| *** rm_you has joined #openstack-lbaas | 00:59 | |
| *** rm_you has joined #openstack-lbaas | 01:00 | |
| *** VijayB has quit IRC | 01:12 | |
| *** sbfox has joined #openstack-lbaas | 01:34 | |
| *** vivek-ebay has joined #openstack-lbaas | 01:55 | |
| *** sbfox has quit IRC | 02:00 | |
| *** VijayB has joined #openstack-lbaas | 02:01 | |
| *** vivek-ebay has quit IRC | 02:38 | |
| *** vivek-ebay has joined #openstack-lbaas | 02:40 | |
| *** VijayB has quit IRC | 02:40 | |
| *** vivek-ebay has quit IRC | 02:44 | |
| *** sbfox has joined #openstack-lbaas | 02:53 | |
| *** sbfox has quit IRC | 02:55 | |
| *** vivek-ebay has joined #openstack-lbaas | 03:30 | |
| *** HenryG_ has joined #openstack-lbaas | 04:14 | |
| *** ptoohill-oo has quit IRC | 04:15 | |
| *** HenryG_ has quit IRC | 04:16 | |
| *** HenryG has quit IRC | 04:16 | |
| *** HenryG has joined #openstack-lbaas | 04:18 | |
| *** HenryG is now known as HenryG_afk | 04:18 | |
| *** sbfox has joined #openstack-lbaas | 04:55 | |
| *** vjay has joined #openstack-lbaas | 05:07 | |
| *** sbfox has quit IRC | 05:17 | |
| *** vivek-ebay has quit IRC | 05:18 | |
| *** blogan_ has quit IRC | 05:55 | |
| *** evgenyf has joined #openstack-lbaas | 06:41 | |
| *** sbalukoff has joined #openstack-lbaas | 06:51 | |
| *** sbfox has joined #openstack-lbaas | 06:57 | |
| *** sbfox has quit IRC | 07:05 | |
| *** Youcef has joined #openstack-lbaas | 07:35 | |
| *** sballe has joined #openstack-lbaas | 08:55 | |
| sballe | Morning :-) Back from vacation but still in Denmark. | 08:56 |
|---|---|---|
| *** evgenyf has quit IRC | 09:20 | |
| *** evgenyf has joined #openstack-lbaas | 09:31 | |
| *** sballe has quit IRC | 10:24 | |
| *** sballe has joined #openstack-lbaas | 11:37 | |
| *** vjay has quit IRC | 11:56 | |
| *** HenryG_afk is now known as HenryG | 12:01 | |
| *** vjay has joined #openstack-lbaas | 12:01 | |
| *** sballe has quit IRC | 12:18 | |
| *** vjay has quit IRC | 13:03 | |
| *** vjay has joined #openstack-lbaas | 13:07 | |
| *** vjay has quit IRC | 13:27 | |
| *** vjay has joined #openstack-lbaas | 13:30 | |
| *** vjay has quit IRC | 13:37 | |
| *** vjay has joined #openstack-lbaas | 13:45 | |
| *** vjay has quit IRC | 13:50 | |
| *** vjay has joined #openstack-lbaas | 13:56 | |
| *** markmcclain has joined #openstack-lbaas | 14:28 | |
| *** xgerman has joined #openstack-lbaas | 15:02 | |
| *** xgerman has quit IRC | 15:05 | |
| *** xgerman has joined #openstack-lbaas | 15:07 | |
| *** sbfox1 has joined #openstack-lbaas | 15:07 | |
| *** sbfox2 has joined #openstack-lbaas | 15:11 | |
| *** sbalukoff has quit IRC | 15:12 | |
| *** sbfox has joined #openstack-lbaas | 15:12 | |
| *** sbfox2 has quit IRC | 15:13 | |
| *** sbfox1 has quit IRC | 15:14 | |
| *** sbfox has quit IRC | 15:24 | |
| *** sbfox has joined #openstack-lbaas | 15:24 | |
| dougwig | morning. rough life you've got. :) | 15:32 |
| blogan | morning | 15:41 |
| rm_work | morning | 15:51 |
| *** sbfox has quit IRC | 15:54 | |
| *** markmcclain has quit IRC | 15:56 | |
| *** evgenyf has quit IRC | 15:57 | |
| dougwig | blogan: you wouldn't have to give a provider for every object. just the first that's created (which could also default, if there's a default.) | 15:59 |
| dougwig | after that, it's optional, since it's an error if it's different. | 15:59 |
| dougwig | i guess if you wait to link them, you'd have to give it on each. double and triple extra belch. | 16:01 |
| dougwig | blech. | 16:01 |
| blogan | so if I create a pool with provider A and then I want to create a listener with the same provider and provider A is not default, wouldn't I still need to say the listener is using provider A? | 16:01 |
| *** sbfox has joined #openstack-lbaas | 16:03 | |
| *** Youcef has quit IRC | 16:04 | |
| dougwig | not if the listener references said pool; you could infer. | 16:05 |
| blogan | so if the listener does not specify the provider, then either the user's intent is that they want the listener to be on provider A or the default provider | 16:09 |
| blogan | which we won't know the exact intent | 16:09 |
| dougwig | well, if it's linking into an existing object tree, there is only one valid intent. but i kind of talked myself out of the auto-magic above when i referenced making them all unlinked and then linking. | 16:11 |
| blogan | yeah you're right on the one valid intent | 16:12 |
| blogan | i don't understand your auto-magic scenario above | 16:12 |
| blogan | ah now i do | 16:13 |
| blogan | if you just create without linking, but then expect to update and link | 16:13 |
| blogan | what happens on the create | 16:13 |
| dougwig | exactly. which, btw, is a HUGE GIANT WART we will have to deal with when converting to many-to-many and multiple providers. | 16:14 |
| blogan | yes indeed | 16:14 |
| blogan | if we ever do | 16:15 |
| *** vivek-ebay has joined #openstack-lbaas | 16:16 | |
| *** VijayB has joined #openstack-lbaas | 16:29 | |
| *** HenryG has quit IRC | 16:30 | |
| *** openstack has joined #openstack-lbaas | 16:32 | |
| *** mlavalle has joined #openstack-lbaas | 16:34 | |
| blogan | mlavalle: saw your email, thanks for the updates, looking good | 16:36 |
| blogan | mlavalle: I have some code updates coming soon | 16:37 |
| mlavalle | blogan: glad to help :-) | 16:37 |
| mlavalle | blogan: great. every time you deploy new code, let me know and we can pass it thorugh the api test | 16:37 |
| blogan | mlavalle: will do | 16:37 |
| *** HenryG has joined #openstack-lbaas | 16:39 | |
| blogan | mestery: where there be discussions on the experimental out-of-tree proposal in the neutron meeting today? | 16:43 |
| blogan | mestery: s/where/will | 16:43 |
| *** sbfox1 has joined #openstack-lbaas | 17:10 | |
| *** sbfox has quit IRC | 17:12 | |
| *** Youcef has joined #openstack-lbaas | 17:13 | |
| dougwig | blogan: there's a gbp "post mortem" at the neutron meeting today. i expect that and the various proposals for how to address it will happen then. | 17:15 |
| *** sballe has joined #openstack-lbaas | 17:20 | |
| *** markmcclain has joined #openstack-lbaas | 17:30 | |
| *** vjay has quit IRC | 17:30 | |
| *** sballe has quit IRC | 17:39 | |
| *** vjay has joined #openstack-lbaas | 17:57 | |
| *** sbalukoff has joined #openstack-lbaas | 17:59 | |
| sbalukoff | Guten Morgen! | 18:03 |
| blogan | oh look who it is | 18:03 |
| blogan | dougwig: oh good ill shall participate | 18:07 |
| blogan | or just lurk | 18:07 |
| dougwig | i was planning to lurk | 18:10 |
| *** sballe has joined #openstack-lbaas | 18:19 | |
| *** sbfox1 has quit IRC | 18:34 | |
| *** sbfox has joined #openstack-lbaas | 18:34 | |
| *** sbfox has quit IRC | 18:34 | |
| *** yfried has joined #openstack-lbaas | 18:38 | |
| yfried | dougwig: ? | 18:38 |
| dougwig | hiya. | 18:38 |
| dougwig | you'll get more opinions here. :) | 18:38 |
| dougwig | most of us don't monitor the neutron channel. | 18:39 |
| dougwig | the db relation model changed majorly between v1 and v2, so the same interface won't work. | 18:39 |
| yfried | dougwig: could you give an example? why can't the interface remain the same? | 18:40 |
| dougwig | your point of it being confusing to have two at the same time is valid; at least if a version something doesn't make it obvious. this is a can of worms here, which is why i wanted you to hear the responses in this channel. | 18:40 |
| *** sballe has quit IRC | 18:40 | |
| yfried | dougwig: it feels to me like you would have neutron and novanetwork at the same time | 18:42 |
| dougwig | sure. lots of little stuff. there are new objects: loadbalancer and listener. health monitors are no longer 1-to-many, so the associate/disassociate commands are gone, the current pool requires a subnet, that got moved to members, the current api/cli assumes pool as a root object, all objects are now independent. | 18:42 |
| yfried | dougwig: you can't have them. that's why all novanetwork api calls are proxied to neutron. can't you do something similar with lbv1 if v2 is enabled? | 18:43 |
| dougwig | err, i can't have what? | 18:44 |
| yfried | dougwig: a deployment with both novanetwork and neutron | 18:44 |
| dougwig | it's a good question, and a constraint that i'd personally support. | 18:45 |
| *** sbfox has joined #openstack-lbaas | 18:47 | |
| yfried | dougwig: personally i'd be more comfortable if you moved the lbaas api to a different module (same way nova api has module for servers and hosts) and have v1 and v2 clients implementing the same api calls for common methods such as create_pool | 18:48 |
| yfried | dougwig: at run time the correct client is loaded and if listeners are called from v1 clients you get an appropriate exception | 18:48 |
| dougwig | how do you reconcile that with such massively different object structures? i mean, the entry point would be named the same, but the json payload would be fundamentally incompatible. | 18:49 |
| dougwig | ahh, so same endpoints, different payloads, fail if you detect the wrong version? | 18:49 |
| dougwig | i think that blogan tried something like that when he first starting adding the new api calls, and it got super messy. i'll leave it to him to comment, but the separate extension change was made at the meetup, after a discussion with mark and kyle. | 18:50 |
| yfried | dougwig: I'm more concerned with pythonclients and cli | 18:51 |
| *** sballe has joined #openstack-lbaas | 18:51 | |
| dougwig | that hasn't merged yet: https://review.openstack.org/#/c/111475/ | 18:52 |
| yfried | dougwig: I don't really care if the client implements a different rest call, json payload and uri, as long as the user executes the same command and the payload/uri are decided at runtime according to the client and the lbaas version active | 18:53 |
| yfried | dougwig: making sure you don't run the risk of creating v1 pools with v2 vip (even if the call seems different) | 18:53 |
| dougwig | i'd suggest bringing that up in that review, which is where that interface lives. | 18:54 |
| dougwig | i'm not a strong enough advocate of the separate clients/commands to give you a good counter-argument. :) | 18:54 |
| *** sballe_ has joined #openstack-lbaas | 18:58 | |
| *** sballe has quit IRC | 19:01 | |
| *** sballe_ has quit IRC | 19:01 | |
| *** HenryG_ has joined #openstack-lbaas | 19:04 | |
| *** HenryG has quit IRC | 19:05 | |
| yfried | dougwig: tnx. I've tried to express my concern on the patch | 19:07 |
| *** VijayB has quit IRC | 19:13 | |
| *** vivek-ebay has quit IRC | 19:23 | |
| *** markmcclain has quit IRC | 19:25 | |
| *** sbfox has quit IRC | 19:30 | |
| *** sballe has joined #openstack-lbaas | 19:36 | |
| sbalukoff | yfried: Sorry, I must me missing half this conversation. What is it you're proposing? | 19:38 |
| yfried | sbalukoff: see my review https://review.openstack.org/#/c/111475/3 | 19:39 |
| yfried | I'd like to make sure that if v1 and v2 can't be active at the same time, and that if you use try to use v1 when v2 is active you'll get an appropriate error. just like you can't use neutron when novanetwork is on and vice versa | 19:41 |
| sbalukoff | So, I think most people who have been working on Neutron LBaaS in the last few months would really like to simply ditch the v1 API, object model, etc. But what we learned at the mid-cycle meetup was that that wasn't going to fly and get accepted. | 19:41 |
| sbalukoff | Not that with the discussion about the neutron-incubator project it looks like we're going to get v2 at all... but I digress. | 19:42 |
| dougwig | right, but he's proposing that it gets disabled if v2 is active. | 19:42 |
| dougwig | which does seem reasonable. | 19:42 |
| sbalukoff | What we've been trying to do is move to the new object model and create a shim so that v1 commands (both API and command-line) operate on the v2 models. | 19:43 |
| sbalukoff | Yep, it seems reasonable to me, too. | 19:43 |
| sbalukoff | And it would simplify things for us considerably. | 19:43 |
| yfried | I'm concerned about users creating v1 objects and trying to match them with v2 objects and unable to figure out what's wrong | 19:44 |
| sbalukoff | Well, v1 objects shouldn't actually exist anymore after the v2 object code gets merged. Rather, v1 API / command-line should operate on v2 objects. | 19:44 |
| yfried | not to mention the automation hell that's going to be raised if instead of importing v2 client in place of v1 we will have to search and replace all existing api calls | 19:45 |
| sbalukoff | But you're right, it is dangerous to mix. | 19:45 |
| yfried | sbalukoff: if that's the case then the proposed prefix of lb/lbaas difference is redundant. simply use the same api calls but direct them to the new version | 19:46 |
| sbalukoff | Well, if the neutron-incubator project comes to light, there's a pretty good chance that we'll lose all enthusiasm to make it backward compatible with v1 at all (be that shim layer or what have you). | 19:46 |
| yfried | sbalukoff: glad to finally be understood. I was beginning to think I'm crazy | 19:47 |
| sbalukoff | yfried: Before we were being told that we're unlikely to get the v2 stuff into Juno, we were being told we couldn't do that in order to get the code into Juno. | 19:47 |
| sbalukoff | So... there's history here that lead to this (IMO sub-optimal) direction.... | 19:48 |
| dougwig | yep. x10 | 19:48 |
| sbalukoff | I actually feel like the discussion around neutron-incubator and whether there's any realistic chance of seeing v2 in Juno needs to be resolved before we can decide these kinds of things. | 19:49 |
| sbalukoff | Before we can decide to change direction like you're proposing. | 19:49 |
| dougwig | it's not a direction change to do the intermediate step of only allowing one set of commands to be active. | 19:49 |
| sbalukoff | I think we're all most interested in getting v2 into Juno by jumping through the hoops set before us if there's a chance (we've been jumping through a LOT of hoops). | 19:50 |
| sbalukoff | But... if that's not possible, then I would be all for doing this the "right" way. :P | 19:50 |
| sbalukoff | dougwig: Practically speaking, if neutron-incubator happens, I don't see the shim layer actually being written. (ie. in favor of simply dumping the v1 API / model for v2 when it "graduates") | 19:51 |
| sbalukoff | And if that's the case, then we've inherently solved this problem. | 19:51 |
| sbalukoff | v1 cannot be active at the same time as v2. | 19:51 |
| dougwig | uhh, no. because they are separate api extensions today, you can load both. | 19:52 |
| dougwig | which leads to this: | 19:52 |
| dougwig | https://www.irccloud.com/pastebin/Gf5wy0Tm | 19:52 |
| sbalukoff | dougwig: Right. We've got a mixed API / CLI right now because we were told that's the only way to get this in Juno. | 19:53 |
| *** vivek-ebay has joined #openstack-lbaas | 19:53 | |
| sbalukoff | My point is, if Juno is off the table, then I don't see a point in doing that. | 19:53 |
| sbalukoff | Does that make sense? | 19:53 |
| dougwig | right. has nothing to do with the shim. and as an incubator, with a separate extension, that wouldn't change. | 19:53 |
| sbalukoff | Ok, I don't think i'm following you. | 19:54 |
| *** sballe has quit IRC | 19:54 | |
| yfried | sbalukoff: I'm no project manager, but if that's the case, I'd rather it not go in Juno. that risk and the reaction of disappointed customers is greater than the gain | 19:54 |
| dougwig | no, because going incubator doesn't make lbaas v1 go away as a supported api | 19:54 |
| dougwig | yfried: openstack lbaas customers are already a disappointed bunch. | 19:55 |
| dougwig | :) | 19:55 |
| sbalukoff | dougwig: +1 | 19:55 |
| yfried | dougwig: wait until they here they are going to get a non-working v2 ... ) | 19:55 |
| sbalukoff | lbaas v1 is basically crap. I don't know of any operator of any decent size that actually uses it. | 19:56 |
| *** markmcclain has joined #openstack-lbaas | 19:56 | |
| *** VijayB_ has joined #openstack-lbaas | 19:56 | |
| yfried | sbalukoff: think about those willing to give v2 a second chance and then failing to understand the API | 19:57 |
| *** vivek-ebay has quit IRC | 19:57 | |
| *** vivek-ebay has joined #openstack-lbaas | 19:57 | |
| sbalukoff | yfried: You raise a good point. | 20:01 |
| blogan | im back, sorry meeting | 20:01 |
| dougwig | yep, he does. | 20:02 |
| blogan | catching up one sec | 20:02 |
| yfried | sbalukoff: tnx. glad I was able to talk to you guys while you are still open for discussion | 20:02 |
| blogan | yfried: you're suggesting that v1 API not be available at teh same time as v2 API and they both use the same prefix correct? | 20:05 |
| blogan | and the reason I did not do the shim is because the logic, edge cases, adn what not were pretty insane | 20:07 |
| blogan | and we didn't have time to get that figured out | 20:07 |
| yfried | blogan: you are correct. not only the same cli names but also the same signatures in the pythonclient | 20:16 |
| yfried | blogan: I understand you guys are working under a crazy time table, but I am thinking of the user experience and automation points of view | 20:17 |
| blogan | so if v1 pool has attributes v2 pool does not and vice versa, wouldn't that make the signatures be different? | 20:17 |
| yfried | blogan: I'm sorry. I didn't understand you last comment | 20:17 |
| blogan | well what do you mean by signature? method signatures or command line signature? or something different? | 20:18 |
| yfried | blogan: let me correct myself - not method signature, but at the very least - method names | 20:18 |
| yfried | same for cli commands | 20:19 |
| blogan | ah okay | 20:19 |
| blogan | yeah that can be done | 20:19 |
| blogan | well | 20:19 |
| blogan | yes it can be done, as long as it is expected the parameters can and will be different | 20:19 |
| blogan | yfried: out of curiousity, how would the client handle v2 neutron vs v3 neutron? the same way? | 20:20 |
| yfried | blogan: only if you must :) thank you | 20:20 |
| yfried | blogan: I believe it should | 20:21 |
| yfried | blogan: from user pov, upgrade should demand the least amount of change to he's setup, scripts and MO | 20:21 |
| blogan | yfried: our main problem was that this is akin to versioning an API, except lbaas is an extension of neutron | 20:22 |
| blogan | yfried: so accessing a new version of a REST API is usually done by adding another resource at the version uri level | 20:22 |
| blogan | yfried: but for the client, the lb can be chosen if v1 is active, and lbaas can be chosen if v2 is active | 20:23 |
| *** crc32 has joined #openstack-lbaas | 20:23 | |
| blogan | that makes me wonder, is there a resource in neutron that will list out the active extensions | 20:24 |
| blogan | and if there is i'm an idiot for not already knowing that | 20:24 |
| dougwig | https://www.irccloud.com/pastebin/zmakiKkv | 20:33 |
| yfried | dougwig: sbalukoff: blogan: I need to go to sleep now. it was a good conv all around. please include me on further LB discussions | 20:34 |
| yfried | dougwig: hehe | 20:34 |
| dougwig | yfried: they always happen in our IRC meeting, the ML, or via reviews. | 20:34 |
| dougwig | we will not ignore your -1's. | 20:35 |
| *** Youcef has quit IRC | 20:42 | |
| yfried | blogan: one last thing - I would like to see API&CLI for getting the active LB version from the SERVER | 20:50 |
| yfried | blogan: do you think you could add that? | 20:50 |
| yfried | it would help users/ automation a great deal | 20:50 |
| blogan | yfried: what exactly do you mean? | 20:55 |
| blogan | you mean an api resource to give the version currently running? | 20:55 |
| * sbalukoff returns from a meeting as well. | 21:00 | |
| *** tmc3inphilly has joined #openstack-lbaas | 21:01 | |
| *** HenryG_ is now known as HenryG | 21:01 | |
| *** vivek-ebay has quit IRC | 21:13 | |
| *** vivek-ebay has joined #openstack-lbaas | 21:24 | |
| *** fnaval has joined #openstack-lbaas | 21:28 | |
| *** sbfox has joined #openstack-lbaas | 21:34 | |
| *** vivek-ebay has quit IRC | 21:37 | |
| *** vivek-ebay has joined #openstack-lbaas | 21:42 | |
| *** vjay has quit IRC | 21:43 | |
| dougwig | anyone curious as to the incubator debate who was not at the neutron meeting just now: http://eavesdrop.openstack.org/meetings/networking/2014/networking.2014-08-11-21.00.log.html | 22:04 |
| blogan | sooo any takers on when a consensus on incubator agreement will happen? | 22:04 |
| blogan | actually I think the incubator will be easy to agree on, its just whether GBP should go in there | 22:04 |
| blogan | can't say I disagree with lbaas v2 going in | 22:04 |
| tmc3inphilly | quite honestly, EVERYTHING except what is core Neutron should go through this process | 22:05 |
| tmc3inphilly | and there should be strict guidlines that determine whether it should be a plugin/addon/driver/etc... or merely a consume of neutron | 22:05 |
| tmc3inphilly | *consumer | 22:06 |
| dougwig | well, the greater problem is that in the context of "the cloud", having a single group for "the network" seems to be... optimistic. i mean, we have like three groups for storage, and one for networking? | 22:06 |
| dougwig | but i also hate duplication/waste, and oslo hasn't caught up yet. | 22:06 |
| dougwig | i like the network umbrella, it's just not ready yet. | 22:06 |
| tmc3inphilly | quite honestly, there should be a group for Layers 0 through 2, a separate group for Layer 3, and sepearate group(s) for layer 4 - 7 | 22:07 |
| tmc3inphilly | switching and network ports is a completely different world from routing | 22:08 |
| dougwig | i don't think all of layer7 should be lumped together. that's a bit broad. ;) | 22:08 |
| tmc3inphilly | no that is why i left it as group(s) :-) | 22:09 |
| dougwig | oh, i see. | 22:09 |
| dougwig | skimmed to fast. :) | 22:09 |
| tmc3inphilly | VPN, FW, LB are all network services/applications but should not be lumped together | 22:09 |
| *** vivek-ebay has quit IRC | 22:10 | |
| tmc3inphilly | else we will end up where Nova was and Neutron is | 22:10 |
| tmc3inphilly | like a turtle on its back | 22:10 |
| blogan | neutron needs some cleaning upf irst for that to happen | 22:10 |
| dougwig | at the same time, those are small enough that they could share infrastructure and reviewers and get some cross-gain. at the cost of independence. and i guess the end of that road is the problem we're in now, too. | 22:10 |
| blogan | lots of hooks in the service cross talk code that doesn't go through the API | 22:10 |
| tmc3inphilly | Agreed | 22:11 |
| *** vivek-ebay has joined #openstack-lbaas | 22:12 | |
| tmc3inphilly | Independence to promote stability and velocity, yet same sets of core reviewers to maintain consistency and perspective | 22:12 |
| tmc3inphilly | i think it is achievable | 22:12 |
| tmc3inphilly | the hard part is selling it to the masses | 22:13 |
| *** fnaval_ has joined #openstack-lbaas | 22:28 | |
| *** fnaval has quit IRC | 22:29 | |
| *** sbfox1 has joined #openstack-lbaas | 22:30 | |
| *** sbfox has quit IRC | 22:30 | |
| *** sbfox1 has quit IRC | 22:37 | |
| *** tmc3inphilly has quit IRC | 22:41 | |
| *** sbfox has joined #openstack-lbaas | 22:45 | |
| sbalukoff | dougwig: Thanks for the link to that transcript. I'll try to start attending the Monday Neutron meetings as well. :/ | 22:59 |
| sbalukoff | After all, there clearly wasn't enough contention in that meeting. | 22:59 |
| sbalukoff | :P | 22:59 |
| blogan | sbalukoff: i wonder if openstack would have exploded had you maded it | 23:00 |
| sbalukoff | Haha! | 23:00 |
| sbalukoff | blogan: for what it's worth, you asked the same questions and raised the same concerns I would have. | 23:00 |
| sbalukoff | You just did it a lot nicer than I would have. ;) | 23:00 |
| blogan | lol | 23:00 |
| blogan | well nicer doesn't always get answers | 23:00 |
| *** vivek-eb_ has joined #openstack-lbaas | 23:00 | |
| sbalukoff | Haha! | 23:00 |
| *** vivek-ebay has quit IRC | 23:00 | |
| sbalukoff | Well, I think there's nobody happy with how things have gone with the GBP discussion (on either side of "the aisle" as it were). | 23:01 |
| blogan | i still don't think the incubator would have solved the situation GBP is in | 23:01 |
| sbalukoff | And I really dislike that it's going to affect how things are going for LBaaS as well. | 23:01 |
| blogan | even though I think it is the right solution for neutron | 23:01 |
| sbalukoff | blogan: I agree | 23:01 |
| sbalukoff | If people can't talk openly about their concerns, then process isn't going to help. | 23:02 |
| blogan | well honestly, even if GBP didn't happen, I'm not so certain v2 would have made it in Juno | 23:02 |
| sbalukoff | blogan: I know, but as it is, there's basically zero chance. | 23:02 |
| sbalukoff | Which is frustrating... | 23:02 |
| sbalukoff | my mind keeps going back to the meeting in Atlanta where I told the core devs we didn't trust them to be able to deliver. | 23:02 |
| sbalukoff | :/ | 23:02 |
| sbalukoff | Anyway... | 23:02 |
| blogan | lol | 23:03 |
| sbalukoff | This also potentially affects where we go with Octavia. | 23:03 |
| blogan | I think we just stay the course with Octavia | 23:03 |
| sbalukoff | blogan: Well, there isn't much of a course right now... not in code anyway, since everyone who is competent at coding has been doing so on Neutron LBaaS. | 23:03 |
| blogan | im sure your opinion is different but lbaas will still be useable when Octavia gets in a state that its useable as well. It may not be in neutron proper at first though | 23:04 |
| sbalukoff | VERY frustrating, because if we'd actually just started with Octavia, we might be somewhere with it now. :/ | 23:04 |
| blogan | sbalukoff: call me an optimist (which many do) but the work done isn't going in the trash can | 23:04 |
| sbalukoff | Oh, sure, it's not completely wasted. But it's certainly not the best way we could have been spending our time. | 23:05 |
| sbalukoff | Anyway... this is something I'd like to discuss at the Wednesday Octavia meeting. | 23:05 |
| sbalukoff | Along with the v0.5 spec. Which... yes, I still haven't sent out yet. (Sorry!) | 23:05 |
| * blogan shakes fist at the sky | 23:06 | |
| blogan | WHY?!?!! | 23:06 |
| blogan | j/k | 23:06 |
| sbalukoff | Heh! | 23:06 |
| blogan | we got a good run down on how we will do what we need to do with our network at rax today | 23:07 |
| sbalukoff | Well, trying to keep up with the GBP discussion, and Friday ended up essentially being a sick day to me. (Massive migrane). But anyway... those aren't excuses and I'm truly sorry it's not in y'all's hands yet. | 23:07 |
| sbalukoff | Oh, good! | 23:07 |
| blogan | but it does boil down so far, i'm sure there will be some caveats we aren't thinking about, to the vip being a floating ip in neutron | 23:09 |
| blogan | but we'll have more details on wednesday | 23:10 |
| rm_work | blogan: http://i.imgur.com/2iRD7fm.jpg :P | 23:11 |
| blogan | could you describe that feeling to me? what it feels like to get a merge in? | 23:11 |
| rm_work | hehehe | 23:11 |
| sbalukoff | LOL | 23:12 |
| rm_work | it's wonderful :P | 23:12 |
| rm_work | also the lady at the bakery looked at me funny | 23:12 |
| blogan | and rightfully so | 23:12 |
| blogan | i don't know if you realize this, but everyone looks at you weird | 23:13 |
| rm_work | ^_^ | 23:13 |
| *** vivek-eb_ has quit IRC | 23:19 | |
| *** vivek-ebay has joined #openstack-lbaas | 23:19 | |
| *** vivek-ebay has quit IRC | 23:20 | |
| *** vivek-ebay has joined #openstack-lbaas | 23:21 | |
| *** vivek-ebay has quit IRC | 23:22 | |
| *** vivek-eb_ has joined #openstack-lbaas | 23:22 | |
| rm_work | http://www.huffingtonpost.com/2014/08/11/robin-williams-dead-dies_n_5670050.html :( | 23:23 |
| *** vivek-eb_ has quit IRC | 23:23 | |
| *** vivek-ebay has joined #openstack-lbaas | 23:25 | |
| *** vivek-ebay has quit IRC | 23:25 | |
| *** vivek-ebay has joined #openstack-lbaas | 23:26 | |
| *** markmcclain has quit IRC | 23:27 | |
| *** vivek-ebay has quit IRC | 23:35 | |
| *** vivek-ebay has joined #openstack-lbaas | 23:36 | |
| *** fnaval has joined #openstack-lbaas | 23:44 | |
| *** ptoohill-oo has joined #openstack-lbaas | 23:47 | |
| *** fnaval_ has quit IRC | 23:47 | |
| *** ptoohill-oo has quit IRC | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
