Tuesday, 2016-02-09

*** doug-fish has joined #openstack-lbaas00:01
sbalukoffjohnsom: Joy!00:04
*** outofmemory is now known as reedip00:05
*** armax has quit IRC00:19
*** armax_ has joined #openstack-lbaas00:19
*** doug-fish has quit IRC00:22
*** doug-fish has joined #openstack-lbaas00:24
*** doug-fish has quit IRC00:25
*** doug-fish has joined #openstack-lbaas00:26
*** doug-fish has quit IRC00:27
*** doug-fish has joined #openstack-lbaas00:30
openstackgerritMerged openstack/octavia: Adding "region and endpoint_type" parameters to barbican_acl.py  https://review.openstack.org/27286100:35
*** doug-fish has quit IRC00:40
*** Varun_Lodaya has joined #openstack-lbaas00:42
*** ajmiller has quit IRC00:44
*** doug-fish has joined #openstack-lbaas00:45
*** allan_h has joined #openstack-lbaas00:55
*** bank_ has quit IRC01:00
*** allan_h has quit IRC01:12
*** jwarendt has quit IRC01:13
*** doug-fish has quit IRC01:13
*** Varun_Lodaya has quit IRC01:14
johnsomWell, removing the workaround passed the gates (whatever that really means)01:21
*** crc32 has quit IRC01:23
*** Aish has quit IRC01:24
*** rooney has joined #openstack-lbaas01:24
*** Aish has joined #openstack-lbaas01:24
*** Aish has quit IRC01:26
*** rooney has quit IRC01:29
*** ducttape_ has joined #openstack-lbaas01:53
*** yamamoto has joined #openstack-lbaas01:59
*** ducttape_ has quit IRC02:05
*** ducttape_ has joined #openstack-lbaas02:05
*** _ducttape_ has joined #openstack-lbaas02:08
*** ducttape_ has quit IRC02:09
*** doug-fish has joined #openstack-lbaas02:14
*** doug-fish has quit IRC02:19
*** madhu_ak has quit IRC02:21
*** yamamoto has quit IRC02:24
*** doug-fish has joined #openstack-lbaas02:25
*** doug-fish has quit IRC02:45
*** doug-fish has joined #openstack-lbaas02:46
*** doug-fish has quit IRC02:50
*** armax_ is now known as armax02:54
*** sbalukoff has quit IRC03:00
*** _ducttape_ has quit IRC03:02
*** ducttape_ has joined #openstack-lbaas03:02
*** ducttape_ has quit IRC03:04
*** ducttape_ has joined #openstack-lbaas03:05
*** yamamoto has joined #openstack-lbaas03:12
*** fnaval has quit IRC03:25
*** doug-fish has joined #openstack-lbaas03:32
*** ajmiller has joined #openstack-lbaas03:34
*** rooney has joined #openstack-lbaas03:36
*** ajmiller has quit IRC03:37
*** fnaval has joined #openstack-lbaas03:37
*** rooney has quit IRC03:40
*** links has joined #openstack-lbaas03:41
*** ducttape_ has quit IRC03:51
*** ducttape_ has joined #openstack-lbaas03:53
*** doug-fish has quit IRC03:56
*** doug-fish has joined #openstack-lbaas03:56
*** doug-fish has quit IRC03:57
*** doug-fis_ has joined #openstack-lbaas04:00
*** ducttape_ has quit IRC04:01
*** doug-fis_ has quit IRC04:04
*** ducttape_ has joined #openstack-lbaas04:09
*** doug-fish has joined #openstack-lbaas04:09
*** armax_ has joined #openstack-lbaas04:18
*** armax has quit IRC04:21
*** armax_ is now known as armax04:21
*** sbalukoff has joined #openstack-lbaas04:30
*** allan_h has joined #openstack-lbaas04:44
*** ducttape_ has quit IRC05:06
*** amotoki has joined #openstack-lbaas05:07
*** amotoki_ has joined #openstack-lbaas05:09
*** amotoki has quit IRC05:12
*** amotoki_ has quit IRC05:15
*** allan_h has quit IRC05:28
*** rooney has joined #openstack-lbaas05:34
*** doug-fish has quit IRC05:48
*** amotoki has joined #openstack-lbaas05:49
*** doug-fish has joined #openstack-lbaas05:55
*** rooney has quit IRC06:03
*** doug-fish has quit IRC06:08
*** numans has joined #openstack-lbaas06:19
*** armax has quit IRC06:49
openstackgerritReedip proposed openstack/neutron-lbaas: [WIP]Add Timeout on Listeners- DO NOT REVIEW  https://review.openstack.org/27389606:56
*** amotoki has quit IRC06:56
*** amotoki has joined #openstack-lbaas06:58
openstackgerritStephen Balukoff proposed openstack/octavia: Update repos for L7 policy / methods  https://review.openstack.org/26552907:11
openstackgerritStephen Balukoff proposed openstack/octavia: Add L7 database structures  https://review.openstack.org/26543007:11
openstackgerritStephen Balukoff proposed openstack/octavia: Update repos for L7 rules / validations  https://review.openstack.org/27664307:11
openstackgerritMerged openstack/octavia: Do not use __builtin__ in python3  https://review.openstack.org/26251607:26
openstackgerritStephen Balukoff proposed openstack/octavia: Add L7 api - policies  https://review.openstack.org/26569007:29
openstackgerritStephen Balukoff proposed openstack/octavia: Add L7 api - rules  https://review.openstack.org/27771807:35
*** doug-fish has joined #openstack-lbaas07:38
*** doug-fis_ has joined #openstack-lbaas07:42
*** doug-fish has quit IRC07:43
*** doug-fis_ has quit IRC07:47
openstackgerritReedip proposed openstack/neutron-lbaas: Add Timeout on Listeners- DO NOT REVIEW  https://review.openstack.org/27389607:49
*** nmagnezi has joined #openstack-lbaas07:49
openstackgerritMerged openstack/octavia: Reverts a workaround that is impacting amp updates  https://review.openstack.org/27760208:19
*** jschwarz has joined #openstack-lbaas08:43
*** jschwarz has quit IRC09:23
*** jschwarz has joined #openstack-lbaas09:26
openstackgerritBrandon Logan proposed openstack/neutron-lbaas: Get Me A LB  https://review.openstack.org/25720109:38
openstackgerritBrandon Logan proposed openstack/neutron-lbaas: Get Me A LB  https://review.openstack.org/25720109:43
openstackgerritStephen Balukoff proposed openstack/octavia: Add L7 controller worker flows and tasks  https://review.openstack.org/27776810:17
*** amotoki has quit IRC10:34
*** amotoki has joined #openstack-lbaas10:49
*** amotoki has quit IRC10:55
*** yamamoto has quit IRC11:05
*** yamamoto has joined #openstack-lbaas11:51
*** rtheis has joined #openstack-lbaas12:23
*** openstackgerrit_ has joined #openstack-lbaas12:39
*** doug-fish has joined #openstack-lbaas12:45
*** yamamoto has quit IRC12:46
*** doug-fish has quit IRC12:47
*** doug-fish has joined #openstack-lbaas12:48
*** ducttape_ has joined #openstack-lbaas12:53
*** numans has quit IRC13:02
*** yamamoto has joined #openstack-lbaas13:03
*** woodster_ has quit IRC13:06
*** ducttape_ has quit IRC13:10
*** yamamoto has quit IRC13:16
*** links has quit IRC13:17
*** yamamoto has joined #openstack-lbaas13:23
*** numans has joined #openstack-lbaas13:29
*** numans has quit IRC13:34
*** openstackgerrit_ has quit IRC13:56
*** numans has joined #openstack-lbaas14:12
*** localloop127 has joined #openstack-lbaas14:15
*** links has joined #openstack-lbaas14:16
*** numans has quit IRC14:20
*** ducttape_ has joined #openstack-lbaas14:21
*** ducttape_ has quit IRC14:21
*** amotoki has joined #openstack-lbaas14:21
*** numans has joined #openstack-lbaas14:22
*** amotoki has quit IRC14:31
*** yamamoto has quit IRC14:32
*** yamamoto has joined #openstack-lbaas14:33
*** amotoki has joined #openstack-lbaas14:35
*** amotoki has quit IRC14:37
*** yamamoto has quit IRC14:37
*** prabampm1 has quit IRC14:39
*** amotoki has joined #openstack-lbaas14:41
*** amotoki has quit IRC14:41
*** nmagnezi has quit IRC14:49
*** amit213 has quit IRC14:51
*** links has quit IRC14:54
*** TrevorV has joined #openstack-lbaas14:54
*** Bjoern has joined #openstack-lbaas14:55
*** cgross has quit IRC14:56
*** cgross has joined #openstack-lbaas14:56
*** TrevorV has quit IRC14:57
*** TrevorV has joined #openstack-lbaas15:01
*** ducttape_ has joined #openstack-lbaas15:01
*** ducttape_ has quit IRC15:02
*** ducttape_ has joined #openstack-lbaas15:02
*** fnaval has quit IRC15:05
*** yamamoto has joined #openstack-lbaas15:07
*** TrevorV has quit IRC15:11
*** TrevorV has joined #openstack-lbaas15:12
*** barclaac has quit IRC15:13
*** barclaac has joined #openstack-lbaas15:13
*** neelashah has joined #openstack-lbaas15:13
*** ajmiller has joined #openstack-lbaas15:17
*** numans has quit IRC15:20
*** piet has joined #openstack-lbaas15:22
*** numans has joined #openstack-lbaas15:23
*** prabampm has joined #openstack-lbaas15:29
*** neelashah has quit IRC15:38
*** numans has quit IRC15:43
*** woodster_ has joined #openstack-lbaas15:45
*** yamamoto has quit IRC15:48
TrevorVAnyone online, can we get some more attention on:16:06
TrevorVIts an openstack/requirements issue that is currently breaking devstack16:07
*** armax has joined #openstack-lbaas16:08
johnsomIt looks like it has been +A'd16:13
TrevorVYeah, johnsom happened right after I reviewed it and linked it ha ha16:20
xgermanI don’t have +A in that project ;-)16:24
*** _ducttape_ has joined #openstack-lbaas16:26
*** jschwarz has quit IRC16:29
*** ducttape_ has quit IRC16:29
*** _ducttape_ has quit IRC16:36
*** ducttape_ has joined #openstack-lbaas16:37
*** piet has quit IRC16:40
*** TrevorV has quit IRC16:40
*** TrevorV has joined #openstack-lbaas16:42
*** neelashah has joined #openstack-lbaas16:44
*** piet has joined #openstack-lbaas16:46
*** neelashah has quit IRC16:54
*** Aish has joined #openstack-lbaas16:59
*** minwang2 has joined #openstack-lbaas16:59
kevinbentonxgerman: hey, how do you set a default flavor?17:03
kevinbentonxgerman: or is it even possible?17:03
*** fnaval has joined #openstack-lbaas17:03
xgermanin that case you would just use the default provider17:03
xgermanjwarendt is working on some more documentation :-)17:03
xgermanflavor is basically to put an end user readable name on providers + some additional flecibility17:04
kevinbentonso then a created thing might not have a flavor at all then?17:04
*** neelashah has joined #openstack-lbaas17:07
*** neelashah1 has joined #openstack-lbaas17:08
kevinbentonxgerman: ^^17:09
xgermannot necessarily17:10
*** piet has quit IRC17:10
*** jwarendt has joined #openstack-lbaas17:11
*** neelashah has quit IRC17:11
kevinbentonxgerman: so the lookup logic is: lookup service provider for flavor, if no flavor, use default service provider17:13
johnsomkevinbenton jwarendt is now online in this channel.  He can probably answer you questions17:13
jwarendtkevinbenton: Correct17:13
jwarendtThat is current logic.17:13
kevinbentoni see, thanks17:14
*** bana_k has joined #openstack-lbaas17:14
kevinbentonjwarendt: load balancer API doesn't allow choosing a service provider directly, does it?17:16
kevinbentonjwarendt: it's always indirect via a flavor17:16
xgermanno, we let you choose it directly17:16
jwarendtkevinbenton: Still has the --provider flag as well, for explicit provider17:16
jwarendtPredated flavors.17:17
kevinbentonjwarendt: other than legacy support, do you think that has a good use case?17:17
kevinbentonxgerman: why wouldn't they use a flavor?17:18
jwarendtI prefer the flavor operator-controlled indirection, obviously.  But flavor requires setup unless we automate the flavor creation/profile bindings based on providers.17:18
kevinbentonmakes sense17:19
xgermankevinbenton they might be evaluating a provider to turn it into a flavor — this allows for a quicker turn around17:19
kevinbentonxgerman: is it not possible to add a flavor but only have operators be able to use it?17:19
kevinbentonlike admin-only flavors or something?17:20
xgermanI don;t think we have that functionality17:20
*** allan_h has joined #openstack-lbaas17:20
kevinbentonxgerman: ok. Makes sense then17:21
kevinbentonDeciding if l3 should allow that as well17:21
kevinbentonProbably should for operator only stuff17:21
xgermanyeah, we like to allow for that flexibility17:23
kevinbentonjwarendt: so is a load balancers flavor and service provider stored in the schema?17:24
xgermanI think we store both17:24
jwarendtThe flavor id is stored with the lb, as is the provider.17:25
kevinbentonIt would seem necessary since a flavor could have multiple servic  profiles17:25
jwarendtThe flavor->service profile bindings are separate db structures; the lbaas only stores the flavor and the chosen provider.17:26
jwarendtAt present.17:26
kevinbentonOk. That's all I needed for now17:27
*** _cjones_ has joined #openstack-lbaas17:34
*** fnaval has quit IRC17:34
*** ajmiller has quit IRC17:39
*** minwang2 has quit IRC17:40
*** ajmiller has joined #openstack-lbaas17:42
*** prabampm has quit IRC17:44
*** fnaval has joined #openstack-lbaas17:58
rm_workis this one of our known intermittent failures? http://logs.openstack.org/91/259391/3/check/gate-neutron-lbaasv2-dsvm-scenario/295e7b5/console.html.gz#_2016-01-08_10_49_18_18317:59
rm_workcould be the third one you mentioned that i haven't actually seen yet?18:00
bloganso slooooowwww18:03
bloganrm_work: yeah i've seen it18:04
*** minwang2 has joined #openstack-lbaas18:05
*** woodster_ has quit IRC18:06
johnsomdougwig Thoughts on the comment here: https://review.openstack.org/#/c/274319/18:15
dougwigi was torn between thinking it was style and wondering if both can exist at the same time, and then being annoyed at having to go look, so fuck it.  i think it's fine/readable as is.18:16
johnsomOk, yeah, I don't have enough background on that.18:16
johnsomI will go solicit the +A then.18:16
johnsomblogan ajmiller ptoohill https://review.openstack.org/#/c/274319/  Gate speedup18:17
ajmillerjohnsom:  coincidentally I had just been looking at it.18:18
*** Aish has quit IRC18:18
johnsomCool, thanks18:18
*** madhu_ak has joined #openstack-lbaas18:19
*** bana_k has quit IRC18:27
*** Aish has joined #openstack-lbaas18:31
*** ajmiller has quit IRC18:33
*** ajmiller has joined #openstack-lbaas18:36
*** bana_k has joined #openstack-lbaas18:38
*** neelashah1 has quit IRC18:38
rm_workyeah i looked at ENABLE/DISABLE a while back and could never figure out exactly what advantage one has over the other, or how it chooses which to use if you used both?18:46
rm_workI think it's style / convenience, blacklist vs. whitelist both supported <_<18:46
TrevorVjohnsom I'm having an interesting issue here...18:53
johnsomOk.  I have ~7 minutes18:54
rm_workptoohill: +2'd, but it's just rax on it right now, waiting for a +A from someone non-rax T_T18:54
TrevorVMy unit tests are failing, saying "<Class> object has no attribute 'assertRaises'"18:54
johnsomYeah, I will take a look later today (I hope)18:54
rm_workah TrevorV where is that18:54
TrevorVIts on EVERY unit test I'm running right now in Pycharm18:54
rm_worksomething probably isn't the right kind of mock/baseclass18:54
rm_workthat seems bad18:54
rm_workdoes it NOT do it in tox?18:54
TrevorVHowever, when I run "tox -e py27" it throws a real error...18:54
rm_workyour environment in pycharm may not be set correctly, sometimes it gets "off" somehow on specific runs18:55
rm_workcheck the options and make sure it's using project default and that the project default is right18:55
openstackgerritMerged openstack/neutron-lbaas: Optimize API gate devstack build time  https://review.openstack.org/27431918:55
rm_worksometimes mine get messed up randomly18:55
TrevorVrm_work I'm not sure entirely where I should be looking though18:55
TrevorVYou mean project interpreter?18:55
TrevorVOr Run/Debug configs?18:56
rm_workclick the name of the run up by the Play/Debug buttons18:56
rm_workrun/debug config18:56
rm_workEdit Configurations...18:56
rm_workis Python Interpreter set to Project Default?18:56
TrevorVIts set to the .tox one18:56
rm_workpy2 or py3, out of curiosity18:56
rm_workwell i am about to head in18:57
TrevorVit doesn't have to be, that's just what I've done so far18:57
rm_workso if you don't figure it out in like 30m18:57
rm_worki'll take a look18:57
rm_worki have def seen that before18:57
rm_workoh hey, voting open for Austin talks?18:57
TrevorVI'm not sure.18:58
*** neelashah has joined #openstack-lbaas18:59
*** Bjoern has quit IRC19:21
TrevorVsbalukoff hey man you around?19:22
dougwigjohnsom: for the octavia meeting this week, can we set aside time to go over all the blueprints/rfes/m-3 bugs for neutron-lbaas and octavia, and where we're at with them? I need to know which to kick into newton.19:27
johnsomI think we have one blueprint in neutron for L719:27
TrevorVThink I found the answer sbalukoff no hurry man19:28
TrevorVWell now I'm seeing a ton of errors that don't quite make sense.19:32
TrevorVAll of a sudden these tests are getting "None" values for things that I doubt have had them for a while19:32
TrevorVAnd a bunch of reports of 500 response codes...19:33
TrevorVIs this due to the mock stuff you guys had issues with a while back?19:33
TrevorVNevermind, I'm just going to revert my simple changes and double check that it wasn't me19:34
*** rooney has joined #openstack-lbaas19:35
TrevorVAlright, looks like its something I did.19:36
TrevorVAlrighty.  Time to figure out what I did wrong.19:36
*** crc32 has joined #openstack-lbaas19:54
mhaydenblogan: ah, you were right on the security group -- port 443 isn't working yet, but i'm still digging20:03
mhaydenblogan: if i do a curl within the network namespace for the LB, i can connect to port 80 *and* port 443 without issue20:04
*** neelashah has quit IRC20:05
*** rcernin has joined #openstack-lbaas20:08
*** neelashah has joined #openstack-lbaas20:08
*** piet has joined #openstack-lbaas20:09
bloganmhayden: ah, just can't from the global namespace?20:11
bloganerr default namespace, whatever its called20:11
mhaydeni'm on another instance on the same vxlan20:11
mhaydenand i can reach the nodes directly on 443 without issues20:11
mhaydenthe traffic *is* making it to the network namespace though20:11
mhaydenverified that with tcpdump20:12
*** piet has quit IRC20:13
*** piet has joined #openstack-lbaas20:13
mhaydeni think haproxy is choking on something20:14
mhaydenthe strace looks odd20:14
*** Varun_Lodaya has joined #openstack-lbaas20:23
bloganmhayden: but if haproxy is able to handle the traffic from within the namespace, it should be working fine20:38
mhaydenhmm, so i backed out the members, pool and listener20:38
mhaydenthen jammed them back in20:38
*** SumitNaiksatam has joined #openstack-lbaas20:38
mhaydenit seems to work!20:38
bloganjust in the haproxy config? or through the api20:39
bloganand it works totally now?20:39
mhaydenthat was... odd20:39
blogansounds like a great user workflow, create all your stuff, then delete htem, then add them back20:39
bloganworks as intended!20:39
mhaydeni'll retry and see what happens20:41
mhaydenthe first time, i copied/pasted the neutron commands20:42
mhaydenthe second time, i did it a little more slowly20:42
mhaydeni wonder if there's a timing bug in there somewhere20:42
mhaydenbut i did find a possible bug -- i can delete members, pool, and listener without deleting the healthmonitor20:43
mhaydenbut then the health monitor is in an undeletable state20:43
mhaydenbecause it has no pools attached20:43
bloganmhayden: on liberty?20:44
bloganmhayden: seems like an obvious bug, wonder if its been fixed20:45
blogansounds kind of familiar, may just need a backport20:45
mhaydenblogan: i'll be glad to open a bug in LP20:45
bloganmhayden: do it!20:45
mhaydenwhich project?20:45
bloganneutron, tag with lbaas20:45
mhaydenin progress..20:45
openstackLaunchpad bug 1450375 in neutron "cannot delete v2 healthmonitor if the hm-associated-pool was deleted first" [Undecided,In progress] - Assigned to LIU Yulong (dragon889)20:46
mhaydenthat looks like it, actually20:46
*** piet has quit IRC20:46
bloganmhayden: well its in progress lol20:49
blogantime to review that sucker, just got resurrected20:50
blogan2 weeks ago...20:50
mhaydeni would be glad to gander, but i'd need to get familiar with neutron's innards first20:51
bloganmhayden: you won't if you value your soul20:51
*** crc32 has quit IRC20:56
*** neelashah has quit IRC21:16
mhaydenblogan: so the security group didn't appear to matter21:20
mhaydenbut i'm going to tear it all down and rebuild it to verify21:21
bloganmhayden: it may not, probably because that vip port doesn't get attached to a VM, so it does't have a host/hypervisor that manages security groups, but i coudl be wrong, i'm just making uneducated guesses here21:22
bloganbut pretty sure its because the vip port is being used on some agent host that obviously wouldn't have a networking agent to manage those ip table rules21:23
TrevorVAnyeone else see this issue and know a fix? https://gist.github.com/anonymous/4ba8be1a065e5962eb0a21:25
johnsomNew one to me21:26
johnsomTrevorV http://stackoverflow.com/questions/12805044/python-console-importerror-pyunicodeucs4-fromstring21:28
johnsomDid you see that?21:28
TrevorVI did.21:28
TrevorVjohnsom I was looking through this initially, but there is no "fix" here.  Just comments about the problem.21:30
TrevorVI'm not compiling something, I'm trying to run tox -e py2721:31
TrevorVWell... that may be me misunderstanding exactly what's happening when I run that.21:31
TrevorVI guess.21:31
johnsomIt sounds like the package was compiled with a different version of python21:31
johnsomptoohill Around?21:45
ptoohilljohnsom: I am21:45
johnsomHow do you feel about moving the jinja template stuff into controller/worker/utils?21:46
ptoohillTotally ok with that, and was something I was thinking about21:46
johnsomThe tasks directory is really for the taskflow tasks21:46
ptoohillAgreed, i was just following others setup21:46
johnsomCool.  It would just keep our directory tree clean21:46
ptoohillmaybe at somepoint we can move all the other templates to a common spot21:47
johnsomOk, that is the only comment I had for the patch.  +2 after that21:47
johnsomYeah, I'm open to starting a wider directory for that now if you're up for it21:47
ptoohillMaybe a common/templates?21:48
johnsomThat was what I was thinking too21:48
ptoohillIll just put my stuff there for now21:48
johnsomStart a trend....21:48
ptoohillI'll get that out21:48
*** neelashah has joined #openstack-lbaas21:49
*** Bjoern has joined #openstack-lbaas21:51
sbalukoffjohnsom: Since the next step in L7 is to update all the jinja templates for generating haproxy configs, should I move them there as well?21:52
johnsomIf you're up for the work, sure that would be cool21:52
sbalukoffEh... I've got to mess with pretty much everything there anyway.21:53
johnsomI expect we can get ptoohill's patch in today, so you shouldn't have to do a dependency21:53
*** rooney has quit IRC21:54
johnsomsbalukoff What is the link for the L7 etherpad?  I keep forgetting to add it to the meeting agenda21:54
ptoohillTesting it out21:55
sbalukoffTrevorV: What was the issue you were running into earlier about which you tried to ping me?  (Sorry-- was out at a doctor's appointment earlier.)21:55
TrevorVUmmm... I forget.21:56
TrevorVYou took too long21:56
sbalukoffYeah, I do that.21:56
ptoohillYou want to -1 the patch johnsom ?21:57
ptoohillforce me to do it :)21:57
ptoohillsbalukoff: If youre going through updating all the jinja configs it would be good to get them in a common place, if you don't mind. But dont think that would be a -1 if you didnt21:58
johnsomYou just want our conflict numbers to go up.... grin22:00
ptoohillhelping out where i can22:00
sbalukoffptoohill: Again, I don't think it'll be significantly more work than I already have to do there. :)22:02
ptoohill:) You say that, then it turns into nightmare22:03
ptoohillJust trying to jinx ya ;)22:03
sbalukoffHeh! Thanks!22:04
ptoohillNo problem, glad I could help!22:04
johnsomsbalukoff Do you want me to assign these shared pools related bugs to you?22:10
sbalukoffjohnsom: Sure!22:12
johnsomOk.  I am going to tag them too.  I just don't want them to get lost, etc.22:12
johnsomdougwig Question about the octavia launchpad bugs.  I see Martin Hickey tagging octavia bugs with "lbaas".  Is octavia repo included in the neutron bug scrub now?22:18
openstackgerritPhillip Toohill proposed openstack/octavia: Allow user-data on amphora creation  https://review.openstack.org/27555822:19
ptoohillAlrighty, That should be good22:19
johnsomptoohill Thanks, looking now22:19
ptoohilli may have missed tests ><22:20
ptoohilljohnsom: Not ready22:20
*** localloop127 has quit IRC22:20
ptoohilltested manual, forgot about units22:20
johnsomHelping to pad my review stats22:21
johnsomYou can set the topic to bug/1541231 as well.  Not sure if it matters, but that is the format in the dev guide22:22
ptoohillYea, not sure why it doesnt pull that with it, i had a rebase conflict, while inside branch and then this happens22:23
ptoohilland it even says it pushed to the branch, i dont get it22:24
ptoohill* [new branch]      HEAD -> refs/publish/master/bug-154123122:24
dougwigjohnsom: no, not to my knowledge22:24
johnsomOk, just trying to be an over-achiever then.22:24
ptoohillit looks like the topic is right hee22:25
johnsomdougwig For reference: https://bugs.launchpad.net/octavia/+bug/1543356  I removed the lbaas as I didn't want folks to confuse the octavia API vs. lbaas22:26
openstackLaunchpad bug 1543356 in octavia "Errors found for octavia api----get listener stats" [Medium,New] - Assigned to Stephen Balukoff (sbalukoff)22:26
*** ducttape_ has quit IRC22:31
openstackgerritPhillip Toohill proposed openstack/octavia: Allow user-data on amphora creation  https://review.openstack.org/27555822:34
openstackgerritGerman Eichberger proposed openstack/neutron-lbaas: [WIP] Adds Cascade option for load balancer drivers  https://review.openstack.org/26823722:34
ptoohillOk johnsom, I think that's good now22:35
xgermanblogan I need some tutoring on that. That LBaaS V2 code is a dungeon22:38
xgermanwe definitely need some devrefs22:38
bloganxgerman: we sure do22:38
xgermanwell, I am lost — not sure why my tests fails? why plugin doesn’t have cascade...22:38
xgermanmight want to give  a talk on how that works :-)22:39
openstackgerritBrandon Logan proposed openstack/neutron-lbaas: Get Me A LB  https://review.openstack.org/25720122:40
bloganxgerman: all of 2 people would show up to that talk22:41
xgermanwell, I might need some tutoring - then whatever we did at the mid cycle doesn’t pass unit tests22:42
johnsom3 people22:43
fnavali'll attend too.22:44
fnavalhmmm. that was a sarcastic remark i think22:44
xgermanguess it’s a crowd :-)22:44
xgermanblogan webinar?22:44
johnsomWe all just like to watch blogan be uncomfortable presenting....22:44
blogani think thats the real reason, xgerman really has cascade delete working22:45
xgermanhonestly, the inner working of LBaaS V2 are completely lost on me22:45
bloganxgerman: its mainling the black magic that is neutron extension loading and service plugin loading22:45
xgermanblack magic there we go - I was supposed to sacrifice a goat before running too?22:46
fnavalyeah i was gonna ask about that.  if we're working on 1 load balancer create call, should there be 1 load balancer delete (everything) call?22:46
bloganxgerman: indeed22:46
*** rtheis has quit IRC22:46
fnavaland update too22:46
bloganfnaval: that is what xgerman is doing and what every UI team wants22:46
bloganupdate is low on priority22:46
fnavalkk blogan thanks22:46
bloganUI team should just use heat22:47
*** TrevorV has quit IRC22:49
xgermanbloagn sure — or ansible22:50
* johnsom throws something heavy and *hot* at blogan22:50
xgermanor somebody ought to start oslo-orchestration22:50
bloganxgerman: do you have your code up?22:50
xgermanyep, but it’s broken and I don’t know why22:50
bloganlink me22:50
xgerman https://review.openstack.org/268237 [WIP] Adds Cascade option for load balancer drivers22:51
bloganbroken on neutron-server start or when trying to actually delete?22:51
xgermanbroken when delete in unit tests22:51
xgermanit just adds the option22:51
xgermandrivers are later22:51
xgermanit will try cascade and i that fails do the old thing22:52
xgermanto some degree22:52
xgermanthis pains me22:53
xgermanthere must be some magic piece I am missing22:54
blogannothing is throwing the TypeError exception though22:55
xgermanit sure did - I had like 30 failures in the unit tests without22:55
xgermanyeah, there must be another place we need to add cascade=True or False22:56
bloganill test it out real quick22:56
xgermanI hope neutron-lib will make that ll ten times cleaner22:57
bloganneutron lib will make a lot of things cleaner but i don't think this will be one of them22:57
*** rcernin has quit IRC22:58
xgermanmaybe pecan then?22:58
xgermanor do we need another project?22:59
bloganoh there's the problem23:04
bloganneed to add **kwargs to the _call_driver_operation method23:04
bloganand pass that to the actualy driver method call23:04
bloganxgerman: also its not using the noop drivers it appears23:08
bloganxgerman: nvm too quick on that assertion23:08
xgermanyeah, this is like a riverwalk with snakes instead of water23:09
ptoohillon fire?23:10
ptoohillthat's a new feature23:10
bloganalright unit test passes xgerman23:14
bloganill push up the changes23:15
xgermanWhat did I do wrong?23:15
blogansince no driver method had taken a kwarg before, the _call_driver_operation didn't expect one, so it needed to be able to do that23:15
bloganthen within the tests you create a load baalncer and listener, but those creation contexts (using with keyword) automatically delete themselves with that context is left, so needed to pass the no_delete kwarg so it doesn't try to clean itself up23:16
openstackgerritBrandon Logan proposed openstack/neutron-lbaas: [WIP] Adds Cascade option for load balancer drivers  https://review.openstack.org/26823723:16
bloganyou'll have to change the logic of how you determine whether a driver supports the cascade and if it does call the old way23:17
xgermanmmh, so TypeError no good23:17
xgermanI guess the driver would ignore the kwarh and just fail since not cascading23:18
xgermanthat logic is in there23:18
bloganTypeError was being thrown because tha tmethod wasn't expecting a keyword argument23:18
xgermanok, I remove type error and we are all good ;-)23:18
bloganbut old drivers that know about the cascade probably do not have their delete method taking **kwargs, so it'll fail on old drivers23:19
bloganthat dont know23:19
xgermanTypeError to the rescue :-)23:20
bloganTypeError is too vague!23:20
blogani guess it'll work damnit23:20
xgermanwell, https://hynek.me/articles/hasattr/ suggests to use TypeError instead of the more fancy stuff23:21
xgermanor in his case AttributeError — guess I am oversimolifying23:21
*** yamamoto_ has joined #openstack-lbaas23:21
blogani was thinking more that you could call a driver.cascade_delete method, and if it raises NotImplemented, then faill back to the old way23:22
xgermanmmh, probably something to ask our resident vendor driver expert dougwig23:23
*** yamamoto_ has quit IRC23:23
bloganwhich sounds more appropriate, you'd just have to make the base load balancer manager have the method cascade_delete and raise NotImplemented23:23
bloganso now every driver gets taht method for free, and only those who do implement it can23:24
xgermanyeah, I can see that… also my way would work in a more ugly fashion...23:24
xgermanThanks, I can add that more clean way23:25
bloganalso, got other comments, ill add them to the review though23:25
bloganspecifically about the extension loading23:25
xgermandang — guess I have my work cut out23:26
bloganadded to the previous PS bc i didnt feel like rewriting23:26
blogannah they're simple, just inside knowledge of the extension loading basically23:26
xgermanoh, I was afraid I would have become like gorge and need to go back to programming school23:26
blogangorge works23:27
bloganjorgem was his programming name, his manager name is gorge23:27
bloganme pushign up that review auto-assigned me to that bug!23:33
bloganxgerman: you tricked me!23:33
ptoohilloh noes23:34
bloganwell i removed myself, but can't readd you xgerman, it can't find your name, nor can it find mine23:35
*** allan_h has quit IRC23:39
*** neelashah has quit IRC23:44
xgermanbug without a name23:46
*** Bjoern has quit IRC23:46
*** yamamoto_ has joined #openstack-lbaas23:50

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