Monday, 2014-08-04

*** banix has quit IRC01:02
*** banix has joined #openstack-lbaas01:16
*** banix has quit IRC01:29
*** banix has joined #openstack-lbaas01:42
*** HenryG is now known as HenryG_afk02:29
*** banix has quit IRC02:29
*** woodster_ has quit IRC02:35
*** banix has joined #openstack-lbaas03:00
*** banix has quit IRC04:23
*** rm_work is now known as rm_work|away06:18
*** evgenyf has joined #openstack-lbaas06:56
*** rm_work|away is now known as rm_work07:11
*** enikanorov__ has quit IRC09:24
*** HenryG_afk is now known as HenryG11:35
*** mestery_ is now known as mestery13:09
*** banix has joined #openstack-lbaas14:04
*** rolledback has joined #openstack-lbaas14:07
*** rolledback has joined #openstack-lbaas14:08
*** evgenyf has quit IRC14:17
*** markmcclain has joined #openstack-lbaas14:22
*** markmcclain has quit IRC14:38
*** markmcclain has joined #openstack-lbaas14:39
*** samuelbercovici has joined #openstack-lbaas14:58
*** samuelbercovici1 has joined #openstack-lbaas15:05
*** samuelbercovici has quit IRC15:08
*** samuelbercovici1 is now known as samuelbercovici15:08
*** sbalukoff has quit IRC15:11
*** xgerman has joined #openstack-lbaas15:11
bloganhello all15:26
mesteryblogan: Is this wiki still accurate for testing out the LBaaS changes:
mesterydougwig: ^^^15:29
*** mlavalle has joined #openstack-lbaas15:37
bloganmestery: yes it is15:41
bloganmestery: as long master doesn't get any new migrations15:42
*** samuelbercovici has quit IRC15:50
mesteryblogan: cool15:52
mlavalleblogan: did you see the email I sent last night?15:55
*** openstackgerrit has quit IRC16:01
*** openstackgerrit_ has joined #openstack-lbaas16:02
*** openstackgerrit_ is now known as openstackgerrit16:03
ctraceymestery et al: what are the plans, if any, for deprecation of object model v1?16:46
ctraceyif/when this lands will v1 be deprecated immediately?16:47
*** rolledback has quit IRC16:48
*** rolledback has joined #openstack-lbaas16:50
*** markmcclain has quit IRC16:52
dougwigctracey: without the shim and horizon, i'm not sure v2 should be the default, so i wouldn't be in favor of immediate deprecation.16:52
bloganmlavalle: yes I did and I plan on trying to reproduce your member delete issue16:52
bloganbut i've had some boot issues this morning16:52
ctraceydougwig: thanks16:53
ctraceyso we will effectively go experimental then live16:53
dougwiglooks like, unless we get a crazy push.16:53
mlavalleblogan: not necessarilly an issue with the LBaaS implementation. Might be a problem with the test script. I will reproduce it tonight and provide more details later. Last night I just ran out of time….16:53
bloganmlavalle: well i'll still try to reproduce it, thanks for giving me more work! you knew i'd fall into your web of deception and lies!16:55
TrevorVFor anyone who can answer; what do we do if we want to send an Octavia specific email?  Basically should the subject include some specific wording?  [Octavia] as a tag even though its not an openstack email?  Any ideas?16:59
bloganTrevorV: [octavia] in subject line17:04
TrevorVShall do17:04
dougwigoctavia is a stackforge project, so [Octavia] is all you need17:13
*** rolledback has quit IRC17:25
*** sbalukoff has joined #openstack-lbaas17:55
*** markmcclain has joined #openstack-lbaas17:55
*** banix has quit IRC18:23
*** markmcclain has quit IRC18:25
*** markmcclain has joined #openstack-lbaas18:26
*** banix has joined #openstack-lbaas18:32
*** xgerman has quit IRC18:35
*** xgerman has joined #openstack-lbaas18:36
*** crc32 has joined #openstack-lbaas18:47
*** rolledback has joined #openstack-lbaas18:49
*** evgenyf has joined #openstack-lbaas18:57
ctraceyblogan and dougwig could I get your thoughts on when you have a moment?19:00
bloganctracey: looking it? thoughts specifically on what the comments are asking or the entire review in general?19:03
ctraceyjust that specific comment. Not sure I am missing something.19:03
ctracey(if you are preoccupied, please ignore)19:03
bloganit sounds like he is saying parent_id is currently set dynamically19:05
bloganlike you have an instance of NeutronCommand and you hae to set parent_id on that isntance for the calls that have a parent to work19:05
ctraceybut i am not sure how has_parent alleviates that19:06
*** xgerman has quit IRC19:08
*** markmcclain has quit IRC19:08
*** xgerman has joined #openstack-lbaas19:10
*** evgenyf has quit IRC19:16
*** evgenyf has joined #openstack-lbaas19:16
*** woodster_ has joined #openstack-lbaas19:24
*** markmcclain has joined #openstack-lbaas19:29
*** fnaval has joined #openstack-lbaas19:46
dougwigctracey: so i see all these changes for parent objects, but if you look at v1, 'neutron lb-member-create', it does handle exactly the same linkage without all that.  why is that all changing?  (not arguing, curiosity.)19:59
ctraceydougwig: no argument with not arguing :)20:04
*** woodster_ is now known as woodster20:04
ctraceydougwig: REQ: curl -i GET20:04
ctraceypool_id seems to be handled in the body20:04
ctraceynot the path20:04
dougwigoh right, it's the nesting.20:04
dougwigthis is only the third time i've asked and you've answered.20:05
dougwigat some point, it'll sink in.  maybe.20:05
ctraceyno worries20:05
ctraceythe issue I am trying to resolve on that one is that I dont see how has_parent makes it any easier20:05
ctraceyone way or another parent_id needs to get set20:05
ctraceybut I could very easily be missing something in the comment20:06
*** rolledback has quit IRC20:10
dougwigmy read was that he was wanting things to be less dynamic and more explicit.  i'm not sure i agreed.20:12
bloganif it requires two variables where one suffices I don't agree with it either20:13
bloganmestery: had a chance to deploy and test out the lbaas code?20:13
mesteryblogan: I've been playing with it, the DB migration requirement means I can't just load the new settings in my local.conf, which caused me some issues :)20:14
mesteryblogan: I have it almost up now, back to where I was last week.20:14
bloganmestery: your local.conf for devstack?20:14
*** rolledback has joined #openstack-lbaas20:16
mesteryblogan: yeah20:17
mesteryblogan: hit an issue20:17
mesteryblogan: the client changes need a rebase I think, as there are conflicts following the instructions on the iki20:17
bloganmestery: ahh wait, there's the review out for the client in gerrit now so the instructions on that probably need to change to reflect that20:18
mesteryblogan: Sweet, can you update so I can try out the latest?20:18
bloganmestery: definitely20:19
mesteryblogan: thanks sir!20:19
*** evgenyf has quit IRC20:28
mlavalleblogan: ahhh! I jst read your last comment…. Now I understand what you said to me at the office…… :-)20:31
bloganmlavalle: lol that was like 4 hours ago20:32
mlavalleblogan: I grew up in Mexico and soap operas there are almost as mellodramatic as your comment :-)…. Yeah, I am the villain here20:34
*** markmcclain has quit IRC20:39
*** banix has quit IRC20:42
*** fnaval has quit IRC20:46
*** fnaval has joined #openstack-lbaas20:46
*** fnaval has quit IRC20:51
*** banix has joined #openstack-lbaas20:52
xgermanI am wondering who Diane Fleming is20:58
xgermanI think Min just took what's in the Google Doc and turned it into a doc -- but it looks like Diane wants to architect a different API?20:59
xgermanjust want to make sure not to annoy important people or at least get sbalukoff to do so :-)21:00
dougwigmestery: would it make the review process quicker if i spooled up a test environment at amazon or the like and just sent you login creds?21:04
dougwigthen we can keep it up to date as we make tweaks.21:04
mesterydougwig: Heh, that would be awesome! :)21:05
*** evgenyf has joined #openstack-lbaas21:07
*** markmcclain has joined #openstack-lbaas21:10
*** fnaval has joined #openstack-lbaas21:11
*** evgenyf has quit IRC21:12
*** rolledback has quit IRC21:15
xgermandougwig - you cna have HPCloud as well...21:17
xgermanno need to enrich AMZN21:17
dougwigthanks.  going to see if i can do it super quick by just spawning an extra instance of our thirdparty-ci slave.  if that's a rathole, i'll take you up on that.21:19
rm_workyeah, I could throw out a RS instance as well21:20
xgermanyeah, let me know and I can stand something up for you...21:20
xgermanrm_work you guys have vms, too?21:20
rm_workturns out we do :P21:20
dougwigi realized the amusement value of saying AWS when so much rackspace/HP cloud is present here.  that'll be a better longer-term answer.21:21
rm_workyeah, tell me what you need and I'll spin it up21:25
rm_workand add people's keys21:25
xgermansame here :-)21:25
rm_workwe can see which one works better :P21:25
dougwigha.  anything with 4gb+ of ram, single instance is fine.21:28
dougwig1VZNjJvSng6/ dougw@dougw-air21:28
dougwigpastebin auto-identify failed me on that one.21:29
blogani come back and see dougwig advocating using AWS21:31
* dougwig hides.21:32
xgermanbooting one up for you doug21:32
dougwigi could tell you all about how it was pure laziness with my third party CI, but I think i'd still get yelled at.21:33
xgermanlet me know if that works or if I made security rules too crazy :-)21:33
bloganctracey: testing the client out, it doesn't look like I can create a member by specifying the pool id in which it belongs, only the pool name21:39
ctraceyblogan: i will double check that21:39
bloganctracey: thanks21:39
ctraceyrefactoring all of that now anyway21:40
blogantenant_id could be the problem there21:41
blogandoesn't work with pool name either, sorry21:41
dougwigfor anyone that wants access to the shared "keep this thing up to date for reviewers to play with" server, or any of you that want to see it all in action, /msg me an ssh public key.21:55
*** vivek-ebay has joined #openstack-lbaas22:01
bloganmestery: finally got a chance to update that wiki page22:01
bloganmestery: though dougwig's solution is probably better22:02
mesteryblogan: thanks sir!22:02
dougwigmestery: can you /msg me an ssh public key?  it's almost done.22:02
mesterydougwig: One minute, apologies, on a call.22:02
dougwigplease just don't re-run, anyone.  there are a few hand tweaks that get blown away at present.22:05
rm_workerr, noted :P22:07
rm_workI just "./ && ./" repeatedly, currently :P22:07
dougwigctracey: everything is working except adding members to pools.22:07
dougwigthat gets a 404.22:07
dougwigmestery: it's ready whenever you are.  adding connection info to the wiki22:09
dougwighostname is ""22:11
dougwigblogan: have you pulled traffic?  trying all the namespaces on the box, 3 result in a hang, one in an immediate failure to connect.  i'll wager the immediate failure is the correct one for the vip, and it's not listening on port 80.22:15
blogandougwig: haha, yeah that will happen.  forgot a simple fix for that, ill be pushing up the fix today for that22:15
bloganlook at the port in neutron for the lbaas vip, ill wager its not active22:16
dougwigha, this demo box means that all of our sins will be open to the world now.22:16
dougwigmsg me a public key, and you can peek.22:16
*** VijayB has joined #openstack-lbaas22:23
*** mlavalle has quit IRC22:39
*** banix has quit IRC22:39
dougwigxgerman: a 16gb instance?  sweet.  :)22:46
rm_workoh man22:46
* rm_work sees if he can spin up a 20gb instance22:46
xgermanI can spin up bigger ones, too :-)22:47
* dougwig makes popcorn, waits to see how big the shared lbaas instance can grow.22:50
dougwigrax and hp fight!  go!22:51
* blogan embraces hp in a loving hug22:55
xgermanthe biggest I can get you is 122,880  MB22:57
xgerman(then I have to work my channels :-)22:58
rm_worknot sure if people would start sending me harshly worded emails if i had one of those on my internal account22:58
xgermanI had to increase my quota once22:59
ctraceymuch much much cleaner, IMO23:02
bloganctracey: yes it is, though a nit i have is to put module functions above all class definitions23:03
bloganthough thats just my personal preference23:04
ctraceyyeah I know...23:04
ctraceynot done yet23:04
ctraceyhave a few more to address and that was one of them23:04
blogandid you get a chacne to look at the member create issue?23:04
ctraceynot yet23:05
ctraceyi need to update my stack23:05
ctraceyto your latest23:05
blogani bet you could generalize that mixin to be ChildResourceMixin or something, so it can be reused, however its not worth doing for this23:05
ctraceyyeah - this is the only parent/child I can see23:06
bloganwith L7 there will be more23:06
bloganbut that can be done then23:06
ctraceymark indicated that they are looking to the unified CLI23:07
ctraceyso maybe that is ready then?23:07
bloganoh yeah23:07
bloganwhere's that work taking place?23:07
ctraceyhere me thinks:
*** mlavalle has joined #openstack-lbaas23:08
mlavalleblogan: I tested manually the listeners operations and delete works fine. it must be the test script or the ReST client in tempest23:09
bloganmlavalle: oh good (for me at least).  I haven't gotten enough time to actually test it out, but I know I had tested it out plenty ebfore23:10
bloganmlavalle: in the email i thought you said it was the memebr delete that wasn't working23:10
mlavalleblogan: crap.... in that case ignore my comments... going back to test23:11
dougwigxgerman: lbaas clearly needs a 120GB instance.  for... something.23:11
bloganmlavalle: lol web of deception and lies continues23:11
blogandougwig: cloud as a service23:12
ctraceyis it safe to pull latest and greatest?23:12
bloganctracey: it is23:12
mlavalleblogan: yeah, you are right.... I guess you are confronting me with the real mlavalle23:12
bloganthe real mlavalle is someone i wouldn't want to meet in dark alley somewhere23:12
mlavalleblogan: yeah, probably me neither23:13
bloganhopefully you don't come across a dark alley of mirrors23:13
ctraceyhere goes nothin ;)23:13
*** markmcclain has quit IRC23:13
ctraceyim betting this wont work23:13
ctraceyjust because....openstack23:13
bloganmost dangerous words in the world "watch this"23:13
bloganunless you qualify it with "this probably won't work"23:14
ctraceyblogan: you may remember me from such failed live demos as Keynote Portland 201323:14
*** mlavalle has quit IRC23:14
bloganis that a troy mcclure reference?23:14
bloganand did you really have a failed live demo in portland?23:15
ctraceyyes and yes23:18
xgermanI hope you being here is not the punishment for that23:18
bloganxgerman: did you happen to see the email in the neutron ML about the docs for lbaas v2?23:19
xgermanno,  but I will start searching23:19
dougwigxgerman: hp cloud needs to host their own ubuntu archive.  :)23:20
bloganxgerman: well the gist of it is that we aren't supposed to get rid of the v1 LBaaS docs, in fact we need those in as well, but need to have the v2 be tagged with experiemental23:21
rm_workdougwig: we do :P23:21
*** rolledback has joined #openstack-lbaas23:21
xgermanoh, ok23:21
bloganxgerman: do you think min can do that fairly quickly?23:21
xgermanI will ask her23:22
bloganxgerman: thanks, let me know23:22
xgermansure thing23:22
rm_workhmm I could create an onMetal server with 512GB of ram :P23:23
xgermanblogan another question: is the google API doc up to date?23:29
xgermanI always tell Min to check the code but...23:29
ctraceyblogan: subnet_id required on members/23:30
bloganxgerman: it should be up to date23:30
ctraceyand found the issue with creating...23:30
*** VijayB has quit IRC23:31
ctraceyprototype changed upstream for one of the functions23:31
bloganctracey: it is required, however there is an argument for it to not be required23:31
bloganbut we'll keep it required for now23:31
*** vivek-ebay has quit IRC23:31
*** vivek-eb_ has joined #openstack-lbaas23:31
bloganyou found the issue with creating?23:31
bloganoh the member23:32
*** vivek-eb_ has quit IRC23:32
*** vivek-ebay has joined #openstack-lbaas23:33
xgermanoh, actually that Diane Flemings who causes Min grieve works at Rackspace (here is a pick if you happen to see here at the Foodtruck Tuesday
bloganxgerman: lol yeah don't know her23:34
xgermanwell, she probably doesn't use the slide?23:34
bloganwell i mean if everyone used the slide there would be world peace right?23:35
bloganim leaving work to go home, ill be on later23:36
bloganadios amigos23:36
xgermanuse the slide!!23:36
*** VijayB has joined #openstack-lbaas23:37
*** vivek-ebay has quit IRC23:57
*** vivek-ebay has joined #openstack-lbaas23:58

Generated by 2.14.0 by Marius Gedminas - find it at!