstack-lbaas | 08:43 | |
| *** rstarmer has quit IRC | 08:52 | |
| *** aojea has quit IRC | 09:12 | |
| *** aojea has joined #openstack-lbaas | 09:12 | |
| *** yamamoto has quit IRC | 09:13 | |
| *** yamamoto has joined #openstack-lbaas | 09:17 | |
| *** pcaruana has joined #openstack-lbaas | 09:28 | |
| *** aojea has quit IRC | 09:51 | |
| *** aojea has joined #openstack-lbaas | 09:55 | |
| *** rcernin has quit IRC | 09:55 | |
| *** aojea_ has joined #openstack-lbaas | 09:58 | |
| *** amotoki_away is now known as amotoki | 10:01 | |
| *** aojea has quit IRC | 10:01 | |
| *** rcernin has joined #openstack-lbaas | 10:03 | |
| *** gans has joined #openstack-lbaas | 10:06 | |
| *** yamamoto has quit IRC | 10:08 | |
| *** rcernin has quit IRC | 10:09 | |
| *** rcernin has joined #openstack-lbaas | 10:11 | |
| *** yamamoto has joined #openstack-lbaas | 10:13 | |
| *** yamamoto has quit IRC | 10:15 | |
| gans | what would be best way to measure latency of VIP? One way i thought of ,making a HEAD call to VIP:PORT | 10:18 |
| *** yamamoto has joined #openstack-lbaas | 10:25 | |
| *** yamamoto has quit IRC | 10:48 | |
| *** yamamoto has joined #openstack-lbaas | 10:49 | |
| *** sanfern has quit IRC | 10:53 | |
| openstackgerrit | Nir Magnezi proposed openstack/octavia master: Correct RBAC Not Authorized status code https://review.openstack.org/477045 | 10:56 |
| *** amit213 has quit IRC | 11:02 | |
| *** gans has quit IRC | 11:09 | |
| *** atoth has joined #openstack-lbaas | 11:11 | |
| *** sanfern has joined #openstack-lbaas | 11:41 | |
| *** pcaruana has quit IRC | 12:02 | |
| *** sanfern has quit IRC | 12:02 | |
| *** pcaruana has joined #openstack-lbaas | 12:02 | |
| *** sanfern has joined #openstack-lbaas | 12:03 | |
| *** chlong_ has quit IRC | 12:15 | |
| *** pcaruana has quit IRC | 12:16 | |
| *** pcaruana has joined #openstack-lbaas | 12:23 | |
| *** pcaruana has quit IRC | 12:23 | |
| *** pcaruana has joined #openstack-lbaas | 12:23 | |
| *** yamamoto has quit IRC | 12:27 | |
| *** pcaruana has quit IRC | 12:29 | |
| *** yamamoto has joined #openstack-lbaas | 12:30 | |
| *** yamamoto has quit IRC | 12:30 | |
| *** pcaruana has joined #openstack-lbaas | 12:37 | |
| *** pcaruana has quit IRC | 12:43 | |
| *** pcaruana has joined #openstack-lbaas | 12:44 | |
| *** yamamoto has joined #openstack-lbaas | 12:49 | |
| *** yamamoto has quit IRC | 12:52 | |
| *** pcaruana has quit IRC | 12:53 | |
| *** pcaruana has joined #openstack-lbaas | 13:06 | |
| *** yamamoto has joined #openstack-lbaas | 13:10 | |
| *** pcaruana has quit IRC | 13:13 | |
| *** pcaruana has joined #openstack-lbaas | 13:14 | |
| *** links has quit IRC | 13:16 | |
| *** leitan has joined #openstack-lbaas | 13:18 | |
| *** links has joined #openstack-lbaas | 13:33 | |
| *** chlong_ has joined #openstack-lbaas | 13:34 | |
| *** pcaruana has quit IRC | 13:40 | |
| *** pcaruana has joined #openstack-lbaas | 13:46 | |
| *** links has quit IRC | 13:56 | |
| *** aojea_ has quit IRC | 14:01 | |
| openstackgerrit | cheng proposed openstack/octavia master: Add allocate vip port when create loadbalancer in server side https://review.openstack.org/463289 | 14:05 |
| *** aojea has joined #openstack-lbaas | 14:05 | |
| *** aojea_ has joined #openstack-lbaas | 14:28 | |
| *** aojea has quit IRC | 14:30 | |
| *** jniesz has joined #openstack-lbaas | 14:48 | |
| *** kobis has quit IRC | 14:55 | |
| *** fnaval has joined #openstack-lbaas | 14:59 | |
| *** KeithMnemonic has joined #openstack-lbaas | 15:04 | |
| *** rcernin has quit IRC | 15:06 | |
| *** links has joined #openstack-lbaas | 15:07 | |
| *** dayou has quit IRC | 15:18 | |
| *** links has quit IRC | 15:36 | |
| *** aojea_ has quit IRC | 15:49 | |
| *** kobis has joined #openstack-lbaas | 16:00 | |
| xgerman_ | johnsom wonder if L7 create can have a 409 — I think we return that before it goes on queue | 16:14 |
| johnsom | xgerman_ Yes it can. I thought we fixed all of those. It's on the l7rule patch | 16:16 |
| johnsom | Line 112 | 16:16 |
| openstackgerrit | Merged openstack/neutron-lbaas-dashboard master: Imported Translations from Zanata https://review.openstack.org/475716 | 16:16 |
| openstackgerrit | Merged openstack/neutron-lbaas-dashboard master: Updated from global requirements https://review.openstack.org/477981 | 16:17 |
| xgerman_ | mmh, I read the rendered https://usercontent.irccloud-cdn.com/file/W3ZZvfq0/Screen%20Shot%202017-06-28%20at%2012.17.11%20PM.png | 16:17 |
| xgerman_ | nope | 16:17 |
| johnsom | That is the old doc | 16:17 |
| xgerman_ | only update has it | 16:17 |
| *** sshank has joined #openstack-lbaas | 16:18 | |
| johnsom | That is the old v1 document, not this one. | 16:18 |
| xgerman_ | well, I clicked on the doc job in the patch… | 16:18 |
| johnsom | It's the api-ref job | 16:18 |
| johnsom | http://docs-draft.openstack.org/17/478217/2/check/gate-octavia-api-ref/78119da//api-ref/build/html/v2/index.html | 16:18 |
| xgerman_ | aha | 16:18 |
| xgerman_ | that link to filtering in Quota does not work | 16:22 |
| johnsom | Does it jump to the top? Are you on Chrome? | 16:26 |
| *** kobis has quit IRC | 16:28 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Add v2 Quota API section https://review.openstack.org/478327 | 16:33 |
| *** armax has joined #openstack-lbaas | 16:34 | |
| *** sshank has quit IRC | 16:35 | |
| *** sshank has joined #openstack-lbaas | 16:35 | |
| openstackgerrit | Merged openstack/python-octaviaclient master: Add TLS-HELLO option for HealthMonitors https://review.openstack.org/476075 | 16:49 |
| *** atoth has quit IRC | 16:54 | |
| *** greghaynes has quit IRC | 16:59 | |
| *** armax has quit IRC | 17:03 | |
| *** greghaynes has joined #openstack-lbaas | 17:11 | |
| *** ipsecguy has quit IRC | 17:14 | |
| *** ipsecguy has joined #openstack-lbaas | 17:15 | |
| rm_work | johnsom: i wonder if we can edit the log format to include request timing? | 17:16 |
| rm_work | i wonder if that's being captured | 17:17 |
| johnsom | hmmm | 17:17 |
| *** aojea has joined #openstack-lbaas | 17:21 | |
| *** aojea has quit IRC | 17:25 | |
| rm_work | well anyway, one thing that will definitely help a ton is having some real deployment system | 17:27 |
| rm_work | so need to figure out why gunicorn and uwsgi don't ... work | 17:27 |
| rm_work | i'm ok with deploying with uwsgi | 17:28 |
| rm_work | so were you going to look at that today or should I? | 17:28 |
| rm_work | johnsom: ^^ | 17:28 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Make developer debugging easier https://review.openstack.org/478575 | 17:28 |
| openstackgerrit | Merged openstack/neutron-lbaas master: Updated from global requirements https://review.openstack.org/477980 | 17:28 |
| johnsom | rm_work I was considering looking at that today. | 17:29 |
| rm_work | kk that'd be good | 17:29 |
| rm_work | ah i need to poke at your patch, i guess i missed a refactor or two lol | 17:30 |
| rm_work | (rbac fix) | 17:30 |
| johnsom | Yeah, not sure why that was... | 17:30 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Don't reload the policy for every API request. https://review.openstack.org/478280 | 17:39 |
| rm_work | anywho, fixed | 17:39 |
| rm_work | probably i read "get_all" but they were the "odd" objects and the get_all global method didn't work so i said whatever | 17:40 |
| rm_work | and didn't realize that means they'd work fine with the single-action-auth method heh | 17:40 |
| *** cody-somerville has joined #openstack-lbaas | 17:41 | |
| *** cody-somerville has quit IRC | 17:41 | |
| *** cody-somerville has joined #openstack-lbaas | 17:41 | |
| *** csomerville has joined #openstack-lbaas | 17:42 | |
| *** cody-somerville has quit IRC | 17:46 | |
| *** links has joined #openstack-lbaas | 17:47 | |
| *** sshank has quit IRC | 17:49 | |
| *** sshank has joined #openstack-lbaas | 17:49 | |
| xgerman_ | some nice rfeading: http://programmingisterrible.com/post/162346490883/how-do-you-cut-a-monolith-in-half | 17:51 |
| xgerman_ | a lot of people are gone Monday so that might be my big OpenStack day ;-) | 17:52 |
| rm_work | johnsom: do we have a deprecated tag for the docs? | 17:55 |
| rm_work | might be useful | 17:55 |
| rm_work | but we can go through and do that later | 17:55 |
| johnsom | https://developer.openstack.org/api-ref/load-balancer/ | 17:56 |
| rm_work | ? | 18:06 |
| *** yamamoto has quit IRC | 18:11 | |
| johnsom | That splash screen lists v1 as deprecated | 18:21 |
| *** reedip_out_of_ar has joined #openstack-lbaas | 18:22 | |
| jniesz | I was thinking of scheduling follow up meeting for active/active L3 on Friday. Other option would be to have it after flavor meeting tomorrow or post holiday | 18:26 |
| jniesz | how do those times work for everybody? | 18:27 |
| johnsom | Friday morning I have a conflict | 18:29 |
| *** links has quit IRC | 18:30 | |
| johnsom | I kind of think flavor and act/act is too much content and needs different prep. So, my vote would be post holiday I guess | 18:30 |
| jniesz | ok, what about Thursday of next week July 6th? | 18:33 |
| jniesz | that gives a week between the two and a day to recover from holiday | 18:33 |
| *** aojea has joined #openstack-lbaas | 18:38 | |
| *** sshank has quit IRC | 18:42 | |
| *** sshank has joined #openstack-lbaas | 18:42 | |
| openstackgerrit | Merged openstack/octavia master: Add v2 L7 Rule API section https://review.openstack.org/478217 | 18:46 |
| johnsom | Works for me | 18:49 |
| jniesz | same time 8 AM PST? | 18:53 |
| *** reedip_out_of_ar has quit IRC | 19:01 | |
| johnsom | Sure | 19:06 |
| *** yamamoto has joined #openstack-lbaas | 19:12 | |
| *** tesseract has quit IRC | 19:13 | |
| *** yamamoto has quit IRC | 19:16 | |
| *** sshank has quit IRC | 19:25 | |
| *** ipsecguy has quit IRC | 19:48 | |
| nmagnezi | o/ | 19:50 |
| johnsom | o/ | 19:51 |
| *** ipsecguy has joined #openstack-lbaas | 19:53 | |
| nmagnezi | i'm still on PTO, but the git conflicts in your patch today bothered me :D | 19:53 |
| johnsom | Ha, thanks | 19:53 |
| nmagnezi | np :) | 19:54 |
| nmagnezi | johnsom, was there a mail about the weekly meeting time? i couldn't find any | 19:55 |
| nmagnezi | johnsom, i voted in the survey | 19:55 |
| johnsom | No, I think we are going to talk about it again today | 19:55 |
| nmagnezi | ok | 19:56 |
| johnsom | Based on the doodle you are the only one that can't do 10am, though I had hoped more people would have participated | 19:57 |
| nmagnezi | sorry to be the only -1 :-) | 19:57 |
| johnsom | Octavia meeting starting soon on #openstack-meeting-alt | 20:00 |
| *** cpuga has joined #openstack-lbaas | 20:05 | |
| *** sshank has joined #openstack-lbaas | 20:07 | |
| *** yamamoto has joined #openstack-lbaas | 20:14 | |
| *** yamamoto has quit IRC | 20:23 | |
| rm_work | johnsom: ah interesting | 20:24 |
| rm_work | johnsom: i meant for attributes, like project-id on a bunch of calls | 20:24 |
| johnsom | rm_work Totally out of context, help me catch up..... | 20:25 |
| rm_work | johnsom: so err, should we be listing v1 as deprecated in the API version return too? | 20:25 |
| rm_work | or fix the api-ref to not say it's deprecated | 20:25 |
| rm_work | [10:55:39] <rm_work>johnsom: do we have a deprecated tag for the docs? | 20:26 |
| rm_work | you linked the splash page that has v1 listed as deprecated | 20:26 |
| rm_work | but i meant on attributes | 20:26 |
| johnsom | Oh, got you now | 20:28 |
| johnsom | I should probably change the API ref screen | 20:28 |
| *** cpuga has quit IRC | 20:28 | |
| johnsom | I need to make a few changes for v1 anyway, I will do it in that patch | 20:29 |
| *** sshank has quit IRC | 20:47 | |
| *** bcafarel has joined #openstack-lbaas | 20:48 | |
| *** leitan has quit IRC | 20:48 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Make developer debugging easier https://review.openstack.org/478575 | 20:51 |
| *** leitan has joined #openstack-lbaas | 20:57 | |
| *** pcaruana has quit IRC | 21:01 | |
| *** leitan has quit IRC | 21:08 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Enable uwsgi support for octavia devstack https://review.openstack.org/478637 | 21:09 |
| johnsom | I will be shocked if that works on the first try, so WIP | 21:13 |
| nmagnezi | johnsom, we should have a gerrit hook to automatically set WIP state for changes that pass CI on ps#1 :D | 21:31 |
| *** diltram has quit IRC | 21:33 | |
| *** fnaval has quit IRC | 21:38 | |
| *** diltram has joined #openstack-lbaas | 21:38 | |
| *** fnaval has joined #openstack-lbaas | 21:38 | |
| rm_work | johnsom: does the PROXY protocol have an associated performance hit? | 21:38 |
| rm_work | wondering if there's a reason we wouldn't just recommend ALWAYS using that over HTTPS/TCP | 21:39 |
| *** sshank has joined #openstack-lbaas | 21:42 | |
| johnsom | It injects data into the start of the connection, so yeah, some hit, plus the backend has to be able to handle it. Normal servers will likely reject the connection | 21:43 |
| *** fnaval has quit IRC | 21:48 | |
| rm_work | ah hmm | 21:52 |
| rm_work | oh | 21:52 |
| johnsom | I don't think it is that widely used really | 21:53 |
| rm_work | need to figure out what is causing my healthmanagers to emit events to a queue that doesn't exist | 21:53 |
| johnsom | Most HTTP connections use x-forwared-for | 21:54 |
| rm_work | it's doing the neutron event emit thing | 21:54 |
| rm_work | but i don't think i have it configured to do that... | 21:54 |
| johnsom | Hmm, that should not happen without the streamer config setting | 21:54 |
| rm_work | 2017-06-28 14:54:44.333 27563 DEBUG octavia.controller.queue.event_queue [-] Emitting data to event streamer {'info_type': 'loadbalancer_stats', 'info_id': u'9063d9c2-e471-4e7f-a78d-747fc0dec57f', 'info_payload': {'bytes_out': 0, 'bytes_in': 0, 'total_connections': 0, 'active_connections': 0, 'request_errors': 0}} emit /opt/octavia/env/lib/python2.7/site-packages/octavia/controller/queue/event_queue.py:70 | 21:54 |
| rm_work | that means it's doing the queue emit right? | 21:55 |
| *** jniesz has quit IRC | 21:55 | |
| johnsom | I haven't really dug into the streamer stuff | 21:56 |
| rm_work | hmmm | 21:56 |
| johnsom | I mean, it should be tied to that config setting right? | 21:56 |
| rm_work | yeah but | 21:57 |
| rm_work | the default seems to be noop | 21:57 |
| rm_work | but it's doing it | 21:57 |
| rm_work | really weird | 21:57 |
| johnsom | Ugh, I hate that code | 21:57 |
| rm_work | yes | 21:58 |
| rm_work | me too | 21:58 |
| rm_work | <_< | 21:59 |
| johnsom | Yeah, that log message is in the actual streamer driver, so not running noop I guess | 21:59 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Enable uwsgi support for octavia devstack https://review.openstack.org/478637 | 22:07 |
| johnsom | They changed the API URL to not have ports so our lbaas default isn't default anymore. | 22:07 |
| johnsom | Ah well, setting it if uwsgi is enabled in devstack | 22:08 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Event Streamer for StatsDb should respect config value https://review.openstack.org/478665 | 22:08 |
| johnsom | Coolness that the octavia gates just worked | 22:08 |
| rm_work | yeah johnsom uwsgi is working well for me downstream | 22:09 |
| rm_work | we going to add it to our requirements? | 22:09 |
| rm_work | ah i should look at the patch | 22:09 |
| johnsom | That is the odd thing, the other projects aren't doing that.... | 22:09 |
| johnsom | Devstack is installing it, so they aren't adding it. | 22:09 |
| rm_work | >_> k | 22:10 |
| rm_work | i had to add it downstream in a patch | 22:10 |
| rm_work | so our image builder would pick it up | 22:10 |
| johnsom | I think they are actually building it to work around a distro issue | 22:10 |
| rm_work | k | 22:10 |
| rm_work | johnsom: so yeah https://review.openstack.org/#/c/478665/ and xgerman_ | 22:10 |
| johnsom | image builder? | 22:10 |
| rm_work | kolla | 22:10 |
| johnsom | I isn't in the amp | 22:10 |
| rm_work | we run the api/worker/etc in docker images | 22:10 |
| rm_work | in kubernetes | 22:10 |
| rm_work | which are built by kolla | 22:11 |
| johnsom | On the fence with adding it to requirements. I hate to deviate from the other projects. Plus, it's not the "default" mode of operation | 22:13 |
| *** chlong_ has quit IRC | 22:18 | |
| rm_work | yeah i don't care either way | 22:22 |
| rm_work | i'm already doing it | 22:22 |
| rm_work | but if it's just part of the devstack plugin | 22:22 |
| rm_work | and devstack already installs uwsgi | 22:22 |
| rm_work | .... | 22:22 |
| johnsom | Right | 22:22 |
| *** aojea has quit IRC | 22:29 | |
| *** ipsecguy has quit IRC | 22:36 | |
| rm_work | xgerman_ / johnsom https://review.openstack.org/#/c/478280/ | 22:42 |
| rm_work | probably need both of you :P | 22:42 |
| johnsom | rm_work https://review.openstack.org/#/c/478327/ | 22:48 |
| johnsom | Since you are pointing out patches.... grin | 22:48 |
| rm_work | ah yeah i just finished reviewing it | 22:49 |
| rm_work | but forgot to +A | 22:49 |
| rm_work | https://review.openstack.org/#/c/478665/ | 22:49 |
| *** ipsecguy has joined #openstack-lbaas | 22:50 | |
| johnsom | Yeah, I was just waiting on the gate runs | 22:50 |
| rm_work | k | 22:50 |
| johnsom | Argh, so something is screwy in neutron land. | 22:50 |
| johnsom | http://logs.openstack.org/37/478637/2/check/gate-neutron-lbaasv2-dsvm-api-ubuntu-xenial/75e4deb/logs/etc/neutron/ | 22:50 |
| johnsom | There are two lbaas config files now.... | 22:50 |
| rm_work | err | 22:51 |
| johnsom | And my iniset didn't work | 22:51 |
| *** ipsecguy has quit IRC | 22:56 | |
| johnsom | Joy, dead code, ok | 22:56 |
| rm_work | lol T_T | 22:57 |
| rm_work | yeah | 22:57 |
| rm_work | man, i am still wondering how anyone is running this thing | 22:57 |
| rm_work | because ... | 22:57 |
| rm_work | these bugs | 22:57 |
| rm_work | i guess if you're running behind n-lbaas most of them aren't noticable <_< | 22:58 |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Enable uwsgi support for octavia devstack https://review.openstack.org/478637 | 22:59 |
| johnsom | Right, stuff works until you poke it | 22:59 |
| *** ipsecguy has joined #openstack-lbaas | 23:00 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Update our docs to point to the new api-ref https://review.openstack.org/478673 | 23:08 |
| *** leitan has joined #openstack-lbaas | 23:09 | |
| *** KeithMnemonic1 has joined #openstack-lbaas | 23:14 | |
| *** leitan has quit IRC | 23:14 | |
| *** KeithMnemonic has quit IRC | 23:17 | |
| *** amotoki is now known as amotoki_away | 23:23 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Update our docs to point to the new api-ref https://review.openstack.org/478673 | 23:25 |
| *** dayou has joined #openstack-lbaas | 23:31 | |
| *** sshank has quit IRC | 23:32 | |
| johnsom | Ok, the v1 docs and API link update patch looks good: https://review.openstack.org/#/c/478673/ | 23:36 |
| johnsom | It is a bit odd as the pages point to the official links, so jumps you out of the gate version of the docs. | 23:36 |
| openstackgerrit | Merged openstack/octavia master: Make developer debugging easier https://review.openstack.org/478575 | 23:47 |
| openstackgerrit | Merged openstack/octavia master: Add v2 Quota API section https://review.openstack.org/478327 | 23:47 |
| rm_work | xgerman_: https://review.openstack.org/#/c/478665/ | 23:52 |
| rm_work | xgerman_: https://review.openstack.org/#/c/478280/ | 23:52 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!
