| *** bharath_ has joined #openstack-lbaas | 00:02 | |
| *** bharath has quit IRC | 00:02 | |
| *** xgerman_ has quit IRC | 00:12 | |
| *** madhu_ak has quit IRC | 00:18 | |
| *** TrevorV has quit IRC | 00:35 | |
| *** TrevorV has joined #openstack-lbaas | 00:36 | |
| *** fnaval has quit IRC | 00:38 | |
| *** Varun_Lodaya has quit IRC | 00:42 | |
| openstackgerrit | Merged stackforge/octavia: Switched from sha265 to sha256 in octavia.conf https://review.openstack.org/174106 | 00:52 |
|---|---|---|
| *** bharath_ has quit IRC | 01:02 | |
| *** BrianShang_ has quit IRC | 01:11 | |
| *** BrianShang has joined #openstack-lbaas | 01:13 | |
| *** Varun_Lodaya has joined #openstack-lbaas | 01:17 | |
| *** crc32 has joined #openstack-lbaas | 01:25 | |
| openstackgerrit | Phillip Toohill proposed stackforge/octavia: Updates service and config for Octavia API https://review.openstack.org/174009 | 01:26 |
| openstackgerrit | Phillip Toohill proposed stackforge/octavia: Updates service and config for Octavia API https://review.openstack.org/174009 | 01:27 |
| *** sbfox has quit IRC | 01:28 | |
| *** madhu_ak has joined #openstack-lbaas | 01:30 | |
| *** Tiancheng has joined #openstack-lbaas | 01:35 | |
| *** SumitNaiksatam has quit IRC | 01:41 | |
| *** madhu_ak has quit IRC | 01:48 | |
| openstackgerrit | Michael Johnson proposed stackforge/octavia: Implements Octavia Controller Worker https://review.openstack.org/151496 | 01:57 |
| openstackgerrit | Michael Johnson proposed stackforge/octavia: Implemented Queue Consumer https://review.openstack.org/149789 | 02:19 |
| johnsom | Courtesy rebase since I checked in the wrong code yesterday.... | 02:24 |
| *** SumitNaiksatam has joined #openstack-lbaas | 02:35 | |
| *** xgerman_ has joined #openstack-lbaas | 02:40 | |
| *** xgerman_ has quit IRC | 02:42 | |
| *** crc32 has quit IRC | 02:43 | |
| openstackgerrit | Brandon Logan proposed stackforge/octavia: Added name as query parameter for listing https://review.openstack.org/174190 | 02:52 |
| *** blogan_ has quit IRC | 02:53 | |
| *** Varun_Lodaya has quit IRC | 03:21 | |
| *** rm_you| has joined #openstack-lbaas | 03:23 | |
| *** rm_you has quit IRC | 03:27 | |
| *** Tiancheng has quit IRC | 03:47 | |
| *** Tiancheng has joined #openstack-lbaas | 03:47 | |
| *** _kiran_ has joined #openstack-lbaas | 03:47 | |
| *** _kiran_ has quit IRC | 03:53 | |
| *** amotoki has joined #openstack-lbaas | 04:07 | |
| *** BrianShang has quit IRC | 04:11 | |
| *** BrianShang has joined #openstack-lbaas | 04:12 | |
| *** Tiancheng has quit IRC | 04:33 | |
| Santosh_NS | ping | 04:43 |
| openstackgerrit | Phillip Toohill proposed stackforge/octavia: Preparing for tempest testing https://review.openstack.org/172199 | 05:22 |
| *** blogan_ has joined #openstack-lbaas | 05:23 | |
| *** BrianShang has quit IRC | 05:29 | |
| *** BrianShang has joined #openstack-lbaas | 05:30 | |
| *** jschwarz has joined #openstack-lbaas | 05:30 | |
| *** kiran_ has joined #openstack-lbaas | 05:47 | |
| *** kiran_ is now known as kiranr | 05:50 | |
| *** kiranr is now known as kiran-r | 05:50 | |
| openstackgerrit | Phillip Toohill proposed stackforge/octavia: Preparing for tempest testing https://review.openstack.org/172199 | 05:53 |
| *** apuimedo has joined #openstack-lbaas | 06:24 | |
| openstackgerrit | Brandon Logan proposed stackforge/octavia: Allow id to be added in POSTs for all entities https://review.openstack.org/174222 | 06:45 |
| *** blogan_ has quit IRC | 06:46 | |
| *** ebagdasa1 has joined #openstack-lbaas | 06:46 | |
| *** ebagdasa has quit IRC | 06:46 | |
| *** ebagdasa1 has quit IRC | 06:53 | |
| *** ebagdasa has joined #openstack-lbaas | 06:53 | |
| *** Tiancheng has joined #openstack-lbaas | 07:07 | |
| *** kobis has joined #openstack-lbaas | 07:07 | |
| *** kobis has quit IRC | 07:11 | |
| *** Tiancheng has quit IRC | 07:14 | |
| *** rdekel has joined #openstack-lbaas | 07:29 | |
| *** chlong has quit IRC | 07:40 | |
| openstackgerrit | Song Li proposed openstack/neutron-lbaas: supports lbaas alembic migration for db2 https://review.openstack.org/174232 | 07:47 |
| *** kobis has joined #openstack-lbaas | 07:55 | |
| *** Tiancheng has joined #openstack-lbaas | 08:08 | |
| *** rm_you| has quit IRC | 08:39 | |
| *** rm_you| has joined #openstack-lbaas | 08:40 | |
| *** mestery has joined #openstack-lbaas | 08:40 | |
| *** mestery_ has quit IRC | 08:43 | |
| *** rm_work is now known as rm_work|away | 09:13 | |
| *** woodster_ has quit IRC | 09:30 | |
| *** rdekel has quit IRC | 10:07 | |
| *** rdekel has joined #openstack-lbaas | 10:57 | |
| *** Tiancheng has quit IRC | 11:06 | |
| *** amotoki_ has joined #openstack-lbaas | 11:15 | |
| *** sbalukoff1 has quit IRC | 11:40 | |
| *** sbalukoff1 has joined #openstack-lbaas | 11:40 | |
| *** john-davidge has joined #openstack-lbaas | 12:23 | |
| *** woodster_ has joined #openstack-lbaas | 13:00 | |
| jschwarz | blogan, ping | 13:05 |
| openstackgerrit | John Schwarz proposed openstack/neutron-lbaas: Prevent deletion of a subnet with lbaas v1 pool https://review.openstack.org/174384 | 13:37 |
| *** xgerman has joined #openstack-lbaas | 13:54 | |
| *** HenryG_ is now known as HenryG | 14:11 | |
| *** kiran-r has quit IRC | 14:27 | |
| *** mlavalle has joined #openstack-lbaas | 14:53 | |
| *** TrevorV_ has joined #openstack-lbaas | 14:58 | |
| *** ajmiller has joined #openstack-lbaas | 15:01 | |
| *** rm_work|away is now known as rm_work | 15:18 | |
| *** fnaval has joined #openstack-lbaas | 15:35 | |
| *** fnaval has quit IRC | 15:35 | |
| *** fnaval has joined #openstack-lbaas | 15:36 | |
| *** TrevorV_ has quit IRC | 15:42 | |
| *** jschwarz has quit IRC | 15:51 | |
| *** kobis has quit IRC | 16:01 | |
| *** rdekel has quit IRC | 16:04 | |
| *** jschwarz has joined #openstack-lbaas | 16:05 | |
| *** madhu_ak has joined #openstack-lbaas | 16:12 | |
| *** madhu_ak has quit IRC | 16:18 | |
| *** madhu_ak has joined #openstack-lbaas | 16:21 | |
| madhu_ak | Morning | 16:23 |
| ptoohill | Mornin' | 16:23 |
| *** _kiran_ has joined #openstack-lbaas | 16:24 | |
| *** _kiran_ is now known as kiran-r | 16:27 | |
| madhu_ak | Yesterday I spent sometime on 'Admin' tempest tests to check whether the creds actually belongs to admin. when I include the testcase - for creating a loadbalancer with missing tenant_id under 'admin' role, the tenant_id of the admin is not matching with the loadbalancer's tenant_id. | 16:44 |
| madhu_ak | I am not sure why that testcase is not working, but it is working fine for other testcases like empty tenant_id and creating a loadbalancer for other tenant. | 16:44 |
| madhu_ak | I am trying to dig in further, but to be on a safer side, I just logged an issue on neutron, so I can get some information from the neutron tempest experts.. | 16:45 |
| *** kiran-r has quit IRC | 16:45 | |
| madhu_ak | https://bugs.launchpad.net/neutron/+bug/1444765 FYI | 16:45 |
| openstack | Launchpad bug 1444765 in neutron "admin's tenant_id is not the same with load_balancer's tenant_id in tempest tests" [Undecided,New] | 16:45 |
| *** sbalukoff1 has quit IRC | 16:46 | |
| madhu_ak | fnaval, xgerman, blogan ^^ | 16:46 |
| xgerman | dougwig | 16:47 |
| xgerman | madhu_ak I think when you create a loadbalancer as admin then it *can* be created for a different user | 16:48 |
| xgerman | not sure if we support loadbalancers which are under the neutron admin user... | 16:48 |
| xgerman | blogan or dougwig would know for sure | 16:49 |
| madhu_ak | exactly, the same feeling about "support loadbalancers which are under the neutron admin user" | 16:49 |
| *** kiran-r has joined #openstack-lbaas | 16:49 | |
| madhu_ak | I am sure, the admin user would be a neutron admin user because, we are using neutron testcases as a base | 16:49 |
| *** amotoki_ has quit IRC | 16:50 | |
| xgerman | yep | 16:50 |
| madhu_ak | interesting | 16:50 |
| *** bharath has joined #openstack-lbaas | 16:52 | |
| dougwig | morning | 16:59 |
| rm_work | mourning | 17:00 |
| * rm_work finds the morning to be worth mourning | 17:00 | |
| *** apuimedo has quit IRC | 17:07 | |
| *** sbfox has joined #openstack-lbaas | 17:07 | |
| *** kiran-r has quit IRC | 17:10 | |
| *** jschwarz has quit IRC | 17:12 | |
| madhu_ak | dougwig: https://bugs.launchpad.net/neutron/+bug/1444765 | 17:19 |
| openstack | Launchpad bug 1444765 in neutron "admin's tenant_id is not the same with load_balancer's tenant_id in tempest tests" [Undecided,New] | 17:19 |
| *** mwang2 has quit IRC | 17:27 | |
| *** john-davidge has quit IRC | 17:38 | |
| fnaval | thanks for filing the bug madhu_ak | 17:38 |
| openstackgerrit | min wang proposed openstack/neutron-lbaas: Admin API tempest for healthmonitor https://review.openstack.org/173542 | 17:47 |
| *** mwang2 has joined #openstack-lbaas | 17:48 | |
| *** sbalukoff1 has joined #openstack-lbaas | 17:55 | |
| *** fnaval has quit IRC | 17:59 | |
| openstackgerrit | Henry Gessau proposed openstack/neutron-lbaas: Add Kilo release milestone https://review.openstack.org/174497 | 18:00 |
| *** fnaval has joined #openstack-lbaas | 18:14 | |
| *** fnaval has quit IRC | 18:14 | |
| *** fnaval has joined #openstack-lbaas | 18:14 | |
| *** fnaval has quit IRC | 18:19 | |
| *** fnaval has joined #openstack-lbaas | 18:20 | |
| *** sbfox has quit IRC | 19:07 | |
| *** madhu_ak_ has joined #openstack-lbaas | 19:13 | |
| *** madhu_ak has quit IRC | 19:14 | |
| *** madhu_ak_ has left #openstack-lbaas | 19:20 | |
| *** madhu_ak_ has joined #openstack-lbaas | 19:20 | |
| *** TrevorV_ has joined #openstack-lbaas | 19:41 | |
| *** TrevorV_ has quit IRC | 19:42 | |
| blogan | afternoon! | 19:46 |
| blogan | dougwig around? | 19:47 |
| dougwig | blogan: aye | 19:47 |
| blogan | look at those reviews? | 19:47 |
| dougwig | yep, commented on both | 19:47 |
| dougwig | the id one would keep the driver much simpler. | 19:47 |
| rm_work | which two? | 19:48 |
| blogan | i agree | 19:48 |
| blogan | and the name one can still get merged bc that will still be useful | 19:49 |
| blogan | https://review.openstack.org/#/c/174222/ | 19:49 |
| blogan | dougwig: just don't like the idea of the id being able to be chosen | 19:49 |
| blogan | but in the future we can make it more of an operator api only ability, if we ever split user and operator api up | 19:49 |
| blogan | rm_work: also, https://review.openstack.org/#/c/174190/ | 19:50 |
| openstackgerrit | German Eichberger proposed stackforge/octavia: Adds plug VIP and plug Port to spec https://review.openstack.org/174579 | 19:50 |
| dougwig | yeah, the idea of setting the id is kinda scary. less so because it's uuids, though. | 19:50 |
| blogan | yeah, and there is a user friendly message when you try to add an id that already exists | 19:52 |
| xgerman | so using an additional id (e.g. neutron_id) is not an option? | 19:52 |
| blogan | xgerman: that would be almost the same as using name | 19:53 |
| xgerman | yep, just saw that | 19:53 |
| xgerman | catching up | 19:53 |
| blogan | xgerman: yeah its something that came up last night, dougwig yelled at me until i cried | 19:54 |
| xgerman | yeah, if I could only scroll back | 19:54 |
| blogan | you would enjoy me crying | 19:54 |
| xgerman | I sure would :-) | 19:55 |
| blogan | i enjoy your happiness xgerman, so i will be sad and cry just for you | 19:55 |
| xgerman | thanks | 19:55 |
| TrevorV | xgerman I left a comment regarding yours in the SSH driver | 19:56 |
| dougwig | the exact thing i said to blogan last night: "so... you can either let me set the id, use name in the url (and have name on all objects), or I have to do a f'ing db migration in neutron-lbaas for this thing." | 19:56 |
| blogan | you cursed at me | 19:57 |
| dougwig | the mapping of neutron object -> octavia object has to happen somewhere. it's handy if I can overload an octavia field for that. | 19:57 |
| dougwig | i cursed at db migrations. | 19:57 |
| blogan | well its obvious you shouldn't ahve to do that in neutron-lbaas | 19:58 |
| blogan | well we do allow vendors to put their own migrations in | 19:58 |
| xgerman | well, won't that clash with the flavor framewokr? | 20:00 |
| johnsom | So the customer facing "name" would only be in the upstream neutron-lbaas DB and on Octavia | 20:06 |
| johnsom | Isn't that going to be a pain for supporting it? | 20:06 |
| blogan | johnsom: i don't follow | 20:07 |
| johnsom | If we provide the customer haproxy logs, how are they going to do the mapping? | 20:07 |
| blogan | if octavia provdies the haproxy logs, how will neutron_lbaas get them? | 20:08 |
| johnsom | Well, we let customers name things, and if we don't have those same names down at the octavia level we will have to look it up in the neutron-lbaas db and map the uuids down to our db | 20:08 |
| johnsom | My assumption is the log access "feature" would be an extension | 20:09 |
| blogan | well if we allow neutron-lbaas to set the id on create, that should solve the issue? | 20:09 |
| johnsom | So in that case you wouldn't be putting the neutron id into the name fields? | 20:11 |
| blogan | johnsom: no, in that case neutron-lbaas would create the entity with the neutron-lbaas id as the octavia id | 20:16 |
| blogan | johnsom: but you bring up a good point that being able to set the id is probably the best way to go | 20:17 |
| blogan | even though it feels icky | 20:17 |
| blogan | but like i said, it'll feel less icky with admin only being able to do that | 20:18 |
| johnsom | It does feel icky. I would rather do a mapping field or the id pass through instead of overloading fields | 20:18 |
| blogan | what do you mean by id pass through? | 20:19 |
| johnsom | neutron-lbaas id = octavia id | 20:19 |
| blogan | isn't that what we're talking about then? by setting the octavia id to teh neutron-lbaas id? | 20:20 |
| johnsom | I am talking about this patch: https://review.openstack.org/#/c/174190/1 where I don't like overloading the name column | 20:22 |
| blogan | oh yeah, that will cause more headaches for us | 20:23 |
| blogan | so are you fine with the other review? | 20:24 |
| johnsom | What is the argument against adding an optional mapping column? | 20:27 |
| johnsom | (catching up a bit on conversation as I don't see it in the channel) | 20:28 |
| blogan | johnsom: it'd be the same as mapping to the name | 20:29 |
| blogan | johnsom: neutron-lbaas would have to make a request first to retrieve the id by that mapping, and with the nested resources, the number of requests multiplies by the number levels of nesting | 20:30 |
| xgerman | ok, got distracted but I see an argument for making the octavia_id the same as the neutron_id | 20:31 |
| johnsom | Yeah, don't want to do that, but I'm missing why our driver wouldn't hide it with joins or views. (probably my SQL over thinking it) | 20:32 |
| xgerman | the only big argument is that if neutron_lbaas refactors we need to change ids | 20:33 |
| xgerman | and match it | 20:33 |
| johnsom | I think we can go forward with the ID coming in from neutron-lbaas. It doesn't overwrite data so could be changed later if it becomes a problem. | 20:34 |
| xgerman | +1 | 20:34 |
| blogan | yeah if a deployer changes ids in neutron-lbaas db, then octavia would need to a sync as well | 20:34 |
| johnsom | That is their migration issue... | 20:34 |
| blogan | true | 20:35 |
| xgerman | well, tight coupling is an anti-pattern ... | 20:35 |
| xgerman | but in that case it probably doesn't matter | 20:36 |
| blogan | i think its necessary in this case, i don't have another solution other than neutron-lbaas having an octavia-id-mapping table | 20:36 |
| johnsom | Our operator api would still be able to create it's own ids. They are uuid not ints | 20:36 |
| rm_work | wait hold on, you aren't talking about sharing LB-ids in octavia and neutron-lbaas are you? :/ | 20:36 |
| blogan | rm_work: yes | 20:36 |
| rm_work | dear god | 20:37 |
| blogan | rm_work: its not great but its the best bad solution right now | 20:37 |
| rm_work | where in this scrollback is the actual *problem* stated? | 20:38 |
| johnsom | I couldn't find it either | 20:39 |
| blogan | the problem is neutron-lbaas would ahve to store the neutron lbaas id in the name field | 20:39 |
| johnsom | -1 | 20:40 |
| blogan | so then it would also need to retrieve by name to get the octavia id | 20:40 |
| xgerman | mapping neutron-lbaas -> octavia | 20:40 |
| blogan | if we created a neutron_id field, we'd still have to query by neutron_id to get the octavia id | 20:41 |
| blogan | because the resource uri is constructed by /loadbalancers/{octavia_lb_id} | 20:41 |
| blogan | not /loadbalancers/{neutron_lbaas_lb_id} | 20:41 |
| xgerman | yep, and REST really doesn't like resources be known by two ids | 20:41 |
| xgerman | and also since 99% of calls come from neutron we could just adopt tehir id | 20:42 |
| blogan | no that would basically break REST | 20:42 |
| blogan | not break it, but you know be illegal | 20:42 |
| xgerman | yep | 20:42 |
| blogan | go against the standard | 20:42 |
| xgerman | so the options are: | 20:44 |
| xgerman | 1) Store octavia id alongside the neutron id in the Neutron DB | 20:44 |
| xgerman | 2) Query Octavia_id witth /loadbalancer?neutron_id=... and then do the update | 20:45 |
| xgerman | 3) Let Octavia operate on neutron_ids, aka neutron_id=octavia_id | 20:45 |
| blogan | 4) Amazingly simple solution we have overlooked | 20:46 |
| johnsom | I say we keep it simple for now and do 3 | 20:49 |
| xgerman | same here... since 90% of our calls are from neutron-lbaas there is no need for a different id | 20:49 |
| TrevorV | xgerman responded again, when you have time | 20:51 |
| xgerman | TrevorV what I am saying is advisory and I guess we have different opinions what's messy and what's not... | 20:53 |
| TrevorV | Well messy or otherwise, I'm talking about performance in the latter portions. No reason to transfer anything off the amphora and then back onto it just to add to the file in any form... | 20:53 |
| xgerman | well, if for some faithful reson you already have eth0 in there and it shows up again and youa dd it again your whole interface file is hosed | 20:55 |
| xgerman | now, we can make the assumption that the amphora is sick anyway if that happens and we should shoot it | 20:55 |
| xgerman | or we can parse and try to fix | 20:56 |
| TrevorV | I don't understand your first argument... eth0 SHOULD be in there, but I do a grep in a command earlier to pull an interface that isn't up. Then I add what I need to that interface, and bring it up. | 20:56 |
| xgerman | well, interfaces CAN go down and still be in the interfaces file | 20:57 |
| TrevorV | Yes, but I wouldn't be able to connect to the device at that point would I? | 20:57 |
| xgerman | well, let's make a better example: eth0 is management network; eth1 is client network | 20:58 |
| xgerman | eth1 goes down for some reason; you get asked to plug eth2 -- you might try to plug eth1 again | 20:58 |
| xgerman | the boom -- but as I said eth1 being down might indicate the thing is sick and we should just forget about it | 20:59 |
| TrevorV | Yes, that's possible, but then that's a "bug" in the grep command. We'd have to add a little more logic to smarten that portion up. | 20:59 |
| TrevorV | So what I could do, xgerman, is modify the grep line to pull any down interface that doesn't already exist in the /etc/network/interfaces file. That would fix the above issue, and its not even a tough change to that existing line | 21:02 |
| xgerman | ok, that would solve that | 21:03 |
| xgerman | but again I don't know how likely that is ;-) | 21:03 |
| TrevorV | Likely or not, if it can happen, and is avoidable, it should be dealt with appropriately, ya know? | 21:04 |
| TrevorV | I didn't think about that though, so thanks, I'll update | 21:04 |
| xgerman | yep, agreed | 21:04 |
| xgerman | I wish blogan could send us the right eth somehow | 21:04 |
| rm_work | blogan: wait, what do other drivers do? dougwig? A10 must use its own IDs internally, right? how does that work? | 21:05 |
| TrevorV | He might be able to xgerman I'll check with him to see if he has that information... He may not though | 21:05 |
| dougwig | whoa, book. | 21:06 |
| xgerman | yep, that would eb more aweosme and prevent us from ever plugging the wrong interface | 21:06 |
| dougwig | sec, scrolling | 21:06 |
| TrevorV | I'm off for the evening though. I'll add a comment to keep track of that change until I get it implemented xgerman | 21:06 |
| rm_work | it seems insane to me that this is even a problem | 21:06 |
| xgerman | it's OpenStack define insane | 21:07 |
| openstackgerrit | Merged stackforge/octavia: Updates service and config for Octavia API https://review.openstack.org/174009 | 21:07 |
| dougwig | i handle it in the a10 because i can reference the object by either internal id or name, so i set the name to be the neutron id, and then can do a single write op against the neutron id. | 21:08 |
| rm_work | hmm | 21:08 |
| dougwig | if we have to do a get first, we introduce fun race conditions when running multiple neutron-servers. i'd rather have a mapping table in the neutron db. | 21:08 |
| blogan | dougwig do your names have to be unique? | 21:08 |
| dougwig | blogan: yes. | 21:08 |
| blogan | xgerman: neutron has no concept of right eth, but! i can send the mac address | 21:09 |
| blogan | i believe | 21:09 |
| dougwig | since we handle /foo/bar/:id1/baz/:id2, if those fields can handle more than just the internal id, it'll all stitch together fine. but you're right, the alternate field would have to be unique. | 21:09 |
| *** apuimedo has joined #openstack-lbaas | 21:09 | |
| dougwig | johnsom: if we do an association table in the neutron db, the neutron id's would not match up to octavia except via that table, which could cause some support headaches, yes. | 21:10 |
| *** sbfox has joined #openstack-lbaas | 21:13 | |
| openstackgerrit | Aishwarya Thangappa proposed openstack/neutron-lbaas: Added admin/non_admin api tests https://review.openstack.org/173423 | 21:16 |
| openstackgerrit | Aishwarya Thangappa proposed openstack/neutron-lbaas: Admin/Non-admin API tempest tests https://review.openstack.org/171832 | 21:16 |
| dougwig | blogan: from what i'm reading, how about if you add a neutron_id to every object, and accept either id or neutron_id in the url lookup? | 21:16 |
| dougwig | i can do the db migration, but i like to avoid db tables when the alternative is straight-forward. the editable id's does give me a case of the wonkies, i'll admit. | 21:17 |
| blogan | id rather not have neutron-lbaas do a migration | 21:18 |
| blogan | adding that neutron_id functionality would be a lot right now, i could probably do it in one day if i got a streak of hours of uninterrupted tiem | 21:20 |
| blogan | but i can't count on that | 21:20 |
| johnsom | I don't think we should modify neutron-lbaas db for this either | 21:20 |
| dougwig | i don't think we need this *today* or anything like that. | 21:21 |
| blogan | you know how i am | 21:21 |
| dougwig | heh, so you'll code it up tonight? :) | 21:21 |
| blogan | ive already started experimetning :( | 21:22 |
| *** madhu_ak_ has quit IRC | 21:22 | |
| dougwig | lol | 21:22 |
| blogan | but first | 21:22 |
| blogan | lets get consensus | 21:23 |
| blogan | which is best allow neutron-lbaas to set octavia id, or have an alt_id field on every entity that can be used in the resource_id as well? | 21:23 |
| openstackgerrit | German Eichberger proposed stackforge/octavia: Adds plug VIP and plug Port to spec https://review.openstack.org/174579 | 21:23 |
| dougwig | #startvote which is best allow neutron-lbaas to set octavia id, or have an alt_id field on every entity that can be used in the resource_id as well? primary, alt | 21:24 |
| blogan | no meeting bot | 21:24 |
| dougwig | #startmeeting octavia | 21:24 |
| openstack | Meeting started Thu Apr 16 21:24:40 2015 UTC and is due to finish in 60 minutes. The chair is dougwig. Information about MeetBot at http://wiki.debian.org/MeetBot. | 21:24 |
| openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 21:24 |
| blogan | oh wait | 21:24 |
| openstack | The meeting name has been set to 'octavia' | 21:24 |
| dougwig | #startvote which is best allow neutron-lbaas to set octavia id, or have an alt_id field on every entity that can be used in the resource_id as well? primary, alt | 21:24 |
| openstack | Begin voting on: which is best allow neutron-lbaas to set octavia id, or have an alt_id field on every entity that can be used in the resource_id as well? Valid vote options are primary, alt. | 21:24 |
| blogan | you smart one | 21:24 |
| openstack | Vote using '#vote OPTION'. Only your last vote counts. | 21:24 |
| dougwig | #vote alt | 21:25 |
| blogan | #vote alt | 21:25 |
| blogan | the obvious answer is alt | 21:25 |
| johnsom | #vote alt | 21:25 |
| blogan | its more a question of work lol | 21:25 |
| xgerman | #vote primary | 21:25 |
| johnsom | Right | 21:25 |
| blogan | oh xgerman! | 21:25 |
| johnsom | It's all about the work | 21:25 |
| blogan | xgerman: you really like primary over alt? | 21:25 |
| dougwig | i can live with either one. i've gotten past feeling wonky about a change before. | 21:25 |
| dougwig | #endvote | 21:26 |
| openstack | Voted on "which is best allow neutron-lbaas to set octavia id, or have an alt_id field on every entity that can be used in the resource_id as well?" Results are | 21:26 |
| dougwig | #endmeeting | 21:26 |
| openstack | alt (3): dougwig, johnsom, blogan | 21:26 |
| openstack | primary (1): xgerman | 21:26 |
| openstack | Meeting ended Thu Apr 16 21:26:11 2015 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 21:26 |
| openstack | Minutes: http://eavesdrop.openstack.org/meetings/octavia/2015/octavia.2015-04-16-21.24.html | 21:26 |
| openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/octavia/2015/octavia.2015-04-16-21.24.txt | 21:26 |
| openstack | Log: http://eavesdrop.openstack.org/meetings/octavia/2015/octavia.2015-04-16-21.24.log.html | 21:26 |
| xgerman | I don't know why we would need different ids then Neutron LBaaS - we are not selling the thing to people who don't have Neutron LBaaS | 21:26 |
| * dougwig waits for someone to report him for bot abuse. | 21:26 | |
| blogan | good thing we had that bot to sum up 4 votes | 21:26 |
| xgerman | yeah, my hand only has 5 fingers | 21:26 |
| dougwig | blogan: it was an overly verbose joke, yes. | 21:26 |
| blogan | xgerman makes a valid point | 21:27 |
| dougwig | yep. | 21:27 |
| johnsom | Yep, this is why I am fine going forward with the neutron-lbaas ids | 21:28 |
| johnsom | We are speculating on the future | 21:28 |
| blogan | ok alt id good for future wehn we get to that point | 21:28 |
| blogan | right now id is fine | 21:29 |
| blogan | im fine with that too | 21:29 |
| dougwig | i said in blogan's review this morning that i was ok with it. :) https://review.openstack.org/#/c/174222/ | 21:29 |
| dougwig | did we just go in a circle? | 21:29 |
| blogan | yes | 21:29 |
| blogan | but to be certain | 21:29 |
| blogan | vote | 21:29 |
| dougwig | vote in that gerrit review this time? :) | 21:30 |
| blogan | ah damnit i didn't mean to add one of those files | 21:30 |
| blogan | i'm renaming octavia-api.py to octavia-api in this review | 21:32 |
| openstackgerrit | Brandon Logan proposed stackforge/octavia: Allow id to be added in POSTs for all entities https://review.openstack.org/174222 | 21:32 |
| blogan | xgerman: no migrations in this one | 21:32 |
| dougwig | relevant here: 3:32 PM <openstackgerrit> Doug Wiegley proposed openstack/neutron-specs: Lbaas, use Octavia as reference implementation https://review.openstack.org/174616 | 21:36 |
| xgerman | dang - realized after I hit enter | 21:37 |
| blogan | xgerman: the ohter one does have it though, so you get a pass :) | 21:37 |
| blogan | btw i think the other review can still be merged too, would just need change the commit message up | 21:38 |
| blogan | it basically just allows filtering by name, and putting name on member and health monitor | 21:38 |
| blogan | dougwig: looks good to me, im sure it will get some backlash though | 21:39 |
| xgerman | ptoohill? | 21:42 |
| ptoohill | ? | 21:42 |
| xgerman | I thought the Neutron LBaaS team can approve their own blueprints? | 21:42 |
| xgerman | ptoohill I was speculating you would be backlash to annoy blogan | 21:43 |
| ptoohill | Oh, of course, i have to! | 21:43 |
| blogan | xgerman: nope neutron drivers team hast o approve | 21:43 |
| ptoohill | :D | 21:44 |
| *** sbfox has quit IRC | 21:46 | |
| dougwig | xgerman: specs are covered by neutron. https://github.com/openstack/neutron/blob/master/doc/source/policies/core-reviewers.rst "For the Liberty release, the Specs core reviewer team will review specs targeted to all neutron git repositories." | 21:47 |
| *** apuimedo has quit IRC | 21:48 | |
| dougwig | blogan: heh, wait until you see the next one. | 21:48 |
| blogan | dougwig: deprecating v1? | 21:48 |
| dougwig | no, openstackgerrit> Doug Wiegley proposed openstack/neutron-specs: Decompose vendor plugins/drivers for neutron-*aas https://review.openstack.org/174619 | 21:50 |
| blogan | dougwig: some people just want to see the world burn | 21:50 |
| dougwig | i'm not even sure you agree with that one, but we'll see. | 21:50 |
| blogan | you sir are one of those | 21:50 |
| xgerman | burn, baby, burn | 21:51 |
| dougwig | i derive my sense of self-worth from how many people are yelling at me, i guess. | 21:51 |
| blogan | is that why you yell at me all the time? | 21:51 |
| blogan | to inflate my self worth? | 21:51 |
| dougwig | it's why i called the tow truck. | 21:51 |
| dougwig | ha, not really, but that would've been awesome if so. | 21:51 |
| blogan | about as awesome as taking my wallet and taking out $225 | 21:52 |
| rm_work | that would be pretty awesome | 21:52 |
| rm_work | then i'd have $225 | 21:52 |
| blogan | and a tow truck on top of you | 21:53 |
| rm_work | also, you'd be rich enough to be carrying $225 cash for some reason | 21:53 |
| dougwig | hey, we offered to take up a collection. you refused. i got at least $20 of entertainment, so i'd have chipped in. | 21:53 |
| johnsom | dougwig lol | 21:53 |
| blogan | so the next midcycle someone else must provide that entertainment! | 21:54 |
| xgerman | dougwig: https://wiki.openstack.org/wiki/Network/Meetings#On_Demand_Agenda | 22:03 |
| dougwig | xgerman: good. remember to put your name (or whoever will take point) with the item. | 22:06 |
| xgerman | done | 22:07 |
| blogan | resurrecting 2 year old bug reports! | 22:13 |
| blogan | actually 3 | 22:13 |
| openstackgerrit | Aishwarya Thangappa proposed openstack/neutron-lbaas: Added admin/non_admin api tests https://review.openstack.org/173423 | 22:22 |
| *** BrianShang_ has joined #openstack-lbaas | 22:37 | |
| *** BrianShang has quit IRC | 22:40 | |
| openstackgerrit | Aishwarya Thangappa proposed openstack/neutron-lbaas: Added admin/non_admin api tests https://review.openstack.org/173423 | 22:50 |
| *** madhu_ak has joined #openstack-lbaas | 22:50 | |
| *** _cjones_ has joined #openstack-lbaas | 22:56 | |
| *** chlong has joined #openstack-lbaas | 23:02 | |
| *** bharath_ has joined #openstack-lbaas | 23:15 | |
| *** bharath has quit IRC | 23:17 | |
| *** madhu_ak has quit IRC | 23:39 | |
| openstackgerrit | Phillip Toohill proposed stackforge/octavia: Use stevedore to load API handler https://review.openstack.org/174648 | 23:49 |
| *** bharath_ has quit IRC | 23:51 | |
| *** xgerman has quit IRC | 23:52 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!