Friday, 2014-07-25

*** fnaval has quit IRC00:01
*** xgerman has joined #openstack-lbaas00:01
*** vivek-ebay has quit IRC00:02
*** xgerman has quit IRC00:09
*** mlavalle has quit IRC01:13
*** mestery_ has joined #openstack-lbaas01:31
*** mestery has quit IRC01:33
*** mestery_ has quit IRC01:34
*** mestery has joined #openstack-lbaas01:35
*** sbfox has joined #openstack-lbaas01:35
*** sbalukoff has quit IRC01:41
*** vivek-ebay has joined #openstack-lbaas01:43
*** mlavalle has joined #openstack-lbaas02:01
*** fnaval has joined #openstack-lbaas02:07
*** sbfox has quit IRC02:15
*** vivek-ebay has quit IRC02:16
*** mestery_ has joined #openstack-lbaas02:26
*** mestery has quit IRC02:26
*** mestery_ is now known as mestery02:27
*** mlavalle has quit IRC02:30
*** mlavalle has joined #openstack-lbaas02:32
*** HenryG is now known as HenryG_afk03:08
*** woodster__ has quit IRC03:35
*** vivek-ebay has joined #openstack-lbaas03:50
blogananyone on?03:52
*** vivek-eb_ has joined #openstack-lbaas04:03
*** vivek-ebay has quit IRC04:06
*** fnaval has quit IRC04:28
*** vivek-eb_ has quit IRC04:49
*** johnw has quit IRC05:33
bloganso i had a feeling passing the sqlalchemy models was goign to bite us in the ass05:45
dougwiguh oh05:58
dougwigthose are not helpful opening words.05:59
dougwig /helpful/hopeful/05:59
dougwigwhat's up?06:00
blogansince sqlalchemy keeps the model isntances updated whenever any change happens, the we aren't able to actually get an original model state for updates06:01
dougwigpass a dict for 'old'?  or there is no clone method or somsuch/06:02
bloganive looked, there's an expunge method on the session that takes it out of the session so it doesn't get updated, but it destroys all the relationships as well06:04
dougwigis it possible to get it, open a transaction, then get a new one for the update?  the transaction should keep it look old.  that's a totally random guess.06:04
bloganthat would require a new session, and i believe the plugin hast o use the one passed to it06:04
bloganbc that first get will get updated on the new one being updated, if they both share the same session06:05
blogani was planning on making data models anyway, just not so soon but i believe thats what I will have to do06:05
bloganalready got them written up, just need to add the code to convert SA models to these popos06:06
dougwigoy, so much for the code hardening a bit.  :)06:06
bloganwell the first one won't chagne06:07
bloganbut yeah its a bit of a downer06:08
bloganfor the 2nd, huge commit which is the main one06:08
dougwigthe model classes themselves are just a bunch of @property's, with fan out for associations, right?  (yeesh, that could get memory messy, not being lazy any longer.)06:10
blogani was going to have to do eager loading anyway06:13
bloganbut i may have found a solution to my problem with help from people ine #sqlalchemy06:14
dougwignice, then i'll stop reading the sqlalchemy docs.06:16
bloganwell i need to test it out first06:16
dougwigyou know, my driver is up to patchset 17.  and i've made exactly one edit: adding a comment to a conf file.  :)06:20
blogani think we have both learned some valuable lessons in this whole review process06:21
dougwigi think there are more to come.  :)06:25
blogandamnit, it did fix some of it, but not all of it06:27
dougwigif you have to go with new model classes for driver calls, you could limit the churn initially by just using them for the 'old' models, and keep using sqa everywhere else (which is 95% of the calls, and 99.99999% of the use cases.)06:30
bloganim not sure if that will result in a big difference in churn, all the models would still have to exist and whenever I retrieved hte "old" model i'd have to convert that over and once that work is complete it really shouldn't be too much more effort for everything else06:32
dougwigok, i won't push on reviews until early next week, then, to give things time to settle down.06:34
bloganyes time to settle down will be needed06:35
*** mlavalle has quit IRC06:43
rm_workblogan: >_>07:20
*** sbalukoff has joined #openstack-lbaas08:14
*** HenryG_afk is now known as HenryG12:42
*** woodster__ has joined #openstack-lbaas12:57
*** fnaval has joined #openstack-lbaas13:04
*** rolledback has joined #openstack-lbaas13:26
*** fnaval has quit IRC13:36
*** rolledback has quit IRC13:38
*** markmcclain has joined #openstack-lbaas14:09
*** markmcclain has quit IRC14:51
*** TrevorV_ has joined #openstack-lbaas15:09
*** rolledback has joined #openstack-lbaas15:18
*** sbfox has joined #openstack-lbaas15:35
*** sbfox1 has joined #openstack-lbaas15:38
*** sbfox has quit IRC15:39
*** fnaval has joined #openstack-lbaas15:42
*** jorgem has joined #openstack-lbaas15:47
*** TrevorV_ has quit IRC15:51
*** fnaval has quit IRC16:05
*** fnaval has joined #openstack-lbaas16:06
dougwigblogan: are the new models in yet?  can we start beating on them?16:27
*** mlavalle has joined #openstack-lbaas16:30
*** rolledba_ has joined #openstack-lbaas16:32
*** rolledback has quit IRC16:32
*** rolledba_ has quit IRC16:57
*** fnaval has quit IRC16:57
*** rolledback has joined #openstack-lbaas17:10
*** crc32 has joined #openstack-lbaas17:13
blogandougwig: no i gotta get some sleep sometime you robot!17:22
blogandougwig: ive also gotta research this quota thing too that oleg brought up17:22
*** rolledback has quit IRC17:25
dougwigyou can sleep after juno-3.17:26
dougwigcan someone else help with the quota thing?17:26
blogananyone is welcome to take a stab at it17:26
*** rolledback has joined #openstack-lbaas17:37
*** rolledback has quit IRC17:41
*** fnaval has joined #openstack-lbaas17:50
bloganalrigh tfigured the quota issue out17:55
bloganone more change to the extension review17:59
* blogan crosses fingers17:59
*** markmcclain has joined #openstack-lbaas18:06
*** markmcclain1 has joined #openstack-lbaas18:09
*** markmcclain has quit IRC18:10
crc32Hey stephen have you heard anything about the "common module" being added where I can populate the X509 functionality code at? I haven't heard from evengy.18:12
crc32Hey sbalukoff have you heard anything about the "common module" being added where I can populate the X509 functionality code at? I haven't heard from evengy.18:12
sbalukoffcrc32: I have not heard anything specific about it, from Evgeny or otherwise.18:16
crc32Do you have a link to the draft your proposing for the paris summit?18:17
sbalukoffcrc32: Yes... hold on...18:18
crc32sbalukoff: So should I wait for evengy to respond or do you know of a place to place my code?18:20
sbalukoffI think my team is planning on submitting this today:
sbalukoffcrc32: To be honest, I don't really have a good idea. My suggestion would be to place it somewhere that you think makes sense given the structure of the code repository, but be open to moving it all elsewhere (hopefully) pretty early in the review cycle.18:21
sbalukoffThat is to say, I don't have a strong opinion on that, but I suspect others might. But let's not wait for them, eh.18:22
crc32yea thats why I'm hesitent to moving it just yet.18:23
crc32I'm pretty much coding it in a "" kind of way18:24
sbalukoffMakes sense to me.18:27
sbalukoffObviously, if you've got a strong opinion on it, you should feel free to argue for it, eh. :)18:27
sbalukoffI would probably back you up. ;)18:27
blogancrc32: i can tell you exactly where you should stick that18:27
TrevorVDon't say anything about anyone's rear... you jerk face18:29
crc32you were saying blowgun?18:29
* crc32 slaps blogan around a bit with a large trout18:30
* blogan eats that trout18:30
crc32where do I push/shove/stick my code?18:31
crc32blogan: You know the code base fairly well whats your impression?18:32
bloganabout where to put the code that calls barbican to retrieve the blob?18:33
crc32no I was gonna let adam decide that. I'm talking more about the X509 mangler.18:33
bloganwell i would think those two should be in the relative same place18:34
crc32I'll be working with adam on the TLS lbaas/barbican cross talk but I'll let him make the religuous decisions.18:34
bloganand i would think that we will probably not be the only advanced services that will want to talk to barbican and do the parsing/mangling18:35
crc32blogan: Are you suggesting dumping the code in oslo?18:35
bloganoh no18:35
blogani was thinking in its own module in the services package, or somewhere else in neutron that they may have something similar18:36
bloganthere is the notifiers package, that nova has a module in18:38
blogani wouldn't exactly consider this only a notifier though18:38
bloganbut they are setting up the nova client with the urls, usernames and api keys18:38
bloganyeah i kinda think the notifiers package would be best, though i'm not too crazy on the name18:39
bloganyou can bring that up on the neutron ML or in the next neutron meeting and ask people their opinions, in the meantime just put it in its own module somewhere so you can easily just move it around and test it out18:40
crc32speaking of names I was gonna put it in "". ;)18:40
bloganyou should make a class caleld that and put it in a or module18:43
*** Youcef has quit IRC18:48
dougwigmy vote would be neutron/services/common/mymodulewhateveritis.py18:55
dougwigi also wouldn't stress the ML so much; people will yell if it's in the wrong place once it's in gerrit.  if you pick a semi-reasonable place, it'll be fine.18:57
*** fnaval has quit IRC19:41
*** sbfox1 has quit IRC20:39
*** sbfox has joined #openstack-lbaas20:40
*** TrevorV_ has joined #openstack-lbaas20:59
*** TrevorV_ has left #openstack-lbaas21:00
*** sbfox1 has joined #openstack-lbaas21:09
*** sbfox has quit IRC21:12
*** sbfox1 has quit IRC21:14
*** sbfox has joined #openstack-lbaas21:15
*** rolledback has joined #openstack-lbaas21:30
*** rolledback has joined #openstack-lbaas21:36
*** fnaval has joined #openstack-lbaas21:46
*** rolledback has quit IRC21:50
*** jorgem has quit IRC21:58
*** sbfox has quit IRC22:05
*** ptoohill_ has joined #openstack-lbaas22:20
*** sbfox has joined #openstack-lbaas22:39
*** mlavalle has quit IRC22:59
*** ptoohill_ has quit IRC23:18
*** fnaval has quit IRC23:47
*** ptoohill_ has joined #openstack-lbaas23:48
*** ptoohill_ has quit IRC23:49
*** ptoohill_ has joined #openstack-lbaas23:49
*** ptoohill_ has quit IRC23:50
*** fnaval_ has joined #openstack-lbaas23:51

Generated by 2.14.0 by Marius Gedminas - find it at!