*** catintheroof has quit IRC | 00:19 | |
*** amotoki has joined #openstack-lbaas | 00:20 | |
*** JudeC has quit IRC | 00:49 | |
openstackgerrit | Erik Olof Gunnar Andersson proposed openstack/neutron-lbaas master: This is an initial attempt of improving the performance of regular api requests; such as lbaas-loadbalancer-list. https://review.openstack.org/455484 | 01:10 |
---|---|---|
*** sanfern has quit IRC | 01:18 | |
*** outofmemory is now known as reedip_ | 01:19 | |
eandersson | rm_work ^ | 01:29 |
eandersson | Super early stuff, but wanted to get something up there to get the full CI suit running. | 01:32 |
*** sanfern has joined #openstack-lbaas | 01:58 | |
*** catintheroof has joined #openstack-lbaas | 02:16 | |
*** fnaval has joined #openstack-lbaas | 02:21 | |
openstackgerrit | Erik Olof Gunnar Andersson proposed openstack/neutron-lbaas master: Improving performance of lbaas api calls https://review.openstack.org/455484 | 02:32 |
*** catintheroof has quit IRC | 02:34 | |
*** catintheroof has joined #openstack-lbaas | 02:35 | |
*** bbzhao has quit IRC | 02:37 | |
*** bbzhao has joined #openstack-lbaas | 02:37 | |
*** catintheroof has quit IRC | 02:40 | |
*** reedip_ has quit IRC | 02:42 | |
*** bzhao has quit IRC | 02:42 | |
*** bzhao has joined #openstack-lbaas | 02:43 | |
*** armax has joined #openstack-lbaas | 03:00 | |
*** fnaval_ has joined #openstack-lbaas | 03:15 | |
*** fnaval has quit IRC | 03:17 | |
*** cody-somerville has joined #openstack-lbaas | 03:24 | |
*** amotoki has quit IRC | 03:26 | |
*** csomerville has joined #openstack-lbaas | 03:26 | |
*** cody-somerville has quit IRC | 03:29 | |
*** amotoki has joined #openstack-lbaas | 03:40 | |
*** fnaval has joined #openstack-lbaas | 03:45 | |
*** chlong has quit IRC | 03:47 | |
*** fnaval_ has quit IRC | 03:48 | |
*** amotoki has quit IRC | 03:50 | |
*** amotoki has joined #openstack-lbaas | 03:58 | |
*** yamamoto_ has quit IRC | 04:09 | |
*** yamamoto has joined #openstack-lbaas | 04:09 | |
openstackgerrit | Merged openstack/octavia master: Loadbalancer v2 needs to accept tenant_id on POST https://review.openstack.org/455442 | 04:12 |
*** amotoki has quit IRC | 04:13 | |
*** amotoki has joined #openstack-lbaas | 04:29 | |
*** rcernin has joined #openstack-lbaas | 05:15 | |
*** gcheresh has joined #openstack-lbaas | 06:09 | |
*** aojea has joined #openstack-lbaas | 06:12 | |
*** voelzmo has joined #openstack-lbaas | 06:26 | |
*** gcheresh has quit IRC | 06:31 | |
*** JudeC has joined #openstack-lbaas | 06:42 | |
*** tesseract has joined #openstack-lbaas | 06:47 | |
openstackgerrit | Merged openstack/octavia master: Optimize the link address https://review.openstack.org/454427 | 06:59 |
*** gcheresh has joined #openstack-lbaas | 07:04 | |
*** pcaruana has joined #openstack-lbaas | 07:06 | |
*** gcheresh has quit IRC | 07:09 | |
openstackgerrit | Merged openstack/octavia master: Adds v2 load balancer API section https://review.openstack.org/438757 | 07:16 |
openstackgerrit | Merged openstack/octavia master: Explicitly list neutron api extensions for tests https://review.openstack.org/454769 | 07:16 |
openstackgerrit | Merged openstack/octavia master: Add Apache License Content in index.rst https://review.openstack.org/455195 | 07:16 |
*** aojea has quit IRC | 07:16 | |
*** aojea has joined #openstack-lbaas | 07:19 | |
openstackgerrit | Adam Harwell proposed openstack/octavia master: Removing dependency on eventlet and oslo.service https://review.openstack.org/454873 | 07:27 |
openstackgerrit | Adam Harwell proposed openstack/octavia master: Install DIB from source so depends-on will work https://review.openstack.org/444707 | 07:29 |
openstackgerrit | Adam Harwell proposed openstack/octavia master: Introduce Octavia v2 API for Health Monitor https://review.openstack.org/406256 | 07:30 |
openstackgerrit | Adam Harwell proposed openstack/octavia master: Implement sorting and pagination for octavia https://review.openstack.org/382147 | 07:37 |
*** kong has quit IRC | 07:39 | |
*** JudeC has quit IRC | 07:48 | |
*** ianychoi has quit IRC | 07:54 | |
*** ianychoi has joined #openstack-lbaas | 07:54 | |
openstackgerrit | Ji Chengke proposed openstack/octavia master: Fix error 500 when using SOURCE_IP and APP_COOKIE https://review.openstack.org/455569 | 07:59 |
openstackgerrit | Ji Chengke proposed openstack/octavia master: Fix error 500 when using SOURCE_IP and APP_COOKIE https://review.openstack.org/455569 | 08:03 |
*** aojea has quit IRC | 08:04 | |
*** aojea has joined #openstack-lbaas | 08:06 | |
*** aojea has quit IRC | 08:07 | |
*** bcafarel has joined #openstack-lbaas | 08:10 | |
openstackgerrit | Adam Harwell proposed openstack/octavia master: Introduce Octavia v2 API for Health Monitor https://review.openstack.org/406256 | 08:14 |
openstackgerrit | Adam Harwell proposed openstack/octavia master: Implement sorting and pagination for octavia https://review.openstack.org/382147 | 08:14 |
*** ianychoi has quit IRC | 08:20 | |
*** ianychoi has joined #openstack-lbaas | 08:21 | |
*** aojea has joined #openstack-lbaas | 08:25 | |
*** aojea has quit IRC | 08:29 | |
*** ianychoi has quit IRC | 08:35 | |
*** ianychoi has joined #openstack-lbaas | 08:38 | |
*** derekjhyang has joined #openstack-lbaas | 08:43 | |
*** fnaval has quit IRC | 08:51 | |
*** fnaval has joined #openstack-lbaas | 08:57 | |
openstackgerrit | Merged openstack/octavia master: Failover logic doesn't handle anti-affinity https://review.openstack.org/452213 | 08:58 |
*** voelzmo has quit IRC | 09:00 | |
*** voelzmo has joined #openstack-lbaas | 09:02 | |
*** openstackgerrit has quit IRC | 09:03 | |
*** fnaval has quit IRC | 09:17 | |
*** fnaval_ has joined #openstack-lbaas | 09:19 | |
*** aojea has joined #openstack-lbaas | 09:26 | |
*** aojea has quit IRC | 09:32 | |
*** fnaval_ has quit IRC | 09:38 | |
*** fnaval has joined #openstack-lbaas | 10:20 | |
*** kong has joined #openstack-lbaas | 10:22 | |
*** fnaval has quit IRC | 10:27 | |
*** aojea has joined #openstack-lbaas | 10:28 | |
*** aojea has quit IRC | 10:33 | |
*** fnaval has joined #openstack-lbaas | 10:35 | |
*** ianychoi has quit IRC | 10:36 | |
*** ianychoi has joined #openstack-lbaas | 10:38 | |
*** fnaval has quit IRC | 10:47 | |
*** gcheresh has joined #openstack-lbaas | 11:05 | |
*** fnaval has joined #openstack-lbaas | 11:10 | |
*** fnaval has quit IRC | 11:20 | |
*** gcheresh has quit IRC | 11:25 | |
*** aojea has joined #openstack-lbaas | 11:29 | |
*** aojea has quit IRC | 11:33 | |
*** gcheresh has joined #openstack-lbaas | 11:48 | |
*** amotoki has quit IRC | 12:11 | |
*** vaishali has quit IRC | 12:12 | |
*** basilAB has quit IRC | 12:14 | |
*** basilAB has joined #openstack-lbaas | 12:18 | |
*** vaishali has joined #openstack-lbaas | 12:18 | |
*** fnaval has joined #openstack-lbaas | 12:21 | |
*** ianychoi has quit IRC | 12:21 | |
*** amotoki has joined #openstack-lbaas | 12:23 | |
*** ianychoi has joined #openstack-lbaas | 12:26 | |
*** fnaval_ has joined #openstack-lbaas | 12:27 | |
*** fnaval has quit IRC | 12:27 | |
*** aojea has joined #openstack-lbaas | 12:30 | |
*** aojea has quit IRC | 12:34 | |
*** amotoki has quit IRC | 12:39 | |
*** amotoki has joined #openstack-lbaas | 12:43 | |
*** amotoki has quit IRC | 12:49 | |
*** fnaval_ has quit IRC | 12:54 | |
*** fnaval has joined #openstack-lbaas | 13:03 | |
*** amotoki has joined #openstack-lbaas | 13:12 | |
*** fnaval has quit IRC | 13:14 | |
*** fnaval has joined #openstack-lbaas | 13:20 | |
*** aojea has joined #openstack-lbaas | 13:31 | |
*** csomerville has quit IRC | 13:34 | |
*** aojea has quit IRC | 13:36 | |
*** ianychoi has quit IRC | 14:05 | |
*** ianychoi has joined #openstack-lbaas | 14:06 | |
*** chlong has joined #openstack-lbaas | 14:22 | |
*** reedip_ has joined #openstack-lbaas | 14:24 | |
*** fnaval has quit IRC | 14:26 | |
-openstackstatus- NOTICE: latest base images have mistakenly put python3 in some places expecting python2 causing widespread failure of docs patches - fixes are underway | 14:26 | |
*** fnaval has joined #openstack-lbaas | 14:27 | |
*** aojea has joined #openstack-lbaas | 14:31 | |
*** aojea has quit IRC | 14:36 | |
*** armax has quit IRC | 14:40 | |
-openstackstatus- NOTICE: we have rolled back centos-7, fedora-25 and ubuntu-xenial images to the previous days release. Feel free to recheck your jobs now. | 14:47 | |
*** ianychoi has quit IRC | 14:48 | |
*** ianychoi has joined #openstack-lbaas | 14:52 | |
*** fnaval has quit IRC | 14:52 | |
*** bbbbzhao_ has joined #openstack-lbaas | 14:52 | |
*** reedip_ has quit IRC | 14:54 | |
*** reedip_ has joined #openstack-lbaas | 14:55 | |
*** fnaval has joined #openstack-lbaas | 15:05 | |
*** rcernin has quit IRC | 15:05 | |
*** armax has joined #openstack-lbaas | 15:06 | |
*** kong has quit IRC | 15:09 | |
*** cody-somerville has joined #openstack-lbaas | 15:28 | |
*** cody-somerville has quit IRC | 15:28 | |
*** cody-somerville has joined #openstack-lbaas | 15:28 | |
*** chlong_ has joined #openstack-lbaas | 15:29 | |
*** chlong has quit IRC | 15:32 | |
*** aojea has joined #openstack-lbaas | 15:32 | |
*** aojea has quit IRC | 15:37 | |
*** openstackgerrit has joined #openstack-lbaas | 15:39 | |
openstackgerrit | Michael Johnson proposed openstack/octavia master: Add v2 listener API section https://review.openstack.org/455755 | 15:39 |
johnsom | WIP, just checkpointing since the parent merged | 15:39 |
openstackgerrit | Michael Johnson proposed openstack/octavia master: Add v2 listener API section https://review.openstack.org/455755 | 15:39 |
reedip_ | \o/ | 15:40 |
johnsom | Hi reedip_ | 15:41 |
reedip_ | tuesday is the best day for Networking in Openstack | 15:41 |
johnsom | Another fine day for Octavia.... | 15:41 |
xgerman | is it? | 15:41 |
johnsom | I hope I can finish this listener section today, but sometimes it turns out to be more work that I had hoped.... | 15:42 |
xgerman | sometimes? | 15:43 |
johnsom | Yep, sometimes it goes pretty quickly. It just depends on how many unique details this API section has.... The insert_headers slowed me down yesterday. | 15:44 |
*** chlong has joined #openstack-lbaas | 15:44 | |
johnsom | Or this part in the old docs that made me go double check the neutron-lbaas code.... | 15:46 |
johnsom | The IP protocol. Valid value is icmp, tcp, udp, or null. No default. | 15:46 |
johnsom | Umm, since when do we load balance ICMP? | 15:46 |
johnsom | Just another problem in the old docs.... | 15:47 |
johnsom | I was especially curious to see how we load balance the null protocol | 15:47 |
xgerman | dev/null protocol? | 15:48 |
*** JudeC has joined #openstack-lbaas | 15:50 | |
reedip_ | No, I think johnsom is stating the None/null protocol :) | 15:51 |
*** voelzmo has quit IRC | 16:02 | |
*** voelzmo has joined #openstack-lbaas | 16:03 | |
*** armax has quit IRC | 16:06 | |
*** ianychoi has quit IRC | 16:07 | |
*** voelzmo has quit IRC | 16:07 | |
*** JudeC has quit IRC | 16:07 | |
*** ianychoi has joined #openstack-lbaas | 16:12 | |
*** csomerville has joined #openstack-lbaas | 16:24 | |
*** cody-somerville has quit IRC | 16:24 | |
*** pcaruana has quit IRC | 16:29 | |
*** aojea has joined #openstack-lbaas | 16:33 | |
*** aojea has quit IRC | 16:38 | |
*** armax has joined #openstack-lbaas | 16:58 | |
*** JudeC has joined #openstack-lbaas | 17:01 | |
*** amotoki has quit IRC | 17:04 | |
*** bbbbzhao_ has quit IRC | 17:08 | |
*** JudeC has quit IRC | 17:11 | |
*** tesseract has quit IRC | 17:26 | |
*** KeithMnemonic1 has joined #openstack-lbaas | 17:26 | |
*** aojea has joined #openstack-lbaas | 17:34 | |
*** aojea has quit IRC | 17:38 | |
*** gcheresh has quit IRC | 17:46 | |
*** armax has quit IRC | 17:47 | |
*** csomerville has quit IRC | 17:50 | |
*** armax has joined #openstack-lbaas | 17:50 | |
*** JudeC has joined #openstack-lbaas | 17:50 | |
*** JudeC has quit IRC | 17:59 | |
*** armax has quit IRC | 18:01 | |
*** JudeC has joined #openstack-lbaas | 18:11 | |
*** diltram_ is now known as diltram | 18:17 | |
*** aojea has joined #openstack-lbaas | 18:35 | |
eandersson | When looking at https://wiki.openstack.org/wiki/Neutron/LBaaS/API_2.0 | 18:37 |
eandersson | apparently listing listeners for each lb is part of the spec when listing load balancers | 18:38 |
eandersson | not sure why | 18:38 |
eandersson | is the listener supposed to be fully loaded, or is it supposed to only have the uuid? | 18:38 |
*** aojea has quit IRC | 18:39 | |
xgerman | johnsom spend some time documenting the load balancer API | 18:49 |
xgerman | unless you juse one of the more inclusive statis calls I would assume we only retunr the id | 18:50 |
eandersson | Currently by simply excluding listeners you improve api performance by 3x | 18:51 |
eandersson | With 160 LBs it takes 21s to list all of them, without listeners it takes less than 7s | 18:52 |
eandersson | Since everything is recursively built a lot of these objects take up a ton of computing | 18:54 |
xgerman | yeah, that lazy loading never works | 18:54 |
eandersson | The solution is to only build what ever was specified in spec. | 18:54 |
xgerman | yes | 18:55 |
eandersson | I put up a very basic commit, but will need to figure a better approach https://review.openstack.org/#/c/455484/ | 18:55 |
eandersson | Also, can't simply exclude listeners, since it's in the speec. | 18:55 |
xgerman | yep | 18:56 |
eandersson | Stats and pools should be safe to exclude though right? since they aren't listed in the api specs | 18:56 |
*** gcheresh has joined #openstack-lbaas | 19:00 | |
*** sanfern has quit IRC | 19:00 | |
*** sanfern has joined #openstack-lbaas | 19:00 | |
*** sanfern has quit IRC | 19:00 | |
johnsom | eandersson Here is a preview of the cleaned up API reference I am working on: http://docs-draft.openstack.org/55/455755/2/check/gate-octavia-api-ref/11275bb//api-ref/build/html/v2/index.html | 19:01 |
eandersson | looks good | 19:02 |
eandersson | Can you include an example pool and listener in the list loadbalancers? | 19:02 |
*** csomerville has joined #openstack-lbaas | 19:02 | |
eandersson | like done for the listeners example int he old documentation | 19:02 |
johnsom | I think they are already there: http://docs-draft.openstack.org/55/455755/2/check/gate-octavia-api-ref/11275bb//api-ref/build/html/v2/index.html?expanded=list-load-balancers-detail#response-example | 19:03 |
eandersson | ah yep | 19:03 |
*** armax has joined #openstack-lbaas | 19:22 | |
openstackgerrit | Adam Harwell proposed openstack/octavia master: WIP: Floating IP Network Driver (spans L3s) https://review.openstack.org/435612 | 19:30 |
*** ianychoi has quit IRC | 19:34 | |
*** ianychoi has joined #openstack-lbaas | 19:35 | |
*** KeithMnemonic1 has quit IRC | 19:47 | |
eandersson | If I replace setattr and getattr the performance penalty for recursive is almost gone | 19:54 |
rm_work | lol | 19:56 |
rm_work | are they breaking the lazy? | 19:57 |
eandersson | Possible, or maybe I am just missing something. | 20:03 |
rm_work | well | 20:04 |
rm_work | if you can replace them and it works... | 20:04 |
rm_work | sounds good to me | 20:04 |
rm_work | i assume you meant "replace them with a different way of making it function the same"? | 20:04 |
eandersson | At the moment I am down to 2s.... from 30s | 20:04 |
eandersson | http://paste.openstack.org/show/KdtsfSvIMno4oSfHO8hH/ | 20:06 |
eandersson | I wish the unit-tests would verify if the api payload is correct :p | 20:07 |
rm_work | can you paste a diff? | 20:09 |
eandersson | I'll push it to https://review.openstack.org/#/c/455484/ in a bit | 20:10 |
rm_work | Ok | 20:10 |
rm_work | I guess you mean this? instance.__dict__[attr_name] | 20:10 |
eandersson | Yea | 20:10 |
rm_work | that is interesting, i've never used that. it's the same functionally as getattr/setattr? | 20:10 |
eandersson | I was just testing to see what da heck was going on | 20:10 |
rm_work | there's another instance of getattr below you missed | 20:11 |
eandersson | http://stackoverflow.com/questions/9790991/why-is-getattr-so-much-slower-than-self-dict-get | 20:11 |
eandersson | I'll re-write the same experimental patch in pycharm :D that was just me hacking in vim | 20:12 |
reedip_ | o/ | 20:12 |
rm_work | ok | 20:13 |
rm_work | eandersson: i may have some comments but I think this is OK | 20:14 |
eandersson | Assuming I am not just a madman and this is not a real fix, we should probably backport this to Mitaka as well. | 20:19 |
openstackgerrit | Nir Magnezi proposed openstack/octavia master: Fix imports in v2 listener types https://review.openstack.org/454177 | 20:22 |
*** armax has quit IRC | 20:26 | |
*** armax has joined #openstack-lbaas | 20:27 | |
rm_work | nmagnezi: were you finishing up changes to the user_group thing soon? | 20:39 |
openstackgerrit | Erik Olof Gunnar Andersson proposed openstack/neutron-lbaas master: Improving performance of lbaas api calls https://review.openstack.org/455484 | 20:41 |
eandersson | Lets see how it does against all the lovely tests. | 20:42 |
rm_work | commented | 20:45 |
eandersson | ops | 20:45 |
eandersson | thanks | 20:45 |
openstackgerrit | Erik Olof Gunnar Andersson proposed openstack/neutron-lbaas master: Improving performance of lbaas api calls https://review.openstack.org/455484 | 20:52 |
*** armax has quit IRC | 20:56 | |
openstackgerrit | Erik Olof Gunnar Andersson proposed openstack/neutron-lbaas master: Improving performance of lbaas api calls https://review.openstack.org/455484 | 21:03 |
*** chlong has quit IRC | 21:09 | |
*** chlong_ has quit IRC | 21:09 | |
*** armax has joined #openstack-lbaas | 21:10 | |
eandersson | I really hope this is a viable fix | 21:13 |
*** catintheroof has joined #openstack-lbaas | 21:19 | |
*** gcheresh has quit IRC | 21:20 | |
*** catintheroof has quit IRC | 21:21 | |
*** catintheroof has joined #openstack-lbaas | 21:21 | |
*** kong has joined #openstack-lbaas | 21:30 | |
openstackgerrit | Michael Johnson proposed openstack/octavia master: Add v2 listener API section https://review.openstack.org/455755 | 21:41 |
johnsom | Ok, that should be the listeners section for review | 21:42 |
xgerman | nice | 21:42 |
*** Sihan has joined #openstack-lbaas | 22:18 | |
Sihan | Hi guys, why are having two lb panel in horizon(loadbalancersv2 and ngloadbalancersv2) together? i am using python-neutron-lbaas-dashboard_2.0.0-0ubuntu1~cloud0_all.deb package. | 22:19 |
Sihan | is there any reason we still keep the old one? | 22:20 |
johnsom | Sihan Hi, the one that should be used is _1481_project_ng_loadbalancersv2_panel.py | 22:22 |
Sihan | I know. | 22:22 |
Sihan | but why do we still keep the old one in the package? | 22:22 |
Sihan | Sorry i am new to lbaas | 22:22 |
johnsom | I am not 100% sure if we should keep the other one around or not. The team working on that aren't working on OpenStack anymore, so we are currently maintaining until we can do more development on it. | 22:23 |
johnsom | We also don't do the ubuntu packaging, so we don't have a lot of influence on that deb | 22:24 |
johnsom | It's on my (long) list of things to look at. | 22:24 |
Sihan | ok, thx John | 22:24 |
Sihan | btw, what 'ng' stand for ? | 22:25 |
Sihan | :) | 22:25 |
johnsom | I think it was "next generation" as it was built using angular | 22:25 |
rm_work | everything is always "next generation" :P | 22:26 |
johnsom | Yep | 22:26 |
rm_work | why would anyone work on something from the current generation | 22:26 |
rm_work | i mean logically that wouldn't make sense | 22:26 |
johnsom | Very innovative around here... Grin | 22:26 |
rm_work | it's already done if it's the current generation | 22:26 |
*** Sihan has quit IRC | 22:48 | |
openstackgerrit | Michael Johnson proposed openstack/octavia master: Add quotas to Octavia v2 API https://review.openstack.org/455849 | 23:19 |
johnsom | rm_work when you have time, keep me honest on those functional tests. I know you found that we were doing it wrong in some places, I don't want to perpetuate that. Also if you are ok with the 202 result code. I think it's a good idea since these are carved in stone it seems, but interested in your opinion. | 23:21 |
*** JudeC has quit IRC | 23:23 | |
*** JudeC_ has joined #openstack-lbaas | 23:29 | |
*** catintheroof has quit IRC | 23:29 | |
*** catintheroof has joined #openstack-lbaas | 23:30 | |
*** catintheroof has quit IRC | 23:30 | |
*** bbzhao has quit IRC | 23:42 | |
*** bbzhao has joined #openstack-lbaas | 23:42 | |
*** fnaval has quit IRC | 23:45 | |
rm_work | k | 23:48 |
rm_work | i'll look at it tomorrow (or later today) probably | 23:48 |
rm_work | stuck on some internal stuff ATM | 23:48 |
johnsom | Thanks. It's a pull over of the v1 version | 23:48 |
johnsom | NP | 23:49 |
*** fnaval has joined #openstack-lbaas | 23:51 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!