| *** dlundquist has quit IRC | 00:06 | |
| *** vivek-ebay has quit IRC | 00:11 | |
| *** german_ has quit IRC | 00:25 | |
| dougwig | sbalukoff: I figured you were on vacation or something, since you were so quiet today. | 00:36 |
|---|---|---|
| dougwig | nice to see my inbox light up. | 00:36 |
| blogan | dougwig: any word on if your bp goes in soon? | 00:44 |
| dougwig | well, let's see. | 00:44 |
| dougwig | mestery: did you get a chance to peek at the driver review? would help with blogan's work. | 00:45 |
| *** woodster__ has quit IRC | 00:55 | |
| VijayB | blogan: Hi Brandon, Craig mentioned that you have been making CLI changes, but I don't see your commits on the branches in https://github.com/craigtracey/python-neutronclient.git - is there another repo you're working on? If so I think we should converge on one repo and work on it.. | 01:09 |
| blogan | VijayB: i haven't been making CLI changes | 01:10 |
| *** fnaval has joined #openstack-lbaas | 01:10 | |
| blogan | VijayB: hows everything else going? | 01:11 |
| sbalukoff | dougwig: Just busy with other things today. :) | 01:11 |
| sbalukoff | Though in a week I will be away from the office for almost a week. | 01:12 |
| dougwig | nice. | 01:12 |
| blogan | sbalukoff: campaign trail? | 01:12 |
| sbalukoff | (Adding a couple days to the 4th holiday and am going to have a really long weekend. :) ) | 01:12 |
| sbalukoff | blogan: Haha! No... | 01:12 |
| VijayB | blogan: ah ok.. I guess Craig got it mixed up.. I'll make those CLI changes in Craig's repo to test out the API.. it needs a bit of work.. | 01:12 |
| sbalukoff | My dad's on that, but that really only starts to heat up in September / October. | 01:12 |
| VijayB | the routes need to be changed a bit.. | 01:12 |
| sbalukoff | And I don't know that I'll go on the trail at all with my dad. | 01:12 |
| blogan | VijayB: which CLI changes? just to implement the new API? | 01:12 |
| ctracey | VijayB: I did not mention that brandon was making CLI changes | 01:13 |
| ctracey | i said that the CLI changes that I have had to change because the API backend changes | 01:13 |
| blogan | sbalukoff: well you could go the black sheep route | 01:13 |
| VijayB | ctracey: ah! ok, my bad - sorry for the confusion guys | 01:13 |
| dougwig | let me know if you do end up in boise, and the beer is on me (trick offer, since i know you don't drink, it'll be free!) :) | 01:13 |
| sbalukoff | blogan: Oh, I'm very much a black sheep in my family. ;) | 01:14 |
| ctracey | which routes need to be changed? | 01:14 |
| VijayB | ctracey: the /lbaas/ needs to go | 01:14 |
| blogan | VijayB: whys that? | 01:14 |
| VijayB | ctracey: we should use /listeners/ directly.. | 01:14 |
| ctracey | what is it now? this is what I was told was the route | 01:15 |
| ctracey | is this code checked into brandon's branch? | 01:15 |
| blogan | VijayB: https://<neutron endpoint>/v2.0/listeners? | 01:15 |
| blogan | is that what you mean? | 01:15 |
| VijayB | blogan: ctracey: sorry, I take that back | 01:16 |
| ctracey | so now this is a root object for v2.0? | 01:16 |
| VijayB | the lbaas must be there | 01:16 |
| blogan | lol Vijay, scaring me | 01:16 |
| ctracey | me too | 01:16 |
| VijayB | blogan: sorry :) Too many things going in my head right now | 01:16 |
| VijayB | D: | 01:16 |
| blogan | i was ready to fight that! | 01:16 |
| sbalukoff | Me too | 01:16 |
| VijayB | blogan: How are you testing your changes?? Can you please paste a sample URI? | 01:17 |
| dougwig | we are waving dead chicken heads at the platters and calling it good. | 01:17 |
| VijayB | say for lbaas-loadbalancer-list | 01:17 |
| blogan | http://localhost:9696/v2.0/lbaas/nodepools/bfc5d141-c160-48ac-8e49-4ca2d725fe32/members.json | 01:17 |
| blogan | btw i hate nodepools too | 01:17 |
| blogan | but it had to be done for now | 01:17 |
| VijayB | blogan: thx | 01:18 |
| VijayB | blogan: nodepools are confusing me :'( | 01:18 |
| blogan | VijayB: its for a lack of a better name | 01:18 |
| VijayB | blogan: ok | 01:18 |
| blogan | if anyone else has a better name be my guest, but it can't be pools | 01:18 |
| ctracey | what are nodepools now? | 01:18 |
| VijayB | blogan: yup :) | 01:18 |
| blogan | pools | 01:18 |
| blogan | ctracey: just pools | 01:18 |
| ctracey | and why not pools? | 01:19 |
| VijayB | ctracey: because neutron uses the same namespace for the routes and resources, we can't use pools because v1.0 api already uses it :( | 01:19 |
| VijayB | so we have to call it by some other name | 01:19 |
| ctracey | wait what? | 01:19 |
| ctracey | i thought his is why we are namespacing it under /lbaas | 01:19 |
| blogan | ctracey: what VijayB said but its not that we can't use it, its just that the properties of the v1.0 pool will be a union with the v2.0 pool | 01:19 |
| VijayB | ctracey: no Mark said that wouldn't make a difference.. | 01:19 |
| ctracey | why is that? | 01:20 |
| blogan | ctracey: me too but that is not the case, i have tested it and tried to get around it but it would require a massive patch in the neutron api code | 01:20 |
| VijayB | blogan: ah ok.. got it.. so when the subattributes are registered by the boilerplate code, they all go with the parent, which is pools in this case.. | 01:20 |
| ctracey | we have been doing this <thing>v2 idiom...which seems silly | 01:20 |
| ctracey | why not lb.v2.<thing> | 01:20 |
| blogan | you mean have a v2 package? | 01:21 |
| ctracey | module | 01:21 |
| ctracey | https://github.com/craigtracey/python-neutronclient/tree/objectmodel/neutronclient/neutron/v2_0/lb/v2 | 01:21 |
| ctracey | as such | 01:21 |
| dougwig | lbpool, lb_pool, pool_lb, the_fucking_deep_end_comma_yo | 01:22 |
| blogan | i believe the reason is because the modules and classes woudl be renamed back to the real names once this all gets completed | 01:22 |
| ctracey | doesnt my idiom make more sense then? | 01:22 |
| blogan | well honestly naming the module v2 or loadbalancerv2 would essentially be the same thing | 01:23 |
| blogan | the old extension module is loadbalancer.py, the old db one is loadbalancer_db.py, are you saying just name those v2.py? | 01:24 |
| ctracey | not really | 01:24 |
| ctracey | my objects are named Listener, Pool, etc | 01:24 |
| ctracey | not <thing>V2 | 01:24 |
| blogan | so you're talking about the models then? | 01:24 |
| ctracey | im talking about namespacing the whole version under a v2 module | 01:25 |
| dougwig | he's talking about namespacing with an alternate character (other than '/'). | 01:26 |
| ctracey | instead of mangling the names with v2 | 01:26 |
| ctracey | (i dont mean 'mangling' in the bad sense...the comp sci sense) | 01:26 |
| blogan | well there's no way to get around mangling the model names | 01:26 |
| blogan | because duplicate model names can't be used in the same declarative base | 01:26 |
| ctracey | correct | 01:27 |
| blogan | even if they're in a different module | 01:27 |
| ctracey | the alchemy models need to be unique (yay databases) | 01:27 |
| ctracey | however, the routes are something different, no? | 01:27 |
| blogan | i guess im still confused, sorry | 01:28 |
| blogan | the routes are defined in the extension? | 01:28 |
| blogan | minus the ? | 01:28 |
| blogan | that was a statement | 01:28 |
| ctracey | correct | 01:29 |
| blogan | so are you saying name the extension v2.py instead of loadbalancerv2.py? | 01:29 |
| ctracey | so i dont understand why we need /lbaas/nodepools vs /lbaas/pools | 01:29 |
| blogan | okay that has nothing to do with the name of hte modules or anything | 01:29 |
| blogan | thats just the dynamic extension building neutron has built in and how extensions are supposed to be created | 01:30 |
| blogan | now it would work fine if we weren't planning on having v1 and v2 running at the same time, and we could go with pools | 01:30 |
| ctracey | doesnt that feel like we are being painted into a corner only for the sake of supporting something that really isnt used? | 01:31 |
| blogan | you mean v1 isn't really being used? | 01:31 |
| ctracey | yes | 01:31 |
| blogan | well from the meeting last week it sounds like there are people running it in production... | 01:32 |
| ctracey | so now we are creating these artificial names that will live on? | 01:32 |
| blogan | oh no they won't live on | 01:32 |
| VijayB | yeah we can't assume that nobody's using it.. | 01:32 |
| blogan | they will rename | 01:32 |
| ctracey | i didnt say we cant assume no one is using it...that is a given | 01:32 |
| ctracey | this just feels very dirty from an API perspective | 01:32 |
| blogan | i was going ot talk about the nodepools thing with mark soon and see what his htoughts are because we don't want to have to rename a resource | 01:32 |
| blogan | a route | 01:32 |
| ctracey | API's should bump versions - not routes | 01:33 |
| blogan | i totally agree | 01:33 |
| blogan | but since this is an extension, and we are bound by neutron we can't really do that | 01:33 |
| blogan | even if we went with /lbaas/v2.0/pools it'd still be the same problem | 01:33 |
| ctracey | i dont know the code well enough | 01:33 |
| blogan | lol i still dont either but i tried to get around this all night the other night | 01:34 |
| ctracey | but that seems like a restriction that will be the "gift" that keeps on giving | 01:34 |
| sbalukoff | Burn it to the ground and start over. | 01:34 |
| sbalukoff | ;) | 01:34 |
| ctracey | regardless...back to the original question... | 01:34 |
| ctracey | i can modify the routes easily in my code | 01:35 |
| blogan | i'll talk to mark and make sure we do not have to use a different name than pools by the time this gets merged in, but for now, for the sake of getting something done it is what i used | 01:35 |
| ctracey | i have been spinning some cycles on getting a test env going | 01:35 |
| sbalukoff | Seriously, though-- given the restrictions Brandon is talking about, I don't see a better way to do this. :/ | 01:35 |
| ctracey | all im sayin is that we shoudl verify that this is a strict restriction | 01:35 |
| ctracey | imho, changing an established api (as would be case with a rename) is a no-no | 01:36 |
| blogan | sbalukoff: all that is really needed to fix it is a patch that puts the attr_map (which is the expected body) in an additional dictionaly with the key being prefix, but that would be a huge change and probbaly a fight in itself to get in | 01:36 |
| sbalukoff | blogan: In other words, definitely not before Juno. | 01:37 |
| *** sbfox has joined #openstack-lbaas | 01:37 | |
| blogan | exactly | 01:37 |
| blogan | but like i said, i don't want this going in with the route nodepool either, but its just a stopgap for now until i can talk to mark about his thoughts on it | 01:38 |
| blogan | if its changed to nodepools, or any bastardized name other than pools, then we'd have a real problem | 01:38 |
| ctracey | well this extends into all of the other like-named objects, doesnt it? | 01:39 |
| ctracey | like members? | 01:39 |
| ctracey | healthmons? | 01:39 |
| blogan | health monitors are not an issue since no attributes have been removed | 01:40 |
| blogan | and same with members for the v2 api | 01:40 |
| ctracey | ok now i am more confused | 01:41 |
| blogan | lol i know | 01:41 |
| blogan | its tough to explain | 01:41 |
| blogan | let me try to better explain it | 01:41 |
| ctracey | if a field is removed, isnt that in our favor? | 01:41 |
| ctracey | just blackhole it | 01:42 |
| blogan | when you send in a create call to v1/pools you are specifying lb_method and subnet_id, if you issue a create call to v2/pools you are specifying lb_algorithm and not a subnet_id since that is removed from the v2 pool | 01:42 |
| ctracey | correct | 01:43 |
| blogan | with v1 and v2 active, the api is expecting lb_method, lb_algorithm, and subnet_id for both v1/pools and v2/pools | 01:43 |
| ctracey | isnt this the point of the alternate route? | 01:44 |
| blogan | i thought so too | 01:44 |
| blogan | but since the neutron api bundles it all expected attributes of a pool into dictionary['pools'] and if 'pools' key already exists it just does a dictionary.update() | 01:45 |
| ctracey | ok | 01:46 |
| ctracey | i think i see the issue | 01:46 |
| ctracey | so here is how I handled it... | 01:46 |
| *** sbfox has quit IRC | 01:46 | |
| ctracey | https://github.com/craigtracey/python-neutronclient/blob/objectmodel/neutronclient/neutron/v2_0/lb/v2/listener.py#L28 | 01:46 |
| ctracey | internally the resource is namespaced | 01:46 |
| blogan | now we can blackhole the lb_algorithm and v1 api, and the lb_method and subnet_id in the v2 api, but i want to make sure with mark that is okay | 01:46 |
| ctracey | but from the front-end it would still be /lbaas | 01:47 |
| ctracey | er /lbaas/listeners | 01:47 |
| ctracey | at the end of the day i think it would be *really* poor form to change the API | 01:47 |
| ctracey | (read: this is the kind of shit that makes developers crap on openstack) | 01:48 |
| blogan | and i promise I'd do that if I didn't have to change the neutron api code | 01:48 |
| blogan | actually i'd do dictionary['lbaas']['pools'] instead of just dictionary['pools'] | 01:49 |
| blogan | bc thats how they store the attribute map | 01:49 |
| ctracey | where is this dictionary? | 01:50 |
| blogan | one sec | 01:50 |
| blogan | https://github.com/openstack/neutron/blob/master/neutron/api/extensions.py | 01:53 |
| blogan | the loop at line 477 | 01:53 |
| VijayB | why in the world am I hitting "DEBUG routes.middleware [-] No route matched for GET /lbaas/nodepools.json from (pid=21602) __call__ /usr/lib/python2.7/dist-packages/routes/middleware.py:97" when trying curl -i http://192.168.45.111:9696/v2.0/lbaas/nodepools.json -X GET -H "X-Auth-Token: $TOK" -H "Content-Type: application/json" -H "Accept: application/json" -H "User-Agent: python-neutronclient" ?? | 01:54 |
| blogan | the attr_map is what contians the attributes expected in the resource body | 01:54 |
| blogan | VijayB: have you enabled the config with the v2 stuff | 01:54 |
| VijayB | blogan: I thought we did that during the hackathon right? Let me check neutron.conf again... | 01:55 |
| blogan | ./stack.sh will clear that out | 01:55 |
| blogan | so everytime you run ./stack.sh it'll reset your config | 01:55 |
| VijayB | oh | 01:55 |
| VijayB | I don't think I ran ./stack.sh again though.. just ./rejoin_stack.sh.. does that do that too? | 01:56 |
| ctracey | sounds like you didnt python setup.py install | 01:56 |
| ctracey | pyc for the win | 01:57 |
| blogan | i dont think so but rejoin_stack.sh isn't reliale for me | 01:57 |
| *** sbfox has joined #openstack-lbaas | 01:57 | |
| ctracey | rejoin stack just drops you into the screens afaik | 01:57 |
| VijayB | ctracey: hmm.. | 01:57 |
| blogan | it'll restart the services as well and drop you into the screens | 01:57 |
| blogan | or start the services that aren't already started | 01:58 |
| VijayB | blogan: yeah it starts them up if they're down.. | 01:58 |
| ctracey | yeah...thats what i mean | 01:58 |
| ctracey | (re)starting a python service does not install new code | 01:58 |
| ctracey | assuming you did a git pull or something? | 01:59 |
| VijayB | still no luck - I must be missing something in neutron.conf... | 01:59 |
| VijayB | blogan: can you please share your neutron.conf, I'll run a quick check against mine.. | 02:00 |
| VijayB | have put up mine at https://etherpad.openstack.org/p/neutronconf1 | 02:04 |
| blogan | VijayB: add this to the end of the service_plugins line under the DEFAULT section | 02:10 |
| blogan | ,neutron.services.loadbalancer.plugin.LoadBalancerPluginv2 | 02:10 |
| VijayB | blogan: ok | 02:11 |
| ctracey | VijayB: https://gist.github.com/brandonlogan/5d28ee177a0dc917289d | 02:11 |
| blogan | lol thanks ctracey | 02:11 |
| VijayB | blogan, ctracey: awesome, thanks a ton! :) | 02:12 |
| VijayB | Now I hit newer issues, which is a lot more relief :p | 02:12 |
| blogan | actually ptoohill wrote a script to do it automatically but it keeps appending to the service_plugins right now so I or him need to fix it | 02:12 |
| *** TrevorV_ has joined #openstack-lbaas | 02:12 | |
| blogan | my bash/sed/awk skills suck | 02:13 |
| VijayB | blogan: ok :) btw, what's ptoohill's real name?? Sorry for asking, I'm catching up on everyone's irc handles | 02:13 |
| ctracey | fire it my way | 02:13 |
| ctracey | i can get it fixed | 02:13 |
| blogan | phil | 02:13 |
| TrevorV_ | VijayB: Its "phillip toohill" | 02:14 |
| VijayB | TrevorV_: cool, thanks! :) | 02:14 |
| TrevorV_ | np | 02:14 |
| blogan | alright ctracey i give up https://gist.github.com/brandonlogan/9f9575d57c1d789a8da7 | 02:22 |
| *** sbfox has quit IRC | 02:26 | |
| VijayB | blogan, ctracey: Finally! I can successfully see no lbaas listeners!! :D | 02:27 |
| VijayB | Thx a ton guys! | 02:27 |
| ctracey | what was issue? | 02:27 |
| ctracey | back - looking at sed | 02:27 |
| VijayB | I had some issues with the db - stale tables... manually deleted them all and reran the upgrade --delta 1 using Brandon's new alembic script.. | 02:27 |
| VijayB | I'll call it a day for now.. catch up with y'all tomorrow! | 02:28 |
| VijayB | Have a good night everyone | 02:29 |
| *** VijayB has quit IRC | 02:31 | |
| *** openstackgerrit has quit IRC | 02:31 | |
| *** vivek-ebay has joined #openstack-lbaas | 02:41 | |
| *** TrevorV_ has quit IRC | 02:43 | |
| *** sbfox has joined #openstack-lbaas | 03:17 | |
| *** sbfox has quit IRC | 03:35 | |
| *** german_ has joined #openstack-lbaas | 03:44 | |
| *** vjay2 has joined #openstack-lbaas | 03:55 | |
| *** german_ has quit IRC | 04:36 | |
| *** ptoohill_ has joined #openstack-lbaas | 05:04 | |
| *** fnaval has quit IRC | 05:05 | |
| *** vivek-ebay has quit IRC | 05:05 | |
| *** blogan is now known as zz_blogan | 05:09 | |
| *** markmcclain has joined #openstack-lbaas | 05:17 | |
| *** markmcclain1 has joined #openstack-lbaas | 05:18 | |
| *** markmcclain has quit IRC | 05:22 | |
| *** ptoohill_ has quit IRC | 05:23 | |
| *** markmcclain1 has quit IRC | 05:39 | |
| *** sbfox has joined #openstack-lbaas | 05:47 | |
| *** sbfox has quit IRC | 06:11 | |
| *** sbfox has joined #openstack-lbaas | 06:34 | |
| *** vjay2 has quit IRC | 06:44 | |
| *** vjay2 has joined #openstack-lbaas | 06:49 | |
| *** sbfox has quit IRC | 07:38 | |
| *** evgenyf has joined #openstack-lbaas | 08:00 | |
| *** openstackgerrit has joined #openstack-lbaas | 08:52 | |
| *** evgenyf has quit IRC | 09:30 | |
| *** evgenyf has joined #openstack-lbaas | 09:45 | |
| openstackgerrit | Stephen Balukoff proposed a change to stackforge/octavia: Populate repository with common initial files. https://review.openstack.org/102482 | 10:00 |
| *** rm_work is now known as rm_work|away | 10:23 | |
| openstackgerrit | Stephen Balukoff proposed a change to stackforge/octavia: Populate repository with common initial files. https://review.openstack.org/102482 | 10:25 |
| *** woodster__ has joined #openstack-lbaas | 12:52 | |
| *** enikanorov_ has quit IRC | 13:00 | |
| *** enikanorov_ has joined #openstack-lbaas | 13:00 | |
| *** fnaval has joined #openstack-lbaas | 13:21 | |
| *** markmcclain has joined #openstack-lbaas | 13:36 | |
| *** markmcclain1 has joined #openstack-lbaas | 13:37 | |
| *** markmcclain1 has quit IRC | 13:38 | |
| *** markmcclain has quit IRC | 13:40 | |
| *** ptoohill_ has joined #openstack-lbaas | 13:46 | |
| *** ptoohill_ has quit IRC | 13:52 | |
| *** fnaval has quit IRC | 13:57 | |
| *** rolledback has joined #openstack-lbaas | 14:13 | |
| *** rolledback has quit IRC | 14:17 | |
| *** rolledback has joined #openstack-lbaas | 14:18 | |
| *** vivek-ebay has joined #openstack-lbaas | 14:33 | |
| *** fnaval has joined #openstack-lbaas | 14:38 | |
| *** fnaval has quit IRC | 14:39 | |
| *** fnaval has joined #openstack-lbaas | 14:39 | |
| *** vivek-ebay has quit IRC | 14:42 | |
| *** markmcclain has joined #openstack-lbaas | 14:48 | |
| *** vivek-eb_ has joined #openstack-lbaas | 14:50 | |
| *** sbfox has joined #openstack-lbaas | 14:53 | |
| *** sbfox has quit IRC | 14:54 | |
| *** TrevorV_ has joined #openstack-lbaas | 15:06 | |
| *** evgenyf has quit IRC | 15:10 | |
| *** rolledback has quit IRC | 15:10 | |
| *** fnaval has quit IRC | 15:12 | |
| *** fnaval has joined #openstack-lbaas | 15:13 | |
| *** zz_blogan is now known as blogan | 15:14 | |
| *** fnaval has quit IRC | 15:17 | |
| *** rolledback has joined #openstack-lbaas | 15:21 | |
| *** vivek-eb_ has quit IRC | 15:21 | |
| *** rolledback has quit IRC | 15:23 | |
| *** rolledback has joined #openstack-lbaas | 15:34 | |
| *** fnaval has joined #openstack-lbaas | 15:37 | |
| *** sbfox has joined #openstack-lbaas | 16:05 | |
| *** crc32 has joined #openstack-lbaas | 16:27 | |
| *** jorgem has joined #openstack-lbaas | 16:34 | |
| *** crc32 has quit IRC | 16:40 | |
| *** jorgem1 has joined #openstack-lbaas | 16:44 | |
| *** jorgem has quit IRC | 16:44 | |
| *** crc32 has joined #openstack-lbaas | 16:48 | |
| *** vivek-ebay has joined #openstack-lbaas | 16:54 | |
| dougwig | good morning | 16:55 |
| sbalukoff | Morning! | 16:57 |
| blogan | afternoon! | 17:02 |
| TrevorV_ | blogan actually waited so he could say that | 17:04 |
| blogan | yes | 17:04 |
| sbalukoff | Heh! | 17:04 |
| ctracey | All: please come to the webex with thoughts on octavia-core/-ptl | 17:11 |
| ctracey | we need some people there in order to get stackforge code accepted via gerrit | 17:11 |
| TrevorV_ | ctracey: right now | 17:13 |
| TrevorV_ | ? | 17:13 |
| blogan | today at 3pm central? crap I'm not going to be able to make it today | 17:14 |
| TrevorV_ | Was there someone other than sbalukoff that wanted to be PTL for Octavia? | 17:15 |
| blogan | still would need to decide on core though | 17:15 |
| TrevorV_ | Oh, right. | 17:15 |
| *** rolledback has quit IRC | 17:17 | |
| *** vjay2 has quit IRC | 17:20 | |
| *** VijayB has joined #openstack-lbaas | 17:21 | |
| ctracey | yeah - we def need to figure this out | 17:22 |
| sbalukoff | Yep. | 17:22 |
| ctracey | i would rather keep the momentum behind things moving | 17:22 |
| sbalukoff | blogan: Can you work with your team to make sure your views on this are represented? | 17:22 |
| sbalukoff | Yep. | 17:22 |
| blogan | is that a requirement for stackforge gerrit acceptance? | 17:22 |
| ctracey | not being able to merge code will be a drag | 17:22 |
| blogan | sbalukoff: yes | 17:22 |
| sbalukoff | blogan: It appears to be. | 17:22 |
| ctracey | is what a requirement? | 17:22 |
| blogan | so why were people saying not to decide on that in the thread on the ML? | 17:22 |
| ctracey | only core or ptl can merge code | 17:23 |
| sbalukoff | blogan: Because the mailing list is public and everyone's an armchair project manager. | 17:23 |
| ctracey | hahaha | 17:23 |
| blogan | lol | 17:23 |
| ctracey | https://bugs.launchpad.net/openstack-ci/+bug/1331034 | 17:23 |
| sbalukoff | I was entertained that the only people shouting down the idea were people who have absolutely nothing to do with neutron lbaas or octavia thus far, and hence don't know the history going into getting octavia to be a stackforge project. | 17:23 |
| *** VijayB has quit IRC | 17:24 | |
| blogan | well i think they were speaking from experience in starting their own project out | 17:24 |
| ctracey | http://ci.openstack.org/stackforge.html#request-an-initial-gerrit-core-group-member | 17:24 |
| *** VijayB_ has joined #openstack-lbaas | 17:24 | |
| ctracey | i have started a project on stackforge | 17:24 |
| sbalukoff | blogan: You're more charitable than I am. ;) | 17:25 |
| ctracey | and we didnt have issues with this | 17:25 |
| ctracey | i dont think there is any concept that this work has been happening for 6+ months | 17:25 |
| blogan | sbalukoff: yes i'm more of a fence rider | 17:25 |
| ctracey | and there are already interested and contributing parties | 17:25 |
| blogan | ctracey: you're probably right, im not arguing against it i was just wondering if that would bite us in the ass | 17:26 |
| ctracey | i cannot see how it would | 17:26 |
| ctracey | i think they are trying to circumvent infighting | 17:26 |
| ctracey | which i do not see a problem with | 17:26 |
| blogan | thats why they call it getting bit in the ass, you don't see it coming | 17:26 |
| *** ptoohill_ has joined #openstack-lbaas | 17:26 | |
| blogan | i dont see a problem with it | 17:27 |
| sbalukoff | What are you talking about, I totally have eyeballs near my asshole. | 17:27 |
| sbalukoff | Oh wait, maybe that's my mouth hole. | 17:27 |
| blogan | i would just like the core people to be very active | 17:27 |
| ctracey | i seriously dont care if there is a single member of core from each company | 17:27 |
| ctracey | blogan: that is the trick | 17:27 |
| sbalukoff | blogan: +1 | 17:27 |
| ctracey | we use reviews as the measure | 17:27 |
| ctracey | on the other project(s) i am on | 17:28 |
| blogan | yeah thats what is normally done i believe | 17:28 |
| ctracey | you review you get street cred | 17:28 |
| *** ptoohill_ has quit IRC | 17:28 | |
| blogan | but since we don't have that one from each active company wouldn't be a bad starting place | 17:28 |
| blogan | i need to learn to use commas | 17:29 |
| *** ptoohill_ has joined #openstack-lbaas | 17:29 | |
| *** ptoohill_ has quit IRC | 17:29 | |
| ctracey | we should do that with a finite time for a "real" election | 17:29 |
| ctracey | in order to bootstrap | 17:29 |
| sbalukoff | That's actually what I would suggest, too! | 17:30 |
| blogan | okay | 17:30 |
| blogan | sounds good to me | 17:30 |
| ctracey | lunching: bbiab | 17:30 |
| dougwig | i'd say skip the mailing list on this from here on out, and just set it up. i didn't hear any other suggestions other than malevolent dictator... umm, i mean stephen, so i'd call it done. and i don't think stackforge requires 2 cores to sign off, btw. i've had stack forge reviews go in with just one. | 17:34 |
| *** sbfox has quit IRC | 17:36 | |
| sbalukoff | I'd also be fine with more than one person from Rackspace in particular being a core since y'all seem to have the most active developers of any of our companies on this project, eh. | 17:37 |
| dougwig | remember mark's advice to keep the core group small. | 17:37 |
| TrevorV_ | What is small dougwig? 4? | 17:38 |
| dougwig | i'd think 3-4, yeah. | 17:38 |
| TrevorV_ | Then I don't think multiple core on RS team would be appropriate... That would make half the core on our time. | 17:38 |
| TrevorV_ | Then again blogan is putting in 16 hour days, so he's got a few time-zones covered :P | 17:39 |
| sbalukoff | Haha! | 17:40 |
| sbalukoff | Ok, that makes sense. | 17:40 |
| *** sbfox has joined #openstack-lbaas | 17:40 | |
| TrevorV_ | If "small" was 6, maybe I'd say 2 could be filled by RS, but I wouldn't say 3-4 ya know? | 17:41 |
| blogan | i think one per company will be fine and if after some time we have enough data to show who is the most active reviewer/contributors then multiple from each company would be appropriate | 17:41 |
| TrevorV_ | blogan I think that makes the list like 8 people large then, right? | 17:41 |
| sbalukoff | Oh, in other news, I heard back from the USPTO office about the "octavia" trademark. They've got my application, and in approximately 3 months, they'll assign it to someone on their team to glance over real quick! | 17:42 |
| sbalukoff | Government efficiency! Huzzah! | 17:42 |
| blogan | 3 months... | 17:42 |
| blogan | if only i could get away with taking me 3 months to look over a document at my job | 17:42 |
| TrevorV_ | sbalukoff: is it going to be YOUR trademark, or is it going to be under "Openstack Foundation"? | 17:42 |
| sbalukoff | From what I recall this is also the usual timeframe. XD | 17:42 |
| dougwig | hell, i usually don't even get 3 minutes. | 17:42 |
| sbalukoff | TrevorV: Intially, I'll be the owner, but once it's granted, I'll be looking into getting it transferred to the OpenStack Foundation. My guess is the OpenStack Foundation isn't going to be interested in maintaining it until octavia is incubated, at least. | 17:43 |
| *** rolledback has joined #openstack-lbaas | 17:43 | |
| TrevorV_ | yeah that makes sense sbalukoff. | 17:44 |
| TrevorV_ | Just wondering | 17:44 |
| sbalukoff | Plus, I'm totally planning on screwing you all over. | 17:44 |
| sbalukoff | ;) | 17:44 |
| TrevorV_ | sbalukoff: I mean you do want that PTL position too, it wouldn't e impossible :D | 17:44 |
| TrevorV_ | be impossible*** | 17:44 |
| sbalukoff | I'm subtly subverting the system right under all y'all's noses. | 17:45 |
| TrevorV_ | My nose isn't all that big, I'll figure you out soon enough o_0 | 17:45 |
| sbalukoff | Hehe! | 17:45 |
| dougwig | isn't that what vendors are supposed to do? | 17:45 |
| sbalukoff | Anyone here heard of Vernon Supreme? | 17:46 |
| sbalukoff | Er... | 17:46 |
| sbalukoff | Vermin Supreme? | 17:46 |
| TrevorV_ | Is that some road-kill pizza type? | 17:46 |
| sbalukoff | He's a presidential candidate. And he's hillarious. | 17:46 |
| TrevorV_ | Hmm. | 17:47 |
| sbalukoff | Anyway, that's the kind of leadership you can expect from me as your supreme ruler. Just sayin' | 17:48 |
| dougwig | before we talk about cores, i'd like to hear thoughts on whether octavia is purely an opensource ref lbaas using haproxy, or whether it's really more of an lbaas/servicevm framework for using soft backends of choice to do at-scale load balancing with nova? (i know stephen's answer.) also, when we spin lbaas out of neutron, will it be yet another incubated | 17:48 |
| dougwig | project, or will we suck it into octavia? | 17:48 |
| sbalukoff | dougwig: Let me amend my answer: Initially, I want it to only have to worry about an haproxy software appliance-based back end (ie. so we get something functional as a reference implementation at the very least). However, I would not rule out being able to use other soft backends, as long as they comply to standards we set in the project. | 17:50 |
| sbalukoff | Not trying to be exclusionary, just trying to be practical. | 17:50 |
| sbalukoff | And hopefully solve some of the (organizational?) problems that have prevented Neutron LBaaS from supporting anything resembling an 'advanced' feature for years now. | 17:51 |
| *** fnaval has quit IRC | 17:52 | |
| sbalukoff | On the spinning lbaas out of neutron: Where it ends up is something I don't have strong feelings about either way. If it gets sucked into Octavia, then this is by definition an expansion of the Octavia project scope, and thus we will need to alter some policies about how Octavia is run to make this friendly to all kinds of LBaaS back-ends. | 17:53 |
| blogan | well I have assume octavia is a purely openstack load balancing backend, as in it consumes other openstack services and uses a soft backend to create an HA scalable operator load balancing solution | 17:53 |
| dougwig | i don't have any issues with haproxy being the first (and only) implementation out of the gate. but it influences design/architecture somewhat if the long-term plan is to be more flexible, and it certainly affects who should be involved. i look at some of the blueprints out there, for adding lbaas lvx or nginx, and those should clearly be soft appliance | 17:53 |
| dougwig | based, not sitting on the neutron node. and every vendor has a more feature-rich soft appliance, which flavors can expose, which can leverage the same thing. | 17:53 |
| sbalukoff | blogan: There's a possibility of using a 3rd party VM to replace / fill the role of the "Octavia VM" in my designs thus far. | 17:54 |
| sbalukoff | But it would involve another driver layer in the controllers. | 17:54 |
| sbalukoff | (Not a bad thing / show stopper. We should probably have that abstracted out anyway.) | 17:54 |
| blogan | yeah I dont care whether they merge into one, though I think if they are both openstack it'd be better if they were separate so that teh lbaas service can focus on user facing API and octavia can focus on what it needs to focus on | 17:54 |
| sbalukoff | Er... sorry, I mean, have a good abstraction layer there anyway. | 17:54 |
| blogan | sbalukoff: i'll ride the fence on the 3rd party VM as well, i originally was thinking that would be a nice to have but I also see the argument that this should only consume openstack services | 17:55 |
| sbalukoff | blogan: Yep, there are good reasons to keep LBaaS separate from Octavia. | 17:56 |
| dougwig | blogan: haproxy isn't an openstack service. | 17:56 |
| blogan | dougwig: good points are not allowed | 17:57 |
| dougwig | can we make that one of our project rules? | 17:57 |
| sbalukoff | Already in there. | 17:57 |
| blogan | i will invoke that rule every time i speak up about something | 17:58 |
| sbalukoff | Along with "every third sentence must reference fish in some way." | 17:58 |
| dougwig | eh, i was pondering stephen's mailing list email over the weekend, and what octavia was to *me*, and really it was glue between neutron, lbaas, serviceVM, and nova. and writing/duplicating the parts of those pieces that weren't mature enough for what we wanted. what actually shuffles the bytes around was secondary in my mind. though that's just my | 17:59 |
| dougwig | opinion. | 17:59 |
| *** sbfox has quit IRC | 18:01 | |
| *** sbfox has joined #openstack-lbaas | 18:02 | |
| *** rm_work|away is now known as rm_work | 18:02 | |
| * VijayB_ in_meeting | 18:03 | |
| dougwig | bringing this back to the question of who should be core, the answers to the question of what is octavia goes directly to that. if it wants to be a purely open-source operator love-fest, that's cool, and i'm probably not a useful core reviewer. if it's more along the lines of what i said above, then we should have at least one vendor in there (and of | 18:04 |
| dougwig | course I think that should be me. :-) ). no hard feelings either way, but that's where my head is at. | 18:04 |
| blogan | dougwig: are you saying you believe it should consume neutron, serviceVM, and nova, and be a backend for lbaas? but you'd like the actual load balancing application to be configurable? | 18:06 |
| TrevorV_ | I guess I'm extremely confused about what Octavia is. Does it extend higher than just a plugin? | 18:07 |
| TrevorV_ | (Am I mixing up terminology??) | 18:07 |
| dougwig | yep. it's just a serviceVM, which could be running haproxy, nginx, f5, radware... mix in flavors, and those softVMs could expose the same features as their hardware cousins. | 18:07 |
| sbalukoff | So, again, if I'm setting the stage, then it'll be a pure opensource love fest until the 1.0 design is realized at least. But I would also advocate not painting ourselves into a corner with that philosophy: That is, we should probably make a driver layer abstraction that the controllers use for talking to VMs. | 18:07 |
| blogan | TrevorV_: it would essentially be another vendor | 18:07 |
| dougwig | sbalukoff: i'm talking philosophy, not 1.0 | 18:08 |
| rm_work | ok, so, just caught up | 18:08 |
| sbalukoff | But! | 18:08 |
| TrevorV_ | Except a vendor that has specific other vendor options/configurations? That doesn't really make sense to me I guess.... | 18:08 |
| blogan | sbalukoff: thats what I would go for as well, focus on that for the first iteration | 18:08 |
| sbalukoff | this also means we may feel free to make 'haproxy' based assumptions in how things work. | 18:08 |
| rm_work | my thought was that Octavia was just a backend -- and that haproxy was our initial provider, but that we could be modular and use nginx or others… but all within the context of novas service VMs | 18:08 |
| dougwig | let me be clear, i don't think any of us have *ever* suggested anything but haproxy for 1.0 | 18:09 |
| blogan | Vendor applications use tons of 3rd party (vendor) software/libraries/systems to accomplish what they need to be done | 18:09 |
| sbalukoff | Part of the problem I see with the current Neutron LBaaS philosophy / methodology is that they have insisted that every back-end support every feature... | 18:09 |
| rm_work | Octavia would never wrap other services like radware/A10/F5 etc the way neutron-lbaas does today | 18:09 |
| sbalukoff | So what's really going on is we're appealing to the least common denominator with that project, which doesn't buy us a rich feature set. | 18:09 |
| rm_work | and I don't see Octavia ever doing what neutron-lbaas does | 18:09 |
| sbalukoff | I DO NOT want those same kinds of restrictions to inhibit octavia. | 18:09 |
| rm_work | which is actually why i was confused about Octavia going into openstack as a top-level project -- i thought neutron-lbaas would be the OpenStack project when it split off | 18:10 |
| rm_work | can Neutron-lbaas split off, become OpenStack: LBaaS, and live side-by-side with OpenStack: Octavia?? | 18:11 |
| dougwig | sbalukoff: that's a silly restriction that we're busy killing at the lbaas level. i think the real concern with vendors in octavia would be it getting derailed. and rm_work: why not? it's just meta-data passing back and forth during configuration as far as neutron or octavia is concerned. | 18:11 |
| blogan | rm_work: it was never suggested Octavia does the user facing API when it was suggested it become a top level openstack project | 18:11 |
| sbalukoff | rm_work: I would say "yes" | 18:11 |
| rm_work | blogan: right | 18:11 |
| rm_work | just making sure that's the "end goal" vision | 18:12 |
| rm_work | is two new openstack projects -- LBaaS and Octavia | 18:12 |
| sbalukoff | dougwig: That's certainly a concern, especailly early in the project. | 18:12 |
| rm_work | where LBaaS is the API and Octavia is still just a backend | 18:12 |
| sbalukoff | rm_work: I like that especially because it means that each project has a lot of autonomy. | 18:12 |
| dougwig | rm_work: there is also decent chance, if serviceVM ever gets off the ground, that octavia becomes obsolete. | 18:12 |
| sbalukoff | dougwig: Based on what I've read about serviceVM, I would say "fat chance" ;) | 18:13 |
| rm_work | I agree, it's just a little weird to have two top-level projects like that, isn't it? maybe not | 18:13 |
| dougwig | sbalukoff: most of them would say the same thing about octavia. :) | 18:13 |
| sbalukoff | But, eh... I suppose it's possible they'll shoot for making a scalable back-end. | 18:13 |
| rm_work | i guess maybe people thought the same thing about nova and neutron | 18:13 |
| blogan | i'm up for the challenge for any and all | 18:14 |
| dougwig | pulling up a bit, i don't think i'm suggesting anything revolutionary. the "scary" bit is the concern that involving vendors will fuck things up, right? | 18:14 |
| blogan | i think vendors can provide a lot of valuable knowledge | 18:14 |
| dougwig | and they can also kill a scalable LB in favor of hardware, and they can offer conflicting requirements that grind things to a halt, and they can just not be in tune with operators. | 18:15 |
| dougwig | let's call out the pros and cons. | 18:15 |
| rm_work | i mean, we're not talking about including Vendors underneat Octavia, are we? | 18:15 |
| blogan | dougwig is a filthy vendor | 18:15 |
| dougwig | rm_work: yep. scroll up to my comments at 10:48 (pacific time) | 18:16 |
| rm_work | just other sevice-vm related options like nginx/haproxy (not what I would call "vendors" in the same sense) | 18:16 |
| rm_work | yeah I am trying to parse all of this | 18:16 |
| blogan | rm_work: i think thats what dougwig is saying, that if its not going to include vendors then he doesn't think he'd be a good fit as a major contributor | 18:16 |
| rm_work | ok, well then my opinion is a clear "no, vendors should not live underneath Octavia" | 18:16 |
| dougwig | rm_work: every LB vendor has a soft appliance, that can do the same thing haproxy or nginx can. | 18:16 |
| rm_work | dougwig: vendors have soft-appliances that live in VMs? | 18:17 |
| rm_work | i mean, in THAT sense, sure, we could include those possibly in the future? | 18:17 |
| sbalukoff | rm_work: Apparently! | 18:17 |
| rm_work | That is news to me | 18:17 |
| sbalukoff | rm_work: Yep, that's what I've been saying. "After version 1.0" | 18:17 |
| rm_work | does F5 have a software LB? lol | 18:17 |
| dougwig | rm_work: example: https://aws.amazon.com/marketplace/pp/B00F3ISXXU/ref=srh_res_product_title?ie=UTF8&sr=0-5&qid=1403720259271 | 18:17 |
| dougwig | rm_work: yep, f5 sure does. | 18:17 |
| blogan | i'd be fine with vendors being able to add their own plugins, as long as the open source soft appliance is the main focus | 18:17 |
| rm_work | yeah, making the LB software that lives in the VM be pluggable makes sense | 18:18 |
| blogan | if its pluggable then nothign stops a vendor from making their own plugin | 18:18 |
| rm_work | i just didn't want hardware vendors under Octavia, because that would be really counter-productive / complicated | 18:18 |
| dougwig | rm_work: yeah, that wouldn't make any sense. | 18:18 |
| rm_work | ok | 18:18 |
| rm_work | then i guess we're on the same page now | 18:18 |
| rm_work | I need to adjust my view of "vendors" to remove the hard-link i had in my brain to "hardware" | 18:19 |
| blogan | dougwig: would you be able to contribute to octavia if it was mainly focused on opensource everything up to 1.0? or would you wait until after that? | 18:19 |
| sbalukoff | KEeping in mind that 1.0 might be a ways out. | 18:20 |
| sbalukoff | (though we hope it won't be.) | 18:20 |
| dougwig | yes, as long as the long-term strategy isn't a lock-out (which isn't any more fun than a lock-in, IMO.) | 18:20 |
| sbalukoff | HAHA | 18:20 |
| sbalukoff | Yes, I think the long-term strategy / philosophy is "no lock-outs" | 18:20 |
| sbalukoff | I'm thinking this decision should become part of a general purpose "Octavia Design Philosophy" document... | 18:21 |
| sbalukoff | Hmmmm.... | 18:21 |
| sbalukoff | I'll write something up, unless y'all object. | 18:21 |
| blogan | an octavia constitution | 18:21 |
| sbalukoff | I like it | 18:21 |
| dougwig | heh. | 18:21 |
| blogan | first amendment: good points are not allowed | 18:22 |
| rm_work | well, i guess we'll see how long that lives. Google's "don't be evil" lasted at least 8 or so years I think | 18:22 |
| sbalukoff | Second: Every third sentence must reference fish. | 18:22 |
| sbalukoff | rm_work: HAHA! | 18:22 |
| blogan | octavia shoudl strive for being the least evil | 18:23 |
| *** vjay2 has joined #openstack-lbaas | 18:23 | |
| sbalukoff | ...with the clear acknowledgement that it will be at least a little evil. | 18:24 |
| dougwig | i might come out of this with schizophrenia, since i was just told that i need to help with serviceVM, and i think they're sending me over to the NFV meetup the first week of july. three groups, three servicevm agendas. | 18:24 |
| dougwig | sbalukoff: making you the PTL wasn't obvious enough on that point? | 18:24 |
| sbalukoff | dougwig: Just wanted to be transparent. :) | 18:25 |
| blogan | he's not the PTL yet | 18:25 |
| dougwig | we know what we're getting. and a loudmouth operator is perfect for the first PTL. | 18:25 |
| dougwig | ooh, a second candidate emerges. | 18:25 |
| blogan | dont inflate his ponytail | 18:25 |
| sbalukoff | HAHA! | 18:25 |
| TrevorV_ | blogan: +1 | 18:25 |
| sbalukoff | I had someone on the plane home complain about my ponytail. | 18:26 |
| blogan | were you draping it over the back of your seat? | 18:26 |
| dougwig | what was the nature of the complaint? | 18:26 |
| sbalukoff | I thought he was serious and responded accordingly and only too late noticed that he had one too. | 18:26 |
| TrevorV_ | Were you draping it over the back of your seat? | 18:26 |
| blogan | whoa TrevorV_ exact wording | 18:26 |
| TrevorV_ | .... | 18:26 |
| sbalukoff | dougwig: Something like 'I think all them damned hippies need to get a haircut' | 18:26 |
| TrevorV_ | I'm kinda... a little... upset about that. | 18:26 |
| TrevorV_ | blogan: get out of my head. | 18:27 |
| dougwig | sbalukoff: well, they do. but. it's a free country. | 18:27 |
| sbalukoff | I think he was making a self-deprecating joke. But in my defense, I'm a complete boob when it comes to snappy come-backs. | 18:27 |
| sbalukoff | I thought of at least 10 in the 5 minutes following the encounter. | 18:28 |
| sbalukoff | Also, his ponytail was tiny and only visible from the back. | 18:28 |
| sbalukoff | Unlike my glorious locks. | 18:28 |
| dougwig | your glorious locks literally ended all conversation. | 18:37 |
| TrevorV_ | Someone else noticed that? | 18:38 |
| TrevorV_ | I wasn't going to be the one to message first either :D | 18:38 |
| sbalukoff | Haha! | 18:39 |
| sbalukoff | Ok, BBIAB. gonna go foodening before the meeting. | 18:41 |
| *** german__ has joined #openstack-lbaas | 18:54 | |
| sbfox | Hey lbaas team, I have a problem with the sqlalchemy upgrade path in icehouse. Can someone give me some help? | 18:56 |
| sbfox | its specific to the members table | 18:56 |
| blogan | sbfox: a lot of us are new in here to this project so our knowledge in icehouse is a bit lacking, but I can try to help | 18:59 |
| sbfox | Thanks! I'm trying to run syncdb to upgrade from havana 2.2 to icehouse. it bombs out with sqlalchemy.exc.OperationalError: (OperationalError) (1061, "Duplicate key name 'uniq_member0pool_id0address0port'") 'ALTER TABLE members ADD CONSTRAINT uniq_member0pool_id0address0port UNIQUE (pool_id, address, protocol_port)' () | 19:00 |
| sbfox | I've checked the schema before running sync db and that contraint does not exist | 19:01 |
| blogan | well that woudl have been the first thing I would look at | 19:02 |
| sbfox | The file responsible for setting the constraint is https://github.com/openstack/neutron/blob/master/neutron/db/loadbalancer/loadbalancer_db.py#L100 and was introduced https://github.com/openstack/neutron/commit/218c249b0dea094944aead116cf07fc19f029852 | 19:03 |
| blogan | this was during the migration though right? | 19:03 |
| blogan | yeah its the alembic migration that is actually doing it | 19:04 |
| *** openstackgerrit has quit IRC | 19:04 | |
| *** openstackgerrit has joined #openstack-lbaas | 19:05 | |
| blogan | hmm im not sure what else to check other than what you already check in that you didn't already ahve that constraint name | 19:05 |
| blogan | you might get better help in #openstack-neutron | 19:06 |
| blogan | sorry sbfox | 19:06 |
| sbfox | I'll give them a go in there, thanks for listening tho :) | 19:07 |
| blogan | lol thats about all i did | 19:07 |
| *** fnaval has joined #openstack-lbaas | 19:16 | |
| *** vjay2 has quit IRC | 19:18 | |
| *** german__ has quit IRC | 19:23 | |
| *** TrevorV__ has joined #openstack-lbaas | 19:23 | |
| *** german__ has joined #openstack-lbaas | 19:24 | |
| *** TrevorV__ has quit IRC | 19:34 | |
| *** rolledback has quit IRC | 19:42 | |
| rm_work | vivek-ebay: if you're around and want to talk Barbican, let me know -- i stalled a little bit waiting to find out what happens with Igor's objection | 19:42 |
| vivek-ebay | Rigth....i saw that... | 19:42 |
| vivek-ebay | lets talk in some time..i am on another call. | 19:42 |
| rm_work | np | 19:43 |
| rm_work | we also have the weekly meeting in 20m | 19:43 |
| *** crc32 has quit IRC | 19:45 | |
| *** crc32 has joined #openstack-lbaas | 19:49 | |
| *** dlundquist has joined #openstack-lbaas | 19:56 | |
| rm_work | can someone link the meeting invite? :P | 19:57 |
| rm_work | pretty please | 19:57 |
| sbalukoff | Webex URL: https://a10networks.webex.com/a10networks/mc | 19:58 |
| sbalukoff | 1. Access code:715 969 090 | 19:58 |
| sbalukoff | 2. If requested, enter your name and email address. | 19:58 |
| sbalukoff | 3. If a password is required, enter the meeting password: Meeting1 | 19:58 |
| sbalukoff | 4. Click "Join". | 19:58 |
| sbalukoff | Also: Call-in toll number (US/Canada): 1-650-429-3300 | 19:58 |
| rm_work | lol this looks different every time | 19:59 |
| *** openstackgerrit has quit IRC | 20:05 | |
| *** openstackgerrit has joined #openstack-lbaas | 20:05 | |
| *** sbfox1 has joined #openstack-lbaas | 20:12 | |
| *** blogan is now known as zz_blogan | 20:13 | |
| *** sbfox1 has quit IRC | 20:13 | |
| dougwig | sorry about the hot mic. ran out of the room without cleaning up. :) | 20:13 |
| *** sbfox1 has joined #openstack-lbaas | 20:13 | |
| *** rolledback has joined #openstack-lbaas | 20:14 | |
| *** sbfox has quit IRC | 20:14 | |
| *** rolledback has quit IRC | 20:15 | |
| *** sbfox has joined #openstack-lbaas | 20:17 | |
| *** sbfox1 has quit IRC | 20:17 | |
| *** rolledback has joined #openstack-lbaas | 20:19 | |
| rm_work | Meeting notes: | 20:20 |
| rm_work | Defacto PTL: sbalukoff | 20:20 |
| rm_work | Meetings: Continue with video conferences for now, keep minutes in IRC | 20:20 |
| rm_work | Official Secretary: TrevorV | 20:21 |
| rm_work | Meeting Time: Still 1pm PST / 3pm CST / 4pm EST | 20:22 |
| rm_work | I'll leave anything further to TrevorV :) | 20:22 |
| *** rolledback has quit IRC | 20:23 | |
| *** rolledback has joined #openstack-lbaas | 20:30 | |
| TrevorV_ | Oh crap, I was supposed to start right now??! Totally not prepared today o_0 | 20:35 |
| rm_work | heh | 20:37 |
| rm_work | I was just doing it right when we started talking about meeting stuff | 20:37 |
| rm_work | before we actually discussed having someone officially do it | 20:38 |
| rm_work | starting next time is prolly good, i was just getting some notes down ;P | 20:38 |
| jorgem1 | Hey everyone! This is your friendly reminder to please update the weekly standup up etherpad ==> https://etherpad.openstack.org/p/neutron-lbaas-weekly-standup | 20:39 |
| *** jorgem1 is now known as jorgem | 20:39 | |
| jorgem | I | 20:39 |
| jorgem | I'll send and email as well :) | 20:39 |
| *** rolledback has quit IRC | 20:44 | |
| rm_work | rm-you | 20:46 |
| ptoohill | the2hill | 20:46 |
| dlundquist | dlundquist | 20:46 |
| german__ | xgerman | 20:46 |
| crc32 | crc32a | 20:46 |
| jorgem | jorgem1106 | 20:46 |
| TrevorV_ | <-- github user: interminator | 20:46 |
| jorgem | Brandon Logan is interim core-reviewer for Rackspace team | 20:50 |
| *** rolledback has joined #openstack-lbaas | 20:50 | |
| sbalukoff | Github user: sbalukoff | 20:52 |
| sbalukoff | (Yeah, I changed it from BlueBoxStephen) | 20:52 |
| sbalukoff | German (I think) is going to be interrim core-reviewer for HP | 20:53 |
| sbalukoff | Vivek and Vijay: Please let us know which of you should be interrim core-reviewer for ebay. | 20:53 |
| sbalukoff | Other notes: We should create (or update if it exists) a "master wiki page" which will contain links to etherpads, blueprints, etc. so that people have a central place to go to know what's going on with and get caught up to speed on both Neutron LBaaS and Octavia. | 20:54 |
| sbalukoff | Stephen will be working on getting a skeleton of structure in place for the Octavia project, as well as writing up the "Octavia Constitution" and design goals, and importing the design work done so far for Octavia into a digestible and reviewable format. | 20:56 |
| sbalukoff | We didn't talk about an official place to record minutes from the meetings. Having them re-iterated in IRC is great, but doesn't work well if you want to find it from history. | 20:56 |
| sbalukoff | I suggest maybe meeting minutes / notes should go in another wiki page or etherpad which gets linked from the master wiki page? | 20:57 |
| *** rolledback has quit IRC | 20:57 | |
| *** rolledback has joined #openstack-lbaas | 20:58 | |
| ptoohill | https://etherpad.openstack.org/p/neutron-lbaas-weekly-standup, this may have been overlooked, posted above by jorgem | 21:00 |
| sbalukoff | Oh yes! That, too! | 21:02 |
| sbalukoff | Please update before tomorrow's IRC meeting! | 21:03 |
| sbalukoff | And dougwig would like more eyes on his logging noop driver gerrit review. | 21:03 |
| dougwig | yep. waiting for stephen's -1 | 21:05 |
| dougwig | :) | 21:05 |
| dougwig | my github is dougwig. (why are we posting github usernames?) | 21:06 |
| ptoohill | dougwig, ctracey will have a 'central' repo that we can fork/PR to. | 21:07 |
| ptoohill | Hes going to add us all | 21:07 |
| dougwig | i see "interim core reviewers". what was the decision there overall? | 21:08 |
| *** VijayB_ has quit IRC | 21:08 | |
| crc32 | for rackspace its going to be blogan | 21:08 |
| crc32 | I thought each company was going to delegate their core reviewer but I'm confused it looks like everyones advocating elections now or something. | 21:09 |
| german__ | I chatted with Susanne and HP will nominate me | 21:09 |
| *** VijayB_ has joined #openstack-lbaas | 21:10 | |
| *** VijayB_ has quit IRC | 21:10 | |
| german__ | core reviewers are not really "elected" they are sort of chosen by the other core reviewers | 21:10 |
| crc32 | seems awkward to nominate while core reviewer just assign from the nomination list. | 21:15 |
| *** sballe has joined #openstack-lbaas | 21:16 | |
| *** vivek-ebay has quit IRC | 21:19 | |
| *** rolledback has quit IRC | 21:21 | |
| *** TrevorV_ has quit IRC | 21:22 | |
| *** HenryG has quit IRC | 21:23 | |
| dougwig | we live for awkward. | 21:25 |
| ctracey | all: here is the org: https://github.com/oslbaas | 21:31 |
| ctracey | please fork and and add your changes to topic branches | 21:31 |
| ctracey | let me know via PM if there are other repos to fork into the repo | 21:32 |
| dlundquist | ctracey: are you going to invite us all to the org? | 21:32 |
| ctracey | i think we should have a few folks with +2 rights | 21:32 |
| ctracey | not to stop progress...more to ensure timely and sane merges | 21:33 |
| ctracey | thoughts? | 21:33 |
| dlundquist | Wouldn't those be the represetitive from each company we chose today? | 21:33 |
| ctracey | sure | 21:35 |
| ctracey | someone pass me a list and I will add people | 21:35 |
| ptoohill | list of the 'core-reviewers' or everyone? | 21:38 |
| *** min has joined #openstack-lbaas | 21:54 | |
| sbalukoff | dougwig: After you were offline, I reiterated some of what we talked about here in IRC. You're still going to have to let us know whether you want / intend to be involved in Octavia given it will probably be a while before it'll see support for 3rd party or proprietary VM images. | 21:55 |
| *** vivek-ebay has joined #openstack-lbaas | 21:56 | |
| sbalukoff | ctracey: I think the list is: blogan from Rackspace, German from HP, we don't have an answer yet from the ebay guys. | 21:56 |
| *** VijayB_ has joined #openstack-lbaas | 21:56 | |
| sbalukoff | ctracey: And I'm defacto dictator-for-life for the forseeable future. | 21:56 |
| sbalukoff | Well until "real" elections happen, which we said should be at Juno. | 21:57 |
| sbalukoff | Anyway, need to go AFK for a few... BBIAB | 21:57 |
| *** markmcclain has quit IRC | 22:08 | |
| VijayB_ | ctracey: vijayendrabvs (github username) | 22:09 |
| vivek-ebay | @ctracey: vivekjain7 is my github username | 22:18 |
| VijayB_ | sbalukoff: Can both Vivek and myself be added to the interim list? | 22:23 |
| sbalukoff | VijayB_: If we do that we should allow two people from each org to get added to the interrim list. :/ | 22:26 |
| sbalukoff | I'm not against this in theory, so long as we work together effectively. | 22:26 |
| sbalukoff | I think it would be bad form to have two people from the same org +2 something to get it merged, though. :P | 22:26 |
| VijayB_ | sbalukoff: yeah pretty much.. I'll leave it to Vivek to decide who among us can be core-review for the interim - I'm fine with either | 22:27 |
| *** min has quit IRC | 22:32 | |
| *** VijayB_ has quit IRC | 22:32 | |
| *** openstackgerrit has quit IRC | 22:35 | |
| *** openstackgerrit has joined #openstack-lbaas | 22:35 | |
| *** markmcclain has joined #openstack-lbaas | 22:45 | |
| *** markmcclain has quit IRC | 22:49 | |
| *** dlundquist has left #openstack-lbaas | 22:52 | |
| *** jorgem has quit IRC | 22:59 | |
| *** markmcclain has joined #openstack-lbaas | 23:01 | |
| *** vivek-ebay has quit IRC | 23:01 | |
| *** vivek-ebay has joined #openstack-lbaas | 23:03 | |
| *** vivek-ebay has quit IRC | 23:03 | |
| *** vivek-ebay has joined #openstack-lbaas | 23:04 | |
| *** sbfox has quit IRC | 23:05 | |
| *** sbfox has joined #openstack-lbaas | 23:20 | |
| *** VijayB has joined #openstack-lbaas | 23:32 | |
| *** sballe has quit IRC | 23:38 | |
| *** fnaval has quit IRC | 23:44 | |
| *** vivek-ebay has quit IRC | 23:49 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!