Friday, 2014-07-11

sbalukoffblogan: Is it worthwhile to review your patches that aren't presently passing CI?00:10
ptoohillsbalukoff, the revision is def still a WIP, think we are just trying to get feedback during this process so changes/fixes can happen sooner rather then later. If you checkout the parent review it is in a "working" state but we would like to hear any thoughts, opinions or ideas for the code if possible00:29
sbalukoffptoohill: Sounds good.00:30
*** vivek-ebay has quit IRC00:54
*** fnaval has joined #openstack-lbaas01:25
*** johnw has quit IRC01:27
*** enikanorov__ has quit IRC01:27
*** rohara has quit IRC01:27
*** zz_blogan has quit IRC01:27
*** Youcef has quit IRC01:27
*** TrevorV has quit IRC01:27
*** jkoelker has quit IRC01:27
*** fnaval has quit IRC01:27
*** openstackgerrit has quit IRC01:27
*** sbalukoff has quit IRC01:27
*** dougwig has quit IRC01:27
*** ptoohill has quit IRC01:27
*** VijayB has quit IRC01:27
*** redrobot has quit IRC01:27
*** enikanorov_ has quit IRC01:27
*** enikanorov__ has joined #openstack-lbaas01:30
*** rohara has joined #openstack-lbaas01:30
*** zz_blogan has joined #openstack-lbaas01:30
*** zz_blogan is now known as blogan01:41
*** enikanorov_ has joined #openstack-lbaas01:58
*** redrobot has joined #openstack-lbaas01:58
*** jkoelker has joined #openstack-lbaas01:58
*** TrevorV has joined #openstack-lbaas01:58
*** openstackgerrit has joined #openstack-lbaas01:58
*** fnaval has joined #openstack-lbaas01:58
*** Youcef has joined #openstack-lbaas01:58
blogansbalukoff: im not sure why those are failing, they're not failing locally02:01
*** VijayB has joined #openstack-lbaas02:01
*** johnw has joined #openstack-lbaas02:03
*** sbalukoff has joined #openstack-lbaas02:05
*** dougwig has joined #openstack-lbaas02:05
*** ptoohill has joined #openstack-lbaas02:05
bloganthere yall are02:06
blogansbalukoff: im not sure why those are failing, they're not failing locally02:06
*** dougwig has left #openstack-lbaas02:06
*** dougwig has joined #openstack-lbaas02:06
blogandougwig: ping02:06
*** dougwig has left #openstack-lbaas02:06
*** ptoohill has left #openstack-lbaas02:06
*** ptoohill has joined #openstack-lbaas02:06
*** dougwig has joined #openstack-lbaas02:07
*** sballe has joined #openstack-lbaas02:07
blogandougwig ping02:07
ptoohillthat was odd02:08
blogandougwig: think i should change the stats call to take an object now?02:09
blogangood excuse to repush code to force another jenkins run since it failed for reasons it shouldn't have02:09
*** jkoelker has quit IRC02:10
dougwiggot it02:10
dougwigyes, for sure.  the tenant scheduling issue is something everyone will have, so the id isn't enough.02:11
blogando you think you could pull that review down and run those tests to see if they fail for you?02:11
dougwigi can also make the change if you don't want to be distracted.02:11
dougwigwhich?  oslbaas?02:12
bloganno directly from gerrit02:12
*** jkoelker has joined #openstack-lbaas02:14
dougwigto make sure i understand, you want me to try unit tests on vanilla neutron + 105609, right?02:15
bloganstupid vpn02:19
blogandougwig did you get my request?02:20
dougwigto make sure i understand, you want me to try unit tests on vanilla neutron + 105609, right?02:20
bloganjust run your noop driver tests02:22
bloganon py26 and py2702:22
dougwigsure, np.   cloning neutron, pulling in dependent, then 609.  standby.02:23
*** blogan_ has joined #openstack-lbaas02:27
blogan_alright screw the vpn02:27
blogan_happen to run those?02:28
dougwigyep, here ya go:02:28
*** blogan is now known as zz_blogan02:30
dougwigat least you know jenkins isn't lying02:31
blogan_jenkins is a liar02:32
blogan_but not in this case02:32
blogan_my local box is a liar02:32
blogan_okay so they renamed the db_base_v202:32
dougwigdoes this help?  "/home/dougw/work/openstack/neutron/.tox/py26/lib/python2.6/site-packages/sqlalchemy/ext/declarative/ SAWarning: This declarative base already contains a class with the same class name and module name as neutron.db.loadbalancer.loadbalancer_dbv2.SessionPersistenceV2, and will be replaced in the string-lookup table."02:34
blogan_thats what i've been seeing as well which i did not understand02:34
blogan_well not on my local box, but in jenkins02:34
dougwigi have a fast failure case.  where/what instrumentation do you want, or i can just dig into it.02:34
blogan_what was the tox command you ran?02:34
dougwigtox -e py2602:35
blogan_lol thats it02:35
dougwigpy27 assplodes (which is usually an import error)02:35
dougwiginteresting, since py26 passed when it was all one review.02:35
blogan_ok so then im getting what appears to be older code from the checkout link in gerrit02:36
blogan_what did you do to checkout the code?02:36
*** VijayB has quit IRC02:36
dougwigif you haven't in awhile, throw a -r into your tox (recreates vent)02:38
dougwighacking changes a lot02:38
blogan_its the code thats not changing02:38
blogan_so why doesnt the checkout url in gerrit give me the new code02:38
dougwigyou get my paste bin?  i just used that script on a virgin linux box and got the same error as jenkins02:38
blogan_yeah i got it, im just annoyed that the checkout command in gerrit is not giving me the same code02:39
blogan_i have to be doing something stupid02:39
dougwigmaster branch in neutron?02:40
dougwiggit screwed me over with oslbaas in a similar way a few weeks ago.  i finally "fixed" it by blowing away my repo and re-cloning.02:41
dougwigwhich is a cop out, i'll admit.02:41
blogan_well even gerrit is showing the same code im pulling down02:42
blogan_so it seems you cloning the review.openstack repo and pulling the changes down gave you the same code jenkins is testing02:44
blogan_im so confused and probably shouldn't be spending this much time trying to figure out why gerrit is not showing the right code but i have to know02:44
blogan_i believe im an idiot02:46
*** fnaval has quit IRC02:48
dougwigas long as you believe.02:49
blogan_i just need to rebase the stupid change request02:50
blogan_and fix the things that changed02:50
dougwignote to self: when 'git review' spews a bunch of errors on your commit message, it still accepts the review.  i unknowingly got myself to patch set 4 on a stupid script change, just trying to get it happy with my git commit message.  i hope all those CIs enjoy the churn.02:51
blogan_did you already change the stats?02:55
dougwigno, the last i saw you said you were doing it, and i didn't hear anything else.  i can start now.02:55
blogan_oh wait02:55
blogan_this is because you originally had the object02:55
blogan_this review doesn't have the change to the id yet02:56
blogan_its the review that depends on this one that does02:56
dougwigare you still pushing back to oslbaas?  if i do the stats change, should i do it there?02:56
blogan_no im not,02:58
blogan_you can just do it in gerrit if you want02:58
dougwigis it in more than one gerrit change?  i can just send you patches.02:58
blogan_only one review would need it changed02:59
blogan_i can do it pretty quickly, it needs to be rebased off this one anyway02:59
blogan_so im going to have to pull it down and change it anyway02:59
blogan_yo dawgin' it03:02
dougwigok, let me pull all four, make the stats change, then i'll send you a patch and let you split it?03:07
*** sballe has quit IRC03:09
*** fnaval has joined #openstack-lbaas03:10
blogan_i already made it03:11
dougwigoh, lol, that would explain why i was starting at what i thought was old code!03:13
dougwigto give you some flavor, this is the first three lines of our icehouse driver stats():03:14
dougwigso that's one more neutron db ref that you just killed.03:14
blogan_though they could have just called the plugin03:16
blogan_ah it was you, you could have called the plugin, but its gone now and doesnt matter03:17
dougwigi think this is the only one that i haven't gotten rid of...  is there a better way?03:17
dougwig(i inherited both of those snippets.)03:18
blogan_looking real quick hold on03:20
blogan_well there'd be another way but it really wouldn't be any better03:22
dougwigat some point, we should just add a method to the lb model to abstract that (i'm not saying you have to do that!)03:23
*** sbfox has joined #openstack-lbaas03:23
blogan_abstract what?03:24
dougwigload_balancer.ip_address already exists, maybe add a property load_balancer.floating_ip_addresses that returns a list of that internal ip's floats.03:25
blogan_isn't it vip_address?03:26
blogan_on loadbalancer03:26
dougwigi'm sure you're right; adjust the example to suit.  :)03:27
dougwigerr, no.  this is on members, not the vip.03:27
dougwigsorry, i meant Member.03:27
blogan_ah okay03:27
dougwigyeah, you can LB to a float or an internal (the ref doesn't, but there's nothing stopping it.)  that's what that snippet is doing.03:28
blogan_i bet you'll be able to get the fixed ips of a load balancer through the vip_port attribute03:28
blogan_nvm im stupid03:29
blogan_you're talking about floating ips03:29
dougwigfor members03:29
blogan_ah, how would that work then?03:30
blogan_how would a member have a floating ip03:30
dougwigyou can plug an LB into your external subnet, talk to members as floats, and the vip is handled directly via hardware.  or you can plumb it inside your private network, talk to members with internal, use floats to the hardware vip.  or you can have the LB straddle both, talk to members via internal, and export the vip itself directly (the last is best.)03:31
dougwigmember is just a nova instance usually; why can't it have a float if you wanted it to?03:31
blogan_yeah but it doesnt have to be, it can be an ip outside the network, but yeah if it is a nova instance, or any openstack addressable entity, i dont see why not03:33
dougwiglet's you do some different things, based on your application.  e.g., you can one-arm the first scenario, so outgoing traffic doesn't go through your LB.03:33
dougwigyeah, if the float lookup fails, we just use the ip handed to us, for that case.03:33
dougwigand it's configurable which we look for.03:34
blogan_sounds like a nice blueprint after all this is done03:34
dougwigyeah, we don't need to feature creep.  i'll use the neutron db for now, and abstract it later.03:35
blogan_good those tests are passing in jenkins now03:37
dougwigwhen you rebase, i get about one email every 10 seconds, from all the CI's.03:37
blogan_lol sorry03:38
blogan_and sometimes i rebase in gerrit, and then pull down and do changes and push back up03:38
blogan_but yeah im getting a ton as well03:38
blogan_i was just thinking it'd be nice to get my review folder back under control03:39
dougwigmy two biggest pet peeves about gerrit: it forgets all feedback on a new push, and it spams me with CI mail that i do not give a shit about, with no easy way to filter it.03:39
dougwigi care about humans and -1's, period.  they're fixing ci usernames, so we'll be able to filter soon.03:40
dougwigalright, i'm going to go find some food...  back later.03:42
blogan_yeah that will be great to filter by usernames03:44
*** HenryG has joined #openstack-lbaas03:56
*** sbfox has quit IRC04:14
*** blogan_ has quit IRC04:28
*** sab has joined #openstack-lbaas04:28
*** sbfox has joined #openstack-lbaas05:12
*** sbfox has quit IRC05:26
*** HenryG has quit IRC05:28
*** fnaval has quit IRC05:46
*** openstack has joined #openstack-lbaas07:41
*** sab has quit IRC09:29
*** sab has joined #openstack-lbaas09:47
*** openstackgerrit has quit IRC11:16
*** openstackgerrit has joined #openstack-lbaas11:17
*** sab has quit IRC11:59
*** mestery has joined #openstack-lbaas12:42
*** HenryG has joined #openstack-lbaas13:43
*** markmcclain has joined #openstack-lbaas13:50
*** xgerman has joined #openstack-lbaas13:54
*** xgerman has quit IRC14:03
*** fnaval has joined #openstack-lbaas14:08
*** fnaval has quit IRC14:38
*** vivek-ebay has joined #openstack-lbaas15:27
*** vivek-ebay has quit IRC16:00
*** sballe has joined #openstack-lbaas16:34
*** dlundquist has joined #openstack-lbaas16:53
*** VijayB_ has joined #openstack-lbaas16:58
*** VijayB_ has quit IRC16:58
*** sbfox has joined #openstack-lbaas17:17
*** VijayB_ has joined #openstack-lbaas17:36
*** markmcclain has quit IRC17:44
*** sab_ has joined #openstack-lbaas17:55
*** sab_ has quit IRC18:00
*** vivek-ebay has joined #openstack-lbaas18:00
*** VijayB_ has quit IRC18:32
*** markmcclain has joined #openstack-lbaas18:35
*** VijayB has joined #openstack-lbaas18:58
*** VijayB has quit IRC19:08
*** vivek-ebay has quit IRC19:18
*** markmcclain has quit IRC19:26
*** vivek-ebay has joined #openstack-lbaas19:59
*** VijayB has joined #openstack-lbaas20:00
*** mestery has quit IRC20:12
*** markmcclain has joined #openstack-lbaas20:52
*** markmcclain1 has joined #openstack-lbaas20:53
*** markmcclain has quit IRC20:56
dougwigquiet day, or did freenode choke again?21:17
*** VijayB_ has joined #openstack-lbaas21:25
*** VijayB_ has quit IRC21:25
*** VijayB_ has joined #openstack-lbaas21:26
*** VijayB has quit IRC21:28
*** VijayB_ has quit IRC21:29
*** VijayB has joined #openstack-lbaas21:32
*** markmcclain1 has quit IRC22:02
*** VijayB has quit IRC22:30
*** vivek-ebay has quit IRC22:58
*** HenryG has quit IRC23:22
*** dlundquist has left #openstack-lbaas23:29
*** sbfox1 has joined #openstack-lbaas23:31
*** sbfox has quit IRC23:35
*** sbfox has joined #openstack-lbaas23:36
*** sbfox1 has quit IRC23:36
*** vivek-ebay has joined #openstack-lbaas23:58

Generated by 2.14.0 by Marius Gedminas - find it at!