Friday, 2016-03-25

*** _ducttape_ has quit IRC00:01
*** fawadkhaliq has quit IRC00:07
*** fawadkhaliq has joined #openstack-lbaas00:07
*** neelashah has joined #openstack-lbaas00:15
*** ihrachys has quit IRC00:16
*** SumitNaiksatam has quit IRC00:26
*** ducttape_ has joined #openstack-lbaas00:28
bana_kis there a way to identify the amps state in ha mode? like is it master or slave00:35
bana_k?00:35
*** ducttape_ has quit IRC00:37
*** fawadkhaliq has quit IRC00:38
*** fawadkhaliq has joined #openstack-lbaas00:39
*** fawadkhaliq has quit IRC00:47
*** fawadkhaliq has joined #openstack-lbaas00:48
*** neelashah has quit IRC00:48
*** ducttape_ has joined #openstack-lbaas00:53
*** fawadkhaliq has quit IRC00:54
sbalukoffbana_k: I think that information is stored in the database.00:55
sbalukoffbana_k: So it should be part of the amp model in any case.00:55
bana_kyea but are we updating that info when master fails and back up becomes master ?00:56
openstackgerritAishwarya Thangappa proposed openstack/octavia: Fixes the SNI issues in master(mitaka) for octavia  https://review.openstack.org/29748001:03
*** bana_k has quit IRC01:21
*** ducttape_ has quit IRC01:24
*** kev0 has quit IRC01:25
openstackgerritMerged openstack/octavia: Adds documentation for the Octavia configuration  https://review.openstack.org/29433801:25
*** minwang2 has quit IRC01:26
*** piet has joined #openstack-lbaas01:26
*** ducttape_ has joined #openstack-lbaas01:27
*** fawadkhaliq has joined #openstack-lbaas01:29
*** ducttape_ has quit IRC01:30
*** ducttape_ has joined #openstack-lbaas01:33
johnsom_banak For human use, you can look in syslog and see the state.  For non-human use, there aren't great options, but two possibilities.01:33
johnsom_1. There is a signal that will cause it to dump a status file.  Not great01:34
johnsom_2. You could make a patch that adds a notify script that updates a state file somewhere.01:35
johnsom_opps I meant bana_k01:43
*** yamamoto_ has joined #openstack-lbaas01:45
*** piet has quit IRC01:45
*** jaff_cheng has joined #openstack-lbaas01:49
johnsom_FYI, for the flow documentation patch I setup a web server with the current output:01:51
johnsom_http://13.91.98.501:52
johnsom_I think there are a few things to adjust, but it looks pretty good.01:52
*** johnsom_ is now known as johnsom01:52
johnsomI did this since it won't build on octavia.io until the taskflow fix is released01:52
*** ducttape_ has quit IRC01:53
rm_workugh01:57
rm_workit IS or ISN'T octavia.io's fault tho?01:57
rm_workit's a bug in taskflow that makes it unable to build?01:57
*** yamamoto_ has quit IRC02:01
*** yamamoto_ has joined #openstack-lbaas02:02
*** woodster_ has quit IRC02:07
*** SumitNaiksatam has joined #openstack-lbaas02:18
*** piet has joined #openstack-lbaas02:26
*** yamamoto_ has quit IRC02:49
*** barra204 has joined #openstack-lbaas02:49
*** ducttape_ has joined #openstack-lbaas02:56
*** alhu_ has quit IRC02:58
*** Purandar has quit IRC03:00
*** fawadkhaliq has quit IRC03:03
*** chenghang has joined #openstack-lbaas03:05
*** jaff_cheng has quit IRC03:07
johnsomrm_work No worries, it was networkx changing out from under taskflow.  I have already put a taskflow patch up.03:10
*** chlong has quit IRC03:12
*** ducttape_ has quit IRC03:12
*** fawadkhaliq has joined #openstack-lbaas03:14
*** chlong has joined #openstack-lbaas03:14
*** neelashah has joined #openstack-lbaas03:25
*** neelashah has quit IRC03:27
*** Purandar has joined #openstack-lbaas03:29
*** jaff_cheng has joined #openstack-lbaas03:31
*** ajmiller has joined #openstack-lbaas03:31
*** chenghang has quit IRC03:33
rm_workjohnsom: I see, ok. I always really just cross my fingers that it keeps working so I don't have to touch it >_>03:34
*** yuanying_ has joined #openstack-lbaas03:35
*** yuanying has quit IRC03:37
*** yuanying has joined #openstack-lbaas03:38
*** yuanyin__ has joined #openstack-lbaas03:40
*** yuanying_ has quit IRC03:40
*** yuanying has quit IRC03:43
*** yamamoto has joined #openstack-lbaas03:44
*** ajmiller_ has joined #openstack-lbaas03:46
*** bana_k has joined #openstack-lbaas03:47
*** yuanyin__ has quit IRC03:49
*** ajmiller has quit IRC03:49
*** yuanying has joined #openstack-lbaas03:51
*** jaff_cheng has quit IRC03:58
*** yuanying has quit IRC03:59
*** links has joined #openstack-lbaas04:01
*** yuanying has joined #openstack-lbaas04:02
openstackgerritBanashankar k proposed openstack/octavia: Adding the topology info at the time of LB AMP association  https://review.openstack.org/28636504:05
openstackgerritStephen Balukoff proposed openstack/octavia: Adding introduction and developer quick start docs  https://review.openstack.org/23217304:05
*** yuanying has quit IRC04:07
*** ajmiller_ has quit IRC04:28
*** eranra has quit IRC04:47
*** bana_k has quit IRC04:47
*** eranra has joined #openstack-lbaas04:47
*** eranra has quit IRC05:01
*** eranra has joined #openstack-lbaas05:05
*** eranra has quit IRC05:06
*** eranra has joined #openstack-lbaas05:07
*** eranra has quit IRC05:08
*** eranra has joined #openstack-lbaas05:09
*** piet has quit IRC05:13
*** alhu has joined #openstack-lbaas05:16
*** alhu has quit IRC05:35
bharathmbana_k: I commented on your patch.. I tried testing it locally and using 'builtins' module resulting in py27 failures.05:38
bharathmrm_work: when you have some time, can you take a look at https://review.openstack.org/#/c/297480/05:40
bharathmI bet you have more experience working on tls/sni stuff in octavia05:41
rm_workmaybe :P05:42
rm_workI will take a look now tho05:42
rm_workthen I need to head to the store in a bit05:42
bharathmSure.. whenever you are free.. :-) thanks05:43
rm_workhmmm05:43
*** links has quit IRC05:43
rm_workI believe EITHER tls_container_id or tls_container.id should work05:43
rm_workas I think it is a model object there05:44
rm_work?05:44
rm_workor is it not05:44
rm_worka model would have tls_container_id as a raw value but then SQLAlchemy should link in tls_container as an object which would have .id05:44
rm_workI think05:44
bharathmusing tls_container.id errored out with "none type object tls_container" or something like that05:44
rm_workbut it has been so long maybe that is totally wrong and those are actually our container class05:44
rm_workhmm05:44
rm_workso what is wrong with this code? I see lots of red from jenkins, is it just unit tests failing that were actually bad originally? or is this actually still not working for you05:45
bharathmI looked at the model and thought the tls_container_id to be right one.. but I may be wrong..05:45
rm_workhmm05:45
rm_workwell the issue is that this is super confusing05:46
rm_workbecause we pass the models around like half the time and cert_manager certificate container classes the other hald05:46
rm_work*other half05:46
rm_workand they are not compatible05:46
bharathmCode is working fine.. Just the tests failing.. But wanted to ask you if we are missing anything05:46
rm_workah ok, so if you've verified that it's working, then i will just look to see if I can think of anything else05:47
rm_workand not worry about whether the arg names are right :P05:47
bharathmCool..05:47
rm_workI find it annoying that we named the fields "listener.tls_certificate_id" but "listener.sni_containers.tls_container_id"05:48
rm_workshould have tried to be more consistent :(05:48
rm_work"listener.sni_certificates.tls_certificate_id"05:48
rm_workor "listener.tls_container_id"05:49
rm_workbut I like the first better05:49
rm_worktoo late now :( just causes confusion05:49
bharathmYeah bugs me too.. I thought that was intentional05:49
rm_workdifferent people worked on different pieces05:49
rm_workI did the base TLS stuff but not SNI05:50
bharathmIndeed.. I would say let it slide for now as it works.. We can revisit and refactor in the future if tls gets more complicated (which I dont think it would)05:50
rm_workI think ORIGINALLY the base TLS stuff *was* "tls_container_id" but we wanted to make it not specifically "container" related05:50
rm_workso it got changed, but SNI didn't05:50
rm_workbecause possibly SNI was mid-development based on the original naming05:51
rm_workugh05:51
rm_workanyway, this looks fine to me I think, though it has been so long I am not sure if there could be something missing still :/05:51
rm_workonce we have scenario tests for this it will be easy peasy05:51
bharathmWas looking at datamodels but I couldn't figure out where it would set tls_container object rather than tls_container_id..05:51
bharathmBut the testcase sample config, it was configured as listener.sni_containers.tls_container[].*05:52
bharathm+1 on scenario tests05:52
*** Purandar has quit IRC05:54
rm_workbharathm: so you just need to fix the tests for this then?05:54
bharathmerr I meant listener.sni_containers[].tls_container.*05:54
rm_workwhy is this moving to a 128 char string?05:54
rm_workthe container ids should be uuids, so05:55
rm_work3605:55
rm_workI wonder why neutronl-baas has 128?05:55
bharathmThe container id from barbican is a url05:55
rm_workoh shit right05:55
rm_workLOL this was the old code when we weren't planning to store that05:55
rm_workand then decided we had to05:55
rm_workyeah this SNI code is based on OLD design05:55
bharathmso in neutron lbaas db its 128 chars however errors out in our octavia api because of column length issues05:55
rm_workI am amazed this is still here, there's no way SNI would work like this05:55
rm_worknot even with JUST octavia05:56
rm_workthe absolute *minimum* length of a container id (really, a container_ref) is like... 4805:56
rm_workhttp://a.co/UUID05:56
rm_workwhich would never even happen05:57
bharathmDo we have a new design ?05:57
bharathmfor handling SNI05:57
rm_worki mean05:57
rm_workthe *current* TLS design05:57
*** fnaval has joined #openstack-lbaas05:58
rm_workbasically SNI is a 1:N clone of the 1:1 TLS stuff05:58
rm_workbut it was based on the *old* TLS design05:58
rm_workwhich we fixed but ignored SNI apparently05:58
rm_workit should essentially be the TLS models/data lifted and put into its own table and linked in as 1:N on a listener05:58
rm_workplus a couple of extra columns05:59
rm_workerr... maybe not even05:59
rm_worksooo, I am a little concerned though about just blindly removing it from the jinja builder05:59
*** cody-somerville has quit IRC06:00
rm_workunless the data from the sni_certs is like... appanded to the tls_cert data that is still passed, then the haproxy config never gets the sni data?06:00
*** cody-somerville has joined #openstack-lbaas06:00
*** cody-somerville has quit IRC06:00
*** cody-somerville has joined #openstack-lbaas06:00
fnavalreview requested pls: https://review.openstack.org/#/c/164828/06:00
rm_worksup fnaval !06:00
fnavalhey rm_work!06:00
rm_workfnaval: are we nearly the same timezone?06:00
fnavalyep06:00
fnavalwell i think off by an hour06:00
rm_workheh06:00
rm_workclose enough06:01
bharathmAhaa.. gotcha.. Well current jinga build doesnt read the SNI from the arg.. Rather it reads directly from the listener model06:01
rm_workbharathm: oh what? errr06:01
rm_workbharathm: then why are we even passing the certs data06:01
rm_workwouldn't it just be able to read that directly as well?06:01
bharathmOops sorry I spoke too soon..06:04
bharathmYou are right.. the correct way is to pass the certs[sni] as arg to build and use it.. I am curious how this worked though.. May be false positive..06:07
bharathmWill try this tomorrow06:07
*** eranra has quit IRC06:08
*** eranra has joined #openstack-lbaas06:09
*** amotoki has joined #openstack-lbaas06:11
*** eranra has quit IRC06:14
openstackgerritMichael Johnson proposed openstack/octavia: Adds a process to generate key flow documentation  https://review.openstack.org/29742306:14
*** eranra has joined #openstack-lbaas06:14
rm_workbharathm: ok yeah, I'd be curious how the model code was handling it -- because it SEEMS like you are right that it could get it from the listener, but... it SHOULD be consistent, and the way you have it right now is not06:15
rm_workposted for posterity on the review06:16
bharathmAgreed.. thanks for the inputs06:17
*** johnsom_ has joined #openstack-lbaas06:33
*** johnsom has quit IRC06:37
*** chlong has quit IRC06:45
*** yuanying has joined #openstack-lbaas06:46
*** fnaval has quit IRC07:13
*** fawadkhaliq has quit IRC07:17
*** kevo has joined #openstack-lbaas07:20
*** bharathm has quit IRC07:38
*** pcaruana has joined #openstack-lbaas07:38
*** bharathm has joined #openstack-lbaas07:42
*** kevo has quit IRC07:45
*** cody-somerville has quit IRC07:55
*** cody-somerville has joined #openstack-lbaas08:10
*** reedip is now known as reedip_away08:13
*** prabampm has joined #openstack-lbaas08:29
*** cody-somerville_ has joined #openstack-lbaas09:38
*** cody-somerville has quit IRC09:39
*** yamamoto has quit IRC10:03
openstackgerritVictor Stinner proposed openstack/neutron-lbaas: Port agent_manager to Python 3  https://review.openstack.org/29708010:04
*** cody-somerville_ has quit IRC10:05
*** lmiccini has quit IRC10:39
*** lmiccini has joined #openstack-lbaas10:40
*** prabampm has quit IRC10:41
*** rtheis has joined #openstack-lbaas11:29
*** yamamoto_ has joined #openstack-lbaas11:35
*** yamamot__ has joined #openstack-lbaas11:35
*** yamamoto_ has quit IRC11:39
*** yamamot__ has quit IRC11:42
*** yamamoto_ has joined #openstack-lbaas11:46
*** yamamoto_ has quit IRC11:55
*** yamamoto_ has joined #openstack-lbaas11:56
*** yamamoto_ has quit IRC12:01
*** ducttape_ has joined #openstack-lbaas12:11
*** prabampm has joined #openstack-lbaas12:26
*** ducttape_ has quit IRC12:32
*** matt-borland has joined #openstack-lbaas12:41
*** yamamoto has joined #openstack-lbaas12:49
*** prabampm has quit IRC13:02
*** piet has joined #openstack-lbaas13:04
*** amotoki has quit IRC13:05
*** neelashah has joined #openstack-lbaas13:12
*** yamamoto has quit IRC13:19
*** yamamoto has joined #openstack-lbaas13:21
openstackgerritMichael Johnson proposed openstack/octavia: Removes port_detach code that was mistakenly added  https://review.openstack.org/29738013:21
*** yamamoto has quit IRC13:28
*** yamamoto has joined #openstack-lbaas13:29
*** piet has quit IRC13:34
openstackgerritMichael Johnson proposed openstack/octavia: Adds a process to generate key flow documentation  https://review.openstack.org/29742313:34
*** yamamoto has quit IRC13:34
*** piet has joined #openstack-lbaas13:34
*** yamamoto has joined #openstack-lbaas13:49
*** ducttape_ has joined #openstack-lbaas13:59
*** piet has quit IRC14:02
*** yamamoto has quit IRC14:07
*** yamamoto has joined #openstack-lbaas14:09
*** yamamoto has quit IRC14:09
*** Bjoern has joined #openstack-lbaas14:17
*** ajmiller has joined #openstack-lbaas14:20
*** piet has joined #openstack-lbaas14:21
*** eranra has quit IRC14:21
*** fnaval has joined #openstack-lbaas14:28
*** minwang2 has joined #openstack-lbaas14:39
*** yamamoto has joined #openstack-lbaas14:42
*** fnaval has quit IRC14:46
*** cody-somerville has joined #openstack-lbaas14:59
*** cody-somerville has quit IRC14:59
*** cody-somerville has joined #openstack-lbaas14:59
*** TrevorV has joined #openstack-lbaas15:06
TrevorVHey johnsom_ or xgerman either of you guys up and running right now?15:07
*** yamamoto has quit IRC15:17
*** neelashah1 has joined #openstack-lbaas15:29
*** neelashah has quit IRC15:31
*** neelashah has joined #openstack-lbaas15:33
*** neelashah1 has quit IRC15:34
*** woodster_ has joined #openstack-lbaas15:39
*** Purandar has joined #openstack-lbaas15:47
johnsom_TrevorV What is up?15:55
*** johnsom_ is now known as johnsom15:56
TrevorVjohnsom I couldn't come up with an answer, but do we expose object limits via the API at all?15:56
TrevorVLike, how many listeners a LB can have?15:57
TrevorVthose kind of limits?15:57
TrevorVDo we HAVE those kinds of limits?15:57
johnsomOctavia does not.  I think that is handled in neutron-lbaas land right now15:57
bloganthere aren't quotas for how many listeners an lb can have15:59
bloganjust a quota for how many listeners a tenant can have, how many lbs a tenant can have etc15:59
TrevorVblogan can you retrieve that via the API?16:00
TrevorVOr is it just deployer known?16:00
bloganneutron/v2.0/quotas16:02
TrevorVThat's neutron-lbaas specific or just lumped together?16:02
*** piet has quit IRC16:06
*** pcaruana has quit IRC16:11
*** piet has joined #openstack-lbaas16:16
*** yamamoto has joined #openstack-lbaas16:17
*** alhu has joined #openstack-lbaas16:19
*** neelashah1 has joined #openstack-lbaas16:27
*** yamamoto has quit IRC16:28
*** neelashah has quit IRC16:28
bloganTrevorV: its neutron16:32
*** bana_k has joined #openstack-lbaas17:01
*** kevo has joined #openstack-lbaas17:08
*** ducttape_ has quit IRC17:09
*** SumitNaiksatam has quit IRC17:14
*** bana_k has quit IRC17:15
*** bana_k has joined #openstack-lbaas17:15
*** alhu has quit IRC17:40
*** SumitNaiksatam has joined #openstack-lbaas17:40
*** alhu has joined #openstack-lbaas17:41
*** ducttape_ has joined #openstack-lbaas17:41
*** piet has quit IRC17:51
*** crc32 has joined #openstack-lbaas18:04
*** Bjoern is now known as Bjoern_zZzZzZzZ18:06
*** piet has joined #openstack-lbaas18:14
*** minwang2 has quit IRC18:21
*** Bjoern_zZzZzZzZ is now known as Bjoern18:46
*** minwang2 has joined #openstack-lbaas18:47
*** Aish has joined #openstack-lbaas18:49
openstackgerritBanashankar k proposed openstack/octavia: Adding the topology info at the time of LB AMP association  https://review.openstack.org/28636518:50
*** alhu has quit IRC19:03
*** alhu has joined #openstack-lbaas19:04
openstackgerritBanashankar k proposed openstack/octavia: Adding the topology info at the time of LB AMP association  https://review.openstack.org/28636519:08
openstackgerritBanashankar k proposed openstack/octavia: Adding the topology info at the time of LB AMP association  https://review.openstack.org/28636519:19
*** crc32 has quit IRC19:21
*** piet has quit IRC20:35
*** rtheis has quit IRC20:58
*** ducttape_ has quit IRC21:00
*** TrevorV has quit IRC21:07
*** Bjoern has quit IRC21:27
*** matt-borland has quit IRC21:28
openstackgerritmin wang proposed openstack/octavia: Replace the os.open method with safer way  https://review.openstack.org/28820821:36
openstackgerritmin wang proposed openstack/octavia: Replace the os.open method with safer way  https://review.openstack.org/28820821:53
*** cody-somerville has quit IRC22:10
*** cody-somerville has joined #openstack-lbaas22:10
*** cody-somerville has quit IRC22:10
*** cody-somerville has joined #openstack-lbaas22:10
openstackgerritmin wang proposed openstack/octavia: Replace the os.open method with safer way  https://review.openstack.org/28820822:19
*** fawadkhaliq has joined #openstack-lbaas22:33
*** neelashah1 has quit IRC22:35
*** alhu has quit IRC22:49
*** fawadkhaliq has quit IRC22:53
*** ducttape_ has joined #openstack-lbaas22:54
*** fawadkhaliq has joined #openstack-lbaas22:56
*** fawadkhaliq has quit IRC23:05
*** piet has joined #openstack-lbaas23:06
*** fawadkhaliq has joined #openstack-lbaas23:12
*** fawadkhaliq has quit IRC23:12
*** fawadkhaliq has joined #openstack-lbaas23:13
*** piet has quit IRC23:21
*** fawadkhaliq has quit IRC23:26
*** fawadkhaliq has joined #openstack-lbaas23:26
*** fnaval has joined #openstack-lbaas23:28
*** fawadkhaliq has quit IRC23:41
*** fawadkhaliq has joined #openstack-lbaas23:41
*** fawadkhaliq has quit IRC23:47
*** fawadkhaliq has joined #openstack-lbaas23:47
*** ducttape_ has quit IRC23:55
*** ajmiller has quit IRC23:58
*** hockeynut has quit IRC23:59

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