Friday, 2017-01-13

rm_worknot sure if it's related to quotas00:00
rm_workdon't have another devstack to test on yet00:00
*** chlong has quit IRC00:01
rm_workoh and FYI00:01
rm_work2017-01-12 23:53:36.558 | +lib/neutron_plugins/services/l3:_neutron_configure_router_v6:416  sudo ip -6 addr replace 2001:db8::2/64 dev br-ex00:01
rm_work2017-01-12 23:53:36.579 | RTNETLINK answers: Permission denied00:01
rm_workthat's where it dies with IPv6 disabled00:01
johnsomWell, I did touch that code, but the exception handling for that case is still there.  I'm going to try it00:01
johnsomHmm, interesting.  That must be new-ish00:01
rm_workasking in qa00:03
rm_workah00:04
johnsomWell, neutron-lbaas handles it ok:00:04
rm_workhttp://docs.openstack.org/developer/devstack/configuration.html#ip-version00:04
rm_workI think that00:04
johnsomneutron lbaas-member-create --subnet tenant-2-subnet --address 172.21.2.12 --protocol-port 80 pool100:04
johnsomneutron CLI is deprecated and will be removed in the future. Use openstack CLI instead.00:04
johnsomMember with address 172.21.2.12 and protocol_port 80 already present in pool e6bae357-aacb-46a5-bcb1-08dcabc0846400:04
johnsomNeutron server returns request_ids: ['req-c6b97221-e700-4e6a-b6ee-88e6977a1c0b']00:04
rm_workjohnsom: oh, i'm going through octavia00:04
johnsomTrying with native octavia next00:04
rm_workcurl -X POST -H Content-type:application/json -H "X-Auth-Token:$MYTOKEN" -d '{"ip_address": "216.58.192.174", "protocol_port": 80, "project_id": "'$MYPROJECT'"}' http://localhost:9876/v1/loadbalancers/${LB_ID}/listeners/${LISTENER1_ID}/pools/${POOL1_ID}/members | python -m json.tool00:05
johnsomYeah, I can repo.  Let's see what I can do with this00:07
rm_workI wonder why it works through n-lbaas? >_>00:09
rm_workit should be issuing the same command00:09
johnsomIt's querying it's own DB00:10
johnsomDifferent API code...00:10
rm_workyeah but then it hits octavia with the same request, no?00:10
johnsomProbably not, I bet there is a validator in the nlbaas API00:10
rm_workwhat would it validate?00:11
rm_workthat the member isn't duplicate?00:11
johnsomI am guessing here...00:11
rm_workhmm00:11
rm_worki mean, the neutron-lbaas octavia driver just sends the same REST requests that I am manually, AFAIK00:12
rm_workthe member create call isn't complex, it's literally IP and PORT00:12
rm_workI mean, you HAVE to be right at some level, because obviously the result is different00:12
rm_workdisagreeing with that isn't really an option :/00:12
*** fnaval has quit IRC00:13
rm_workbut it doesn't make a lot of sense to me00:13
johnsomI think it just gets caught before the call is made to octavia.00:13
rm_workso, it lies?00:13
rm_work"yeah, created your member fine!" *didn't do anything*00:13
johnsomSo, yeah, it's walking off the end of that if elif for duplicates.  Existing code, probably never run.00:14
rm_workwhich would also be a bug in n-lbaas IMO00:14
rm_workit should return an error code00:14
johnsomThe duplicate column is "uq_member_pool_id_address_protocol_port" evidently00:14
rm_workright00:15
rm_workmakes sense00:15
rm_workso, does having the same member twice make sense?00:15
rm_workit DOESN'T, right?00:15
johnsomIt does not make sense00:15
rm_workright, you can just double the weight if you want to have it "twice"00:16
johnsomI'm just going to simplify this exception handler00:16
rm_workit can be a different CR00:16
johnsomRight00:16
rm_workquotas looks good from my testing otherwise00:16
*** cody-somerville has joined #openstack-lbaas00:17
rm_workI haven't gone through the code line-by-line yet tho :/ ugh00:17
rm_workI need to grab breakfast before I can deal with that00:17
johnsomGrin.  Ok, I am going to post one with the fix for that.  Probably should check why the functional test for members didn't catch it, will do that now too.00:18
johnsomAhhh, it passes functional test (there is one) because sqllite handles that differently.00:27
johnsomJoy00:27
*** cody-somerville has quit IRC00:42
rm_workrofl00:42
rm_workwoo sqlite00:42
rm_workalright, diving into this00:42
rm_workumm, are these default quotas right?00:43
rm_work20 LBs 10 pools? lol00:43
rm_workthat's ... not useful00:43
rm_workor is pools PER LB?00:43
rm_work... I didn't actually test that00:43
johnsomPer project.00:43
johnsomSo, yeah, the defaults are bogus....00:43
rm_workOK so... you can have 20 LBs but only 10 of them can have pools? :P00:44
johnsomShould we just set them unlimited?00:44
johnsomShared pools man....00:44
johnsomgrin00:44
rm_workcan't share between LBs :P00:44
johnsomRemember, I inherited this patch00:44
rm_workright I know very well00:44
rm_workI inherited it too00:44
rm_workand managed to escape somehow :P00:44
rm_workif only pools were sharable across LBs...00:44
johnsomShould we just set them unlimited?00:44
johnsomThat is kind of how I lean00:45
rm_workby default? ... maybe00:45
rm_workor maybe it'd be safe to have LBs default to like 10 and everything ELSE be unlimited00:50
rm_workbut it's pretty arbitrary whatever number you use00:50
johnsomYeah, I'm going to take the backwards-compat clause and say unlimited for everyone....00:51
openstackgerritMichael Johnson proposed openstack/octavia: Add quota support to Octavia  https://review.openstack.org/36079400:51
johnsomOk, fixed those00:51
rm_workdamn, same error with IP_VERSION=400:52
openstackgerritMerged openstack/octavia: Updated from global requirements  https://review.openstack.org/41935101:01
*** kevo has quit IRC01:27
rm_workjohnsom: couple more questions01:36
rm_workjohnsom: on the quota review01:36
rm_work(posted)01:36
johnsomSure01:36
*** cody-somerville has joined #openstack-lbaas01:40
*** cody-somerville has quit IRC01:40
*** cody-somerville has joined #openstack-lbaas01:40
*** gongysh has joined #openstack-lbaas01:40
johnsomrm_work Can you take a quick look at my answers and give me feedback, then I will do another spin01:45
rm_workOK then it's just the constants :P01:50
rm_workthen I'm good for +201:51
johnsomI can change to docs over too if you think it's the right thing.01:52
rm_workI think it's the right thing, but01:52
rm_workI don't think it's your responsibility here :/01:52
johnsomOk, well, let's do it now...  Get-er-done01:53
rm_worklol alright, fine by me :P01:53
rm_workbut it's super tedious01:53
rm_worksince i think a blind find/replace is too much01:53
johnsomYeah.  The should the request examples still be single quotes?01:54
johnsomI think so01:54
rm_workerr01:56
rm_worklike what line01:56
rm_workah01:57
rm_workno01:57
rm_workrequests with single-quotes in JSON will fail01:57
*** yamamoto_ has joined #openstack-lbaas01:57
johnsomGreat, now this too01:58
johnsomhttp://logs.openstack.org/94/360794/64/check/gate-octavia-v1-dsvm-scenario-ubuntu-xenial-nv/c99268b/console.html#_2017-01-13_01_42_50_66432801:58
rm_worklol wut01:58
johnsomA quick google comes up with neutron having the same problem.  They solved it by turning debug off for the test.  The code just has a TODO in it....01:59
rm_workthat's AFTER the tests pass01:59
rm_workwhat test even failed01:59
johnsomIt's not a test, it's the subunit stuff01:59
rm_workare we TOO VERBOSE?01:59
rm_worksubunit can't parse all the crap we spit out?01:59
johnsomhttps://github.com/testing-cabal/subunit/blob/master/python/subunit/v2.py#L20402:00
johnsomI don't know.  There is voodoo in that stuff for me still02:00
rm_workumm, lol02:01
rm_workyeah huh02:01
johnsomLooking at the testr file it's parsing, The biggest blocks I see are the response body stuff02:02
johnsomYeah, the log output from "test_load_balancer_tree_minimal" is lengthy with all of the req/resp text02:04
*** bana_k has quit IRC02:07
*** ducttape_ has joined #openstack-lbaas02:19
openstackgerritJoe Mills proposed openstack/neutron-lbaas: api test: centralize health monitor protocol  https://review.openstack.org/41932402:27
*** cody-somerville has quit IRC02:37
rm_workalright, I'll be +2 whenever the next patchset goes up02:40
rm_workis the subunit fail on anything voting?02:41
rm_workI'd kinda be for ... fixing it in a followup CR02:41
johnsomCool, just a few more minutes.  Fixing some other inconsistencies.02:41
rm_worklooks like that was on the scenarios...02:41
johnsomThe listener TLS stuff in this doc is all messed up too...  That will be for another day02:41
rm_workah k02:41
rm_worksad :/02:41
johnsomYeah, it's in non-voting scenario tests02:42
rm_workI've had a reprieve from looking at TLS since we aren't supporting it here for MVP, but02:42
rm_workI need to dig into that again eventually and get the weirdness sorted02:42
johnsomWell, soon I'm going to take on our api ref (next week maybe) so I will have to straighten it out then02:43
openstackgerritMichael Johnson proposed openstack/octavia: Add quota support to Octavia  https://review.openstack.org/36079402:45
johnsomOk, there it is....02:45
johnsomI need to sign off for the evening.  Catch you tomorrow02:45
*** yuanying has quit IRC02:52
*** amotoki has quit IRC02:53
*** ducttape_ has quit IRC02:59
*** ducttape_ has joined #openstack-lbaas03:00
*** ducttape_ has quit IRC03:04
*** ducttape_ has joined #openstack-lbaas03:18
*** gongysh has quit IRC03:29
*** ducttape_ has quit IRC03:42
*** ducttape_ has joined #openstack-lbaas03:42
*** ducttape_ has quit IRC03:44
*** ducttape_ has joined #openstack-lbaas03:44
*** yuanying has joined #openstack-lbaas03:52
*** links has joined #openstack-lbaas03:54
*** ducttape_ has quit IRC04:08
*** ducttape_ has joined #openstack-lbaas04:09
*** ducttape_ has quit IRC04:11
*** ducttape_ has joined #openstack-lbaas04:12
*** armax has quit IRC04:17
*** ducttape_ has quit IRC04:18
*** amotoki has joined #openstack-lbaas04:46
*** beardedeagle has joined #openstack-lbaas05:18
*** beardedeagle has quit IRC05:24
*** amotoki_ has joined #openstack-lbaas05:27
*** amotoki_ has quit IRC05:28
*** gongysh has joined #openstack-lbaas05:30
*** sticker has quit IRC05:31
*** sferna1_ has joined #openstack-lbaas05:34
*** yamamoto_ has quit IRC05:37
*** anilvenkata has joined #openstack-lbaas06:00
*** yuanying has quit IRC06:06
*** yuanying has joined #openstack-lbaas06:07
*** yuanying has quit IRC06:11
*** yuanying_ has joined #openstack-lbaas06:11
*** yuanying_ has quit IRC06:12
*** yuanying has joined #openstack-lbaas06:12
*** yuanying has quit IRC06:17
*** yamamoto_ has joined #openstack-lbaas06:21
*** madhu_ak has joined #openstack-lbaas06:45
*** gcheresh_ has joined #openstack-lbaas06:48
*** gcheresh has joined #openstack-lbaas06:52
*** gcheresh_ has quit IRC06:55
*** gongysh has quit IRC06:57
*** anilvenkata has quit IRC06:59
*** madhu_ak has quit IRC07:06
*** tesseract has joined #openstack-lbaas07:11
*** gcheresh_ has joined #openstack-lbaas07:12
*** gcheresh has quit IRC07:13
*** yuanying has joined #openstack-lbaas07:19
*** sindhu has quit IRC07:28
*** ptoohill has quit IRC07:30
*** dlundquist has quit IRC07:30
*** mhayden has quit IRC07:30
*** dlundquist has joined #openstack-lbaas07:33
*** ptoohill has joined #openstack-lbaas07:34
*** sindhu has joined #openstack-lbaas07:34
*** mhayden has joined #openstack-lbaas07:36
*** AlexeyAbashkin has joined #openstack-lbaas07:39
*** gcheresh_ has quit IRC08:06
*** kevo has joined #openstack-lbaas08:16
*** eezhova has joined #openstack-lbaas08:18
*** kevo has quit IRC09:07
*** ri0 has joined #openstack-lbaas09:08
*** kobis has quit IRC09:15
*** amotoki has quit IRC09:18
*** ri0 has quit IRC09:19
openstackgerritJoe Mills proposed openstack/neutron-lbaas: scenario test: configure scenario test protocols  https://review.openstack.org/41820109:29
*** amotoki has joined #openstack-lbaas09:36
*** openstackgerrit has quit IRC10:18
*** ipsecguy has quit IRC10:52
*** sferna1_ has quit IRC11:37
*** ipsecguy has joined #openstack-lbaas11:51
*** yamamoto_ has quit IRC12:11
*** catintheroof has joined #openstack-lbaas12:30
*** catintheroof has quit IRC12:31
*** catintheroof has joined #openstack-lbaas12:32
*** gcheresh_ has joined #openstack-lbaas12:55
*** yamamoto has joined #openstack-lbaas13:04
*** yamamoto_ has joined #openstack-lbaas13:05
*** yamamoto has quit IRC13:09
*** amotoki has quit IRC13:19
*** ducttape_ has joined #openstack-lbaas13:31
*** yamamoto_ has quit IRC13:37
*** AlexeyAbashkin has quit IRC13:37
*** yamamoto has joined #openstack-lbaas13:40
*** AlexeyAbashkin has joined #openstack-lbaas13:40
*** reedip_outofmemo has joined #openstack-lbaas13:43
*** sferna1_ has joined #openstack-lbaas13:49
*** gcheresh_ has quit IRC13:49
*** ducttape_ has quit IRC13:55
*** beardedeagle has joined #openstack-lbaas14:11
*** yamamoto has quit IRC14:35
*** chlong has joined #openstack-lbaas14:45
*** amotoki has joined #openstack-lbaas14:49
*** yamamoto has joined #openstack-lbaas14:50
*** yamamoto has quit IRC14:54
*** beagles is now known as beagles_afk14:54
*** ducttape_ has joined #openstack-lbaas15:04
*** ducttape_ has quit IRC15:09
*** armax has joined #openstack-lbaas15:26
*** fnaval has joined #openstack-lbaas15:33
*** ducttape_ has joined #openstack-lbaas15:33
*** jsheeren has joined #openstack-lbaas15:34
*** jsheeren has left #openstack-lbaas15:37
*** kobis has joined #openstack-lbaas16:29
*** pcaruana has joined #openstack-lbaas16:36
*** chlong has quit IRC16:38
xgermano/16:40
johnsomMorning16:42
*** eezhova has quit IRC16:42
*** chlong has joined #openstack-lbaas16:43
*** chlong has quit IRC16:48
*** AlexeyAbashkin has quit IRC16:49
*** bana_k has joined #openstack-lbaas16:50
*** reedip_outofmemo has quit IRC16:51
rm_worko/16:59
rm_worknight16:59
*** chlong has joined #openstack-lbaas16:59
johnsomTurning in?17:06
*** chlong has quit IRC17:09
*** bana_k has quit IRC17:15
*** eezhova has joined #openstack-lbaas17:26
*** eezhova has quit IRC17:38
*** _ducttape_ has joined #openstack-lbaas17:41
*** ducttape_ has quit IRC17:44
*** kevo has joined #openstack-lbaas17:45
*** eezhova has joined #openstack-lbaas17:49
*** openstackgerrit has joined #openstack-lbaas18:11
openstackgerritMerged openstack/octavia: Add quota support to Octavia  https://review.openstack.org/36079418:11
*** cody-somerville has joined #openstack-lbaas18:11
*** links has quit IRC18:12
*** bana_k has joined #openstack-lbaas18:15
*** beardedeagle has quit IRC18:24
*** _ducttape_ has quit IRC18:25
*** pcaruana has quit IRC18:26
*** ducttape_ has joined #openstack-lbaas18:26
*** pcaruana has joined #openstack-lbaas18:39
*** tesseract has quit IRC18:53
openstackgerritMichael Johnson proposed openstack/octavia: Fix octavia multinode setup for flavor and image  https://review.openstack.org/42013318:55
*** openstack has joined #openstack-lbaas19:02
*** pcaruana has quit IRC19:02
*** eezhova has quit IRC19:05
*** chlong has joined #openstack-lbaas19:14
*** eezhova has joined #openstack-lbaas19:30
*** ankur-gupta-f1 has left #openstack-lbaas19:34
openstackgerritMichael Johnson proposed openstack/octavia: Fix octavia multinode setup for flavor and image  https://review.openstack.org/42013319:43
*** openstack has joined #openstack-lbaas20:01
*** ankur-gupta-f2 has joined #openstack-lbaas20:01
*** beardedeagle has joined #openstack-lbaas20:10
*** openstackstatus has quit IRC20:27
*** openstack has joined #openstack-lbaas20:31
openstackgerritShashank Kumar Shankar proposed openstack/octavia: [Super WIP] Introduce Octavia v2 API for pools  https://review.openstack.org/40592220:44
*** cody-somerville has quit IRC20:58
*** cody-somerville has joined #openstack-lbaas21:54
*** cody-somerville has quit IRC21:54
*** cody-somerville has joined #openstack-lbaas21:54
*** yamamoto has joined #openstack-lbaas21:56
*** diltram has quit IRC21:56
openstackgerritSindhu Devale proposed openstack/octavia: Introduce Octavia v2 API for listeners  https://review.openstack.org/40506022:04
openstackgerritShashank Kumar Shankar proposed openstack/octavia: Introduce Octavia v2 API for pools  https://review.openstack.org/40592222:05
openstackgerritSindhu Devale proposed openstack/octavia: Align Octavia API to n-lbaasv2 for L7Policy  https://review.openstack.org/40632822:08
*** cody-somerville has quit IRC22:09
openstackgerritSindhu Devale proposed openstack/octavia: Align Octavia API to n-lbaasv2 for L7Rules  https://review.openstack.org/40633622:10
*** yamamoto has quit IRC22:11
*** yamamoto has joined #openstack-lbaas22:13
openstackgerritShashank Kumar Shankar proposed openstack/octavia: [WIP] Introduce Octavia v2 API for Pools with Type, Model change  https://review.openstack.org/42019522:14
rm_workjohnsom: yep, but i'm back!22:14
rm_workman, it's lively in here today22:14
rm_work<_<22:14
johnsomYeah, we have rebooted the API work, so there should be lots of activity there....22:14
*** yamamoto has quit IRC22:15
johnsomrm_work When you have a minute: https://review.openstack.org/420133   Nova puked on one of the tests, but the multinode passed in the first go round22:15
johnsomWe were getting multinode failures because the flavor wasn't in the secondary node conf22:16
rm_workah22:16
rm_workkk22:16
rm_worki was about to go through and help just fix merge conflicts on the large number of patches that fell over due to the quota merge22:17
*** yamamoto has joined #openstack-lbaas22:20
*** yamamoto has quit IRC22:22
*** yamamoto has joined #openstack-lbaas22:24
rm_workjohnsom: yeah i see. this seems right22:25
rm_workmind if I +2 with a recheck?22:25
johnsomI already hit recheck, so just a +222:25
rm_workah k22:25
rm_workin that case i'll just wait :P22:25
johnsomOk, fair enough.  ~30 minutes22:26
*** yamamoto has quit IRC22:26
*** ducttape_ has quit IRC22:27
*** catintheroof has quit IRC22:38
*** ducttape_ has joined #openstack-lbaas22:59
*** beardedeagle has quit IRC23:09
*** fnaval has quit IRC23:18
*** chlong has quit IRC23:38
*** eezhova has quit IRC23:39
*** harlowja has quit IRC23:56

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!