Monday, 2017-04-24

*** JudeC has joined #openstack-lbaas00:01
*** JudeC has quit IRC01:26
*** KeithMnemonic has joined #openstack-lbaas01:28
*** yamamoto has quit IRC01:33
*** yamamoto has joined #openstack-lbaas01:33
*** amotoki has quit IRC01:36
*** yamamoto has quit IRC02:03
*** armax has quit IRC02:19
*** KeithMnemonic has quit IRC02:19
*** yamamoto has joined #openstack-lbaas02:28
*** Guest76746 is now known as med_02:37
*** med_ has joined #openstack-lbaas02:38
*** yamamoto has quit IRC02:50
*** yamamoto has joined #openstack-lbaas02:53
*** yamamoto has quit IRC03:04
*** aojea has joined #openstack-lbaas03:14
*** aojea has quit IRC03:18
*** csomerville has joined #openstack-lbaas03:27
*** cody-somerville has quit IRC03:31
*** thomasem_ has quit IRC04:04
*** yamamoto has joined #openstack-lbaas04:07
*** JudeC has joined #openstack-lbaas04:26
*** oomichi has quit IRC04:46
*** thomasem_ has joined #openstack-lbaas05:00
*** aojea has joined #openstack-lbaas05:07
*** fnaval has joined #openstack-lbaas05:19
*** gcheresh_ has joined #openstack-lbaas05:24
openstackgerritJude Cross proposed openstack/python-octaviaclient master: [WIP] Add rest of load balancer commands
*** aojea has quit IRC05:45
*** aojea has joined #openstack-lbaas05:47
*** aojea has quit IRC05:52
*** JudeC has quit IRC05:59
*** kobis has joined #openstack-lbaas06:07
*** thomasem_ has quit IRC06:16
*** thomasem_ has joined #openstack-lbaas06:20
*** aojea has joined #openstack-lbaas06:48
*** voelzmo has joined #openstack-lbaas06:53
*** aojea has quit IRC06:54
*** fnaval has quit IRC06:56
*** voelzmo has quit IRC07:02
*** voelzmo has joined #openstack-lbaas07:12
*** JudeC has joined #openstack-lbaas07:20
*** pcaruana has joined #openstack-lbaas07:24
*** rcernin has joined #openstack-lbaas07:43
*** JudeC has quit IRC07:44
*** amotoki has joined #openstack-lbaas08:18
*** rcernin has quit IRC08:23
*** aojea has joined #openstack-lbaas08:50
*** aojea has quit IRC08:55
*** aojea has joined #openstack-lbaas10:14
rm_workgoing to sleep now, catch you in a few hours11:10
nmagnezigood night :)11:10
rm_workok i feel a little dumb11:39
rm_worki mean, besides that i'm still not asleep11:40
rm_worksolved the problem with the wsme response type11:40
rm_workby realizing it's never an issue >_<11:40
rm_workfix inc11:41
rm_workalso added a few bug tags12:00
openstackgerritAdam Harwell proposed openstack/octavia master: Add support for single-create for APIv2
rm_workOK now bbl12:01
*** catintheroof has joined #openstack-lbaas12:37
nmagnezirm_work, I'll just update my devstack node with the latest changes you added before I start :D12:41
rm_workk erm12:41
rm_workyeah uh12:42
rm_workso about that12:42
nmagnezirm_work, no worries, just let me know when it's ready :-)12:42
rm_worklol nah i mean12:46
rm_workyou can12:46
rm_worki am about to post another fix12:46
rm_workbecause i started to sleep12:46
rm_workand was like12:46
rm_workah, why don't i just finish up the quotas stuff right now really quick12:46
nmagnezirm_work, meanwhile I'm gonna review this small patch:
openstackgerritAdam Harwell proposed openstack/octavia master: Add support for single-create for APIv2
nmagnezirm_work, trying to figure out how to test with curl again :P12:49
rm_workit's funny because it doesn't matter12:49
rm_workbecause i completely change all that code in the single-create patch anyway12:49
rm_workwell, the minimal listener stuff at least12:51
rm_workerr, minimalL712:51
rm_workanywho, quotas should be there12:51
rm_workI need to figure out functional testing for them :/12:51
rm_workright now the quota testing is braindead, just makes all quota checks fail, which isn't useful for checking if the Members quota is bad while everything else is good12:52
rm_workfor example12:52
*** jerrygb has joined #openstack-lbaas13:02
nmagnezirm_work, got it, so not sure why to have that small patch in the first place :P13:05
nmagnezirm_work, anyhow, I tried to test but i get the same result when I created a listener (this is how to test right?) both with your patch and with master.. strange13:06
*** cpuga has joined #openstack-lbaas13:24
*** cpuga_ has joined #openstack-lbaas13:25
nmagnezixgerman, o/13:26
nmagnezixgerman, I was trying to review
nmagnezixgerman, but couldn't get it to work like i expected,13:26
nmagnezixgerman, created a l7 policy linked it to a listener and queried for that listener13:27
xgermanit looks sane to me and it didn’t break the gates ;-)13:27
nmagnezixgerman, didn't see any diff from the result i get when using the master branch13:27
nmagnezixgerman, what did i miss ? :<13:27
nmagnezixgerman, LOL13:27
*** amotoki has quit IRC13:27
xgermanI see some things regarding the TLS refs13:28
xgermanbut we can’t test that since TLS is broken ATM13:28
nmagnezixgerman, you patch in this subject is ready for review?13:28
*** cpuga has quit IRC13:28
* nmagnezi spins up devstack13:29
nmagnezixgerman, , right?13:29
xgermanI think BBQ changed the format of their payload…13:30
nmagnezixgerman, gotta love openstack O_o13:39
nmagnezixgerman, stacking.. so it will take some time to be ready. meanwhile I'll try to understand what exactly changed13:40
xgermanon the TLS fron they now return six.text_type which can’t be turned into a byte array without decoding13:41
nmagnezixgerman, btw don't we have any test that covers this?13:41
xgermanmmh, I could probably force it to the two new types they (claim to) return13:42
nmagnezixgerman, <<- this?13:53
nmagnezixgerman, so from what I understand about this if they return six.binary_type but if they return six.text_type we'll need to decode13:58
xgermanyep, and this is what my code does13:58
nmagnezixgerman, there is no longer a way to make sure thet always return six.binary_type?13:58
xgermannope, that’s what they changed13:58
nmagnezixgerman, i'm trying to understand "I could probably force it to the two new types they (claim to) return" :-)13:59
nmagnezithis is what they return anyways, no?13:59
xgermanI can add some six.text_type(<TLS stuff>) to the tests13:59
xgermanour tests go off a constant14:00
rm_worknmagnezi: on that small patch, the only real difference is that the l7policy will only return as an ID14:00
rm_workinstead of as a full object14:00
nmagnezixgerman, meaning you'll use _conver in tests to cover both cases in the same test?14:02
nmagnezirm_work, maybe i should have used --debug in the cli client to see that diff14:03
xgermanI have kept it to the original tests — didn’t try to force it to text_type but I could14:03
nmagnezixgerman, maybe we should. so tests will cover cases where we get the text_type as well14:05
xgermank, will do14:05
rm_worknmagnezi: err, dunno about the CLI14:08
rm_workuse curl14:08
nmagnezixgerman, aye. I commented on the patch. I will also run the actual code when my devstack node is ready and report14:09
xgermansounds good14:09
nmagnezirm_work,  but I like CLI :<14:09
rm_workthe CLI doesn't *work*14:11
rm_workanyway, what I REALLY am hoping is that some folks can give the single-create a workout this week14:18
rm_workI'd like to get that + paging/sorting merged before the summit14:19
rm_workand also possibly filtering (which is my next target)14:19
rm_workthat'll allow the client stuff Jude is working on to fully function14:19
openstackgerritMerged openstack/octavia master: Correct some v2 Listener response entries
openstackgerritAdam Harwell proposed openstack/octavia master: Install DIB from source so depends-on will work
*** amotoki has joined #openstack-lbaas14:28
rm_workrebased that one, it had some weird failure14:28
rm_workoh right, sleep14:32
*** amotoki has quit IRC14:33
*** kobis has quit IRC14:39
*** kobis has joined #openstack-lbaas14:39
*** blogan has joined #openstack-lbaas14:40
*** kobis has quit IRC14:40
*** armax has joined #openstack-lbaas14:47
xgerman^^ rm_work what’s this unicode and py35 about14:48
*** armax has quit IRC14:57
*** gcheresh_ has quit IRC15:02
rm_workwhat patch is that part of15:07
rm_workhuh, master15:07
rm_workthat's, uhh15:07
rm_workI think that was done wrong15:08
xgermanyeah, not sure…15:10
rm_workyeah so15:10
rm_workwe kinda cloned what glance did15:10
rm_workand they've fixed it15:10
rm_workI feel like this basic exception thing should be boilerplate in oslo15:11
rm_worknot implemented EVERYWHERE15:11
xgerman^^ harlowja15:11
rm_workvery slightly differently to do the same thing15:11
xgermanok, I will ready a patch15:11
johnsomJosh isn't the oslo PTL anymore...15:12
rm_workactually maybe we cloned the neutron one, xgerman15:13
xgermanI think they all do the same15:13
rm_worklooks maybe more similar15:13
rm_work(they also fixed it, also differently)15:13
rm_workmaybe I'd do their version15:13
rm_worksince it looks more like ours15:13
rm_workjust proving that *every* project reimplements the *exact* same functionality15:14
rm_workjust ........ no two the same15:14
xgermanok, patch is forthcomong15:15
johnsomrm_work RE: quotas and your functional test comment earlier.  Did we not pull the functional tests over when we added quotas to the objects?15:16
rm_workwe did15:17
rm_worki think15:17
rm_workbut they're kinda braindead15:17
johnsomAh, so what is missing?15:17
rm_workso they work for testing individual objects15:17
rm_workbut trying to test a tree...15:17
rm_worki can't easily say "we're only over quota on members", for example15:17
rm_workI mean, it's not going to be that hard to fix, I think15:18
rm_workjust have to do it15:18
johnsomRight, ok.  Yeah single-call I may have not updated the testing for ti15:18
rm_workah and the quota test for listeners is missing15:18
rm_workwhich explains why we didn't catch that it wasn't ever added15:18
johnsomJust wanted to learn if I did something braindead when I fixed the braindead tests...  Grin15:18
rm_worki'm adding the missing listener quota test15:22
rm_workand then seeing if i can rewrite this quota handling15:22
rm_worki meant to sleep15:22
rm_workbut just kept working15:22
rm_workah LB quota test was missing too15:24
rm_workugh our quota exception blows15:25
rm_workno idea WHICH quota15:25
*** csomerville has quit IRC15:26
rm_worki guess that normally would be obvious15:26
*** csomerville has joined #openstack-lbaas15:26
openstackgerritGerman Eichberger proposed openstack/octavia master: Fixes unicode problems in p35
*** KeithMnemonic has joined #openstack-lbaas15:30
rm_workLOL I derped on the copy/paste15:39
rm_workbut it's ok i'm about to fix it15:39
rm_workmy test was behaving oddly, realized it's because I fixed the issue where it'd blindly say everything was fine even if you were checking the wrong quota :P15:40
rm_workhmmm this is odd15:43
rm_workL7Rule doesn't have quotas :P15:43
rm_workbut there's a test for it15:43
rm_workso I guess technically our quota check just does nothing for other objects lol15:45
*** armax has joined #openstack-lbaas15:54
*** amotoki has joined #openstack-lbaas15:56
rm_workyeah that was easier than I thought16:01
*** voelzmo has quit IRC16:02
openstackgerritGerman Eichberger proposed openstack/octavia master: Fixes Octavia not working with Barbican
openstackgerritAdam Harwell proposed openstack/octavia master: Add support for single-create for APIv2
rm_workjohnsom: ^^ quota issue fixed16:06
rm_workor rather, the testing16:06
rm_workerr, actually yes, also another quota issue16:07
rm_workI feel like our quota system is a little wacky16:07
rm_workbut I am not going to dig toooo deeply into that right at the moment16:07
rm_workI have plenty of other stuff to work on >_>16:07
xgermanwe need to get the TLS stuff straightened out… so we can use it in the lab ;-)16:11
rm_workI'll look at that shortly too16:11
rm_workI'd REALLY like to show off a working single-create16:11
rm_workwith TLS16:12
*** jerrygb has quit IRC16:16
openstackgerritAdam Harwell proposed openstack/octavia master: Implement sorting and pagination for octavia
rm_workxgerman: what were you doing to get that unicode bug to actually trigger <_<16:25
rm_workI'm not sure how to reproduce16:25
rm_worktesting seemed fine on both py27 and py35 obviously16:25
xgermanI was running tox -epep816:25
rm_workbecause our gates are still working16:25
rm_workerr wut16:25
xgermanon oython316:25
xgermanso it should slide by our gates16:26
rm_workis there an easy tox command to do that16:26
rm_workor do i need to edit tox.ini16:26
*** jerrygb has joined #openstack-lbaas16:33
johnsomThe only way I know to do it is to edit the tox.ini and add the override16:34
rm_workyeah, did that16:41
rm_worki feel like if i were a tox guru there's a slicker way to do this16:42
rm_workok BACK LATER16:45
*** sanfern has joined #openstack-lbaas17:02
*** cpuga_ has quit IRC17:13
*** cpuga has joined #openstack-lbaas17:13
xgermanafk for a bit17:20
harlowjaxgerman may be what u want?17:31
xgermanGotta check that out17:32
*** gcheresh_ has joined #openstack-lbaas17:39
*** sanfern has quit IRC17:48
*** eandersson has joined #openstack-lbaas17:54
*** amotoki has quit IRC17:54
*** chlong has joined #openstack-lbaas17:59
*** catintheroof has quit IRC18:01
xgermanwell, if I use that in18:13
xgerman__unicode__ as exception_to_unicode(self) I could get an inifnite loop18:14
*** JudeC has joined #openstack-lbaas18:16
harlowjathere used to be a base exception in oslo.utils for everyone18:18
harlowjabut never seemed to last18:19
harlowja is the closest that exists18:19
xgermanwell, I could use thst to rewrite something else…18:20
JudeCHey johnsom: Do you have any resources on the differences/similarities of the Octavia v1 vs v2 API? I'm trying to understand why there are 2 APIs and what I need to do as far as the client is concerned.18:32
xgermanJudeC Octavia’s V1 API was <only> for the communication lbaasv2 Octavia plugin <-> Octavia —- so shouldn’t have any bearing on V2 (though it’s similar)18:35
JudeCxgerman: thank you.18:37
*** catintheroof has joined #openstack-lbaas18:38
openstackgerritGerman Eichberger proposed openstack/octavia master: Fixes Octavia not working with Barbican
*** cpuga_ has joined #openstack-lbaas19:05
*** cpuga has quit IRC19:05
JudeCrest_method:: GET /v2.0/lbaas/loadbalancers/{loadbalancer_id}/stats19:10
JudeCdoes this not actually exist in the v2 api yet?19:10
johnsomJudeC: right, ignore v1 and status/stats isn't working yet19:13
JudeCah ok just making sure I am not crazy19:14
openstackgerritGerman Eichberger proposed openstack/neutron-lbaas master: Octavia Proxy Plugin
*** jerrygb has quit IRC19:42
*** pcaruana has quit IRC20:01
*** harlowja has quit IRC20:03
*** jerrygb has joined #openstack-lbaas20:17
*** jerrygb has quit IRC20:21
openstackgerritGerman Eichberger proposed openstack/octavia master: Fixes Octavia not working with Barbican
openstackgerritGerman Eichberger proposed openstack/octavia master: Fixes mysterious gdbm error
xgerman^^ I did a ton of simple fixes so approve away20:39
*** gcheresh_ has quit IRC20:44
*** harlowja has joined #openstack-lbaas20:59
*** catintheroof has quit IRC21:15
*** aojea has quit IRC21:26
rm_workThat's an interesting one xgerman21:48
xgermanYep . Was driving me insane21:49
nmagnezixgerman, looks like you are about to get you sanity back :)22:03
*** fnaval has joined #openstack-lbaas23:03
*** cpuga_ has quit IRC23:04
*** fnaval has quit IRC23:18
*** fnaval has joined #openstack-lbaas23:25
*** fnaval has quit IRC23:26
*** fnaval has joined #openstack-lbaas23:26
*** fnaval has quit IRC23:26
openstackgerritJude Cross proposed openstack/python-octaviaclient master: [WIP] Add rest of load balancer commands

Generated by 2.14.0 by Marius Gedminas - find it at!