| openstackgerrit | Adam Harwell proposed openstack/octavia master: Remove singleton pattern from API controllers https://review.openstack.org/457813 | 00:00 |
|---|---|---|
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Introduce Octavia v2 API for Health Monitor https://review.openstack.org/406256 | 00:01 |
| *** reedip has joined #openstack-lbaas | 00:08 | |
| johnsom | Ok, I am going to start preping dinner. Ping me if there are things I should take a look at. | 00:12 |
| *** JudeC has quit IRC | 00:20 | |
| *** JudeC has joined #openstack-lbaas | 00:22 | |
| rm_work | k | 00:23 |
| *** amotoki has quit IRC | 00:45 | |
| *** amotoki has joined #openstack-lbaas | 00:50 | |
| *** johnsom has quit IRC | 00:53 | |
| *** johnsom has joined #openstack-lbaas | 00:53 | |
| *** ajo has quit IRC | 00:54 | |
| *** amotoki has quit IRC | 00:59 | |
| *** ajo has joined #openstack-lbaas | 01:00 | |
| rm_work | almooooost.... | 01:05 |
| *** JudeC has quit IRC | 01:10 | |
| *** amotoki has joined #openstack-lbaas | 01:14 | |
| openstackgerrit | Merged openstack/neutron-lbaas master: Use OS_AUTH_URL instead of baked in defaults https://review.openstack.org/457781 | 01:14 |
| openstackgerrit | Adam Harwell proposed openstack/neutron-lbaas master: Trying a different strategy for fixing py3 errors in testing https://review.openstack.org/457394 | 01:21 |
| rm_work | ok rebased and +A | 01:23 |
| *** aojea has joined #openstack-lbaas | 01:30 | |
| *** fnaval has quit IRC | 01:32 | |
| *** aojea has quit IRC | 01:36 | |
| *** gongysh has joined #openstack-lbaas | 01:54 | |
| *** amotoki has quit IRC | 02:35 | |
| *** amotoki has joined #openstack-lbaas | 02:50 | |
| *** fnaval has joined #openstack-lbaas | 02:54 | |
| rm_work | johnsom: i see you got most of the rechecks | 02:58 |
| rm_work | I was just going looking and noticed they all floated to the top :P | 02:58 |
| johnsom | Yeah, had a minute after dinner | 02:58 |
| *** csomerville has quit IRC | 02:59 | |
| *** csomerville has joined #openstack-lbaas | 03:00 | |
| rm_work | testing the py3 docs fix now | 03:01 |
| rm_work | since we have no gate for it... | 03:01 |
| *** sanfern has joined #openstack-lbaas | 03:02 | |
| *** JudeC has joined #openstack-lbaas | 03:06 | |
| *** ducnc has quit IRC | 03:06 | |
| *** ducnc1 has joined #openstack-lbaas | 03:07 | |
| rm_work | ok, i hit merge on a few n-lbaas patches that look good to go | 03:08 |
| rm_work | looks like what I'd call "bugs" and testing-only, so whatever | 03:09 |
| *** ducnc1 is now known as ducnc | 03:09 | |
| *** JudeC has quit IRC | 03:10 | |
| *** amotoki has quit IRC | 03:15 | |
| *** JudeC has joined #openstack-lbaas | 03:17 | |
| *** links has joined #openstack-lbaas | 03:19 | |
| *** jerrygb has joined #openstack-lbaas | 03:21 | |
| *** aojea has joined #openstack-lbaas | 03:32 | |
| *** amotoki has joined #openstack-lbaas | 03:35 | |
| *** aojea has quit IRC | 03:37 | |
| *** reedip has quit IRC | 03:50 | |
| openstackgerrit | Merged openstack/neutron-lbaas master: Trying a different strategy for fixing py3 errors in testing https://review.openstack.org/457394 | 03:53 |
| *** ducnc1 has joined #openstack-lbaas | 03:56 | |
| *** gongysh has quit IRC | 03:56 | |
| *** ducnc has quit IRC | 03:56 | |
| *** ducnc1 is now known as ducnc | 03:56 | |
| *** chlong has quit IRC | 03:58 | |
| *** amotoki_ has joined #openstack-lbaas | 03:58 | |
| *** amotoki has quit IRC | 04:00 | |
| openstackgerrit | Merged openstack/neutron-lbaas master: Fix doc generation for Python3 https://review.openstack.org/457225 | 04:01 |
| openstackgerrit | Merged openstack/neutron-lbaas master: api test: Centralize listener/pool protocol https://review.openstack.org/415073 | 04:01 |
| openstackgerrit | Merged openstack/neutron-lbaas master: api test: Add config options for listener/pool protocol https://review.openstack.org/415076 | 04:01 |
| openstackgerrit | Merged openstack/neutron-lbaas master: api test: centralize health monitor protocol https://review.openstack.org/419324 | 04:01 |
| *** amotoki_ has quit IRC | 04:06 | |
| openstackgerrit | Merged openstack/octavia master: Fix dumb error message typo https://review.openstack.org/456950 | 04:13 |
| *** amotoki has joined #openstack-lbaas | 04:19 | |
| *** aojea has joined #openstack-lbaas | 04:33 | |
| *** links has quit IRC | 04:36 | |
| *** aojea has quit IRC | 04:38 | |
| *** links has joined #openstack-lbaas | 04:53 | |
| *** fnaval has quit IRC | 04:55 | |
| *** gongysh has joined #openstack-lbaas | 05:01 | |
| *** gcheresh_ has joined #openstack-lbaas | 05:18 | |
| *** amotoki_ has joined #openstack-lbaas | 05:23 | |
| *** links has quit IRC | 05:25 | |
| *** jerrygb has quit IRC | 05:25 | |
| *** amotoki has quit IRC | 05:26 | |
| johnsom | Yeah, I manually tested the py3 docs patch | 05:27 |
| johnsom | I just want to pull out the tox restrictions where they are implemented | 05:28 |
| *** aojea has joined #openstack-lbaas | 05:32 | |
| *** links has joined #openstack-lbaas | 05:37 | |
| *** reedip_afk is now known as reedip | 05:40 | |
| *** armax has quit IRC | 05:48 | |
| *** JudeC has quit IRC | 05:53 | |
| *** aojea has quit IRC | 06:07 | |
| *** aojea has joined #openstack-lbaas | 06:07 | |
| *** rcernin has joined #openstack-lbaas | 06:08 | |
| *** aojea has quit IRC | 06:12 | |
| *** reedip has quit IRC | 06:23 | |
| *** reedip has joined #openstack-lbaas | 06:24 | |
| *** kobis has joined #openstack-lbaas | 06:25 | |
| *** voelzmo has joined #openstack-lbaas | 06:25 | |
| *** pcaruana has joined #openstack-lbaas | 06:33 | |
| *** JudeC has joined #openstack-lbaas | 06:56 | |
| *** tesseract has joined #openstack-lbaas | 07:01 | |
| openstackgerrit | M V P Nitesh proposed openstack/octavia master: Fix doc generation for Python3 https://review.openstack.org/457224 | 07:20 |
| openstackgerrit | YAMAMOTO Takashi proposed openstack/neutron-lbaas master: scenario test: configure scenario test protocols https://review.openstack.org/418201 | 07:20 |
| *** aojea has joined #openstack-lbaas | 07:24 | |
| rm_work | nmagnezi: err, did you run functional on that test locally? :P looks like a few fails | 07:25 |
| rm_work | http://logs.openstack.org/98/429398/34/check/gate-octavia-tox-functional-ubuntu-xenial/d1a4391/testr_results.html.gz | 07:25 |
| rm_work | just need to change the tests actually :) | 07:26 |
| rm_work | commented, just a one character change in one place lol | 07:27 |
| *** JudeC has quit IRC | 07:35 | |
| *** jerrygb has joined #openstack-lbaas | 07:55 | |
| *** jerrygb has quit IRC | 08:00 | |
| *** ducnc has quit IRC | 08:06 | |
| *** cody-somerville has joined #openstack-lbaas | 08:09 | |
| *** cody-somerville has quit IRC | 08:09 | |
| *** cody-somerville has joined #openstack-lbaas | 08:09 | |
| *** csomerville has quit IRC | 08:11 | |
| *** amotoki_ has quit IRC | 08:22 | |
| openstackgerrit | Merged openstack/neutron-lbaas master: scenario test: configure scenario test protocols https://review.openstack.org/418201 | 09:10 |
| *** dlundquist has quit IRC | 09:23 | |
| *** dlundquist has joined #openstack-lbaas | 09:23 | |
| *** amotoki has joined #openstack-lbaas | 09:29 | |
| *** amotoki has quit IRC | 09:45 | |
| *** amotoki has joined #openstack-lbaas | 09:58 | |
| *** belharar has joined #openstack-lbaas | 10:39 | |
| *** sanfern has quit IRC | 10:55 | |
| *** dmk0202 has joined #openstack-lbaas | 11:02 | |
| *** dmk0202 has quit IRC | 11:21 | |
| *** dmk0202 has joined #openstack-lbaas | 11:21 | |
| *** gongysh has quit IRC | 11:35 | |
| *** amotoki has quit IRC | 11:50 | |
| *** amotoki has joined #openstack-lbaas | 11:57 | |
| *** amotoki has quit IRC | 12:02 | |
| *** voelzmo has quit IRC | 12:23 | |
| *** kobis has quit IRC | 12:23 | |
| *** voelzmo has joined #openstack-lbaas | 12:23 | |
| *** kobis has joined #openstack-lbaas | 12:23 | |
| *** kobis has quit IRC | 12:23 | |
| *** voelzmo has quit IRC | 12:24 | |
| *** voelzmo has joined #openstack-lbaas | 12:24 | |
| *** sanfern has joined #openstack-lbaas | 12:30 | |
| *** krypto has joined #openstack-lbaas | 12:36 | |
| *** gongysh has joined #openstack-lbaas | 12:41 | |
| *** catintheroof has joined #openstack-lbaas | 12:48 | |
| *** kobis has joined #openstack-lbaas | 12:52 | |
| *** jerrygb has joined #openstack-lbaas | 12:53 | |
| *** vlaza has joined #openstack-lbaas | 12:59 | |
| vlaza | Hi everyone, can someone please point me to a functional guide to install Octavia on Ocata/RHEL_7 ? seems I cannot find the right way to do it.. | 13:00 |
| *** reedip_ has joined #openstack-lbaas | 13:06 | |
| *** reedip_ has quit IRC | 13:10 | |
| *** leitan has joined #openstack-lbaas | 13:17 | |
| leitan | Hi guys, is there any clear instructions on how to install / run octavia not in devstack? or its just build the image, upload+tag, and launch it ? | 13:18 |
| *** leitan has quit IRC | 13:31 | |
| *** voelzmo has quit IRC | 13:38 | |
| *** voelzmo has joined #openstack-lbaas | 13:39 | |
| *** reedip_ has joined #openstack-lbaas | 13:42 | |
| *** voelzmo has quit IRC | 13:43 | |
| *** krypto has quit IRC | 13:43 | |
| *** voelzmo has joined #openstack-lbaas | 13:46 | |
| *** kobis has quit IRC | 13:58 | |
| *** links has quit IRC | 14:02 | |
| *** chlong has joined #openstack-lbaas | 14:03 | |
| *** kobis has joined #openstack-lbaas | 14:04 | |
| *** amotoki has joined #openstack-lbaas | 14:14 | |
| *** amotoki has quit IRC | 14:15 | |
| *** kobis has quit IRC | 14:16 | |
| openstackgerrit | Nir Magnezi proposed openstack/octavia master: Auto detect haproxy user_group https://review.openstack.org/429398 | 14:23 |
| *** belharar_ has joined #openstack-lbaas | 14:28 | |
| *** krypto has joined #openstack-lbaas | 14:30 | |
| *** belharar has quit IRC | 14:30 | |
| *** vlaza has quit IRC | 14:31 | |
| *** krypto has quit IRC | 14:36 | |
| *** fnaval has joined #openstack-lbaas | 14:37 | |
| *** krypto has joined #openstack-lbaas | 14:37 | |
| *** amotoki has joined #openstack-lbaas | 14:40 | |
| *** sanfern has quit IRC | 14:43 | |
| *** gcheresh_ has quit IRC | 14:51 | |
| johnsom | vlaza leitan All we have right now is the quick start guide here: https://docs.openstack.org/developer/octavia/guides/dev-quick-start.html | 14:55 |
| johnsom | This lays out the steps required to setup octavia. You can also reference our devstack plugin script here: https://github.com/openstack/octavia/blob/master/devstack/plugin.sh | 14:55 |
| johnsom | We hope to have a more step-by-step guide sometime during Pike | 14:56 |
| *** kobis has joined #openstack-lbaas | 15:01 | |
| *** rcernin has quit IRC | 15:03 | |
| *** belharar_ has quit IRC | 15:03 | |
| *** voelzmo has quit IRC | 15:04 | |
| *** amotoki has quit IRC | 15:19 | |
| *** kobis has quit IRC | 15:24 | |
| *** kobis has joined #openstack-lbaas | 15:24 | |
| *** armax has joined #openstack-lbaas | 15:24 | |
| *** pcaruana has quit IRC | 15:28 | |
| *** gcheresh_ has joined #openstack-lbaas | 15:36 | |
| *** vlaza has joined #openstack-lbaas | 15:47 | |
| *** aojea has quit IRC | 15:51 | |
| *** sanfern has joined #openstack-lbaas | 15:53 | |
| *** tesseract has quit IRC | 15:59 | |
| rm_work | nmagnezi: ugh, missed another one T_T | 16:03 |
| rm_work | johnsom: I wish I knew where diltram was on the lxd gate :/ | 16:04 |
| rm_work | that would have been really nice to have, but it will be a ton of work to continue that and we had NO time for knowledge transfer :( | 16:05 |
| johnsom | Yeah, I think he was waiting on an ubuntu backport to come through | 16:06 |
| johnsom | I may look into it at some point | 16:06 |
| rm_work | I wonder if he's still reachable | 16:08 |
| rm_work | just to tell us what is going on | 16:08 |
| rm_work | so we can pick up his work | 16:09 |
| rm_work | diltram: T_T | 16:09 |
| *** dmk0202 has quit IRC | 16:09 | |
| rm_work | xgerman: https://review.openstack.org/#/c/406256/ | 16:11 |
| rm_work | nmagnezi: also https://review.openstack.org/#/c/454177/ | 16:13 |
| rm_work | johnsom: if you agree with me, you should remove your +2 as it is masking my -1 ^^ | 16:13 |
| rm_work | if not, we should discuss | 16:14 |
| rm_work | wait, now I am confused because I thought when I commented it was on some other changes... | 16:14 |
| johnsom | It needs to handle that input though right? | 16:15 |
| johnsom | Yeah, that is just the imports | 16:15 |
| rm_work | where is the thing i was commenting on, wtf | 16:15 |
| rm_work | ah, should that be a minimal response, or full response? for l7policy and pool | 16:16 |
| rm_work | maybe that is what I meant | 16:16 |
| rm_work | oh, no | 16:16 |
| johnsom | It's minimal, I'm pretty sure I opened a bug for that | 16:17 |
| rm_work | they ARE used in the POST | 16:17 |
| rm_work | see line 122/123 | 16:17 |
| rm_work | IMO if we have a patch to fix this, it should *fix this* not just do the imports, when in reality, we don't even NEED the imports | 16:17 |
| rm_work | because I don't think we can take those things as part of the POST | 16:18 |
| rm_work | and we don't need them for the MinimalResponses | 16:18 |
| rm_work | I think in the end, those imports will actually *go away*] | 16:18 |
| johnsom | Why can't you take them as part of the POST for single call create? | 16:18 |
| rm_work | for single call create sure? | 16:19 |
| rm_work | but this model is also for just Listener | 16:19 |
| rm_work | I mean... are we going to do it that way? | 16:19 |
| rm_work | can you do a partial Single-Create? | 16:19 |
| rm_work | if you start from the Listener endpoint | 16:19 |
| rm_work | can you do Listener+Pool+l7? | 16:19 |
| johnsom | I think so, yes | 16:20 |
| rm_work | ok | 16:20 |
| rm_work | in that case, we need to HANDLE them | 16:20 |
| rm_work | but you want to do that in a different patch? | 16:20 |
| rm_work | this just seems like a sad patch, only does 2 characters change | 16:20 |
| rm_work | when there is a lot more work here | 16:20 |
| johnsom | Yeah, I think the single call stuff is going to be a few patches | 16:20 |
| rm_work | but, sure, I guess this is fine | 16:21 |
| johnsom | I agree, this is super small | 16:21 |
| rm_work | whatever, I can just merge this | 16:21 |
| rm_work | and then start on fixing the rest | 16:21 |
| rm_work | that is my next thing I think | 16:21 |
| rm_work | is the single-call | 16:21 |
| rm_work | well, and trying to get the paging/sorting merged finally | 16:21 |
| rm_work | basically I don't think that patch will DO anything | 16:22 |
| *** gongysh has quit IRC | 16:22 | |
| rm_work | since we are changing the v1 to v2 for something that isn't actually looked at | 16:22 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Implement sorting and pagination for octavia https://review.openstack.org/382147 | 16:23 |
| rm_work | ^^ just a rebase | 16:23 |
| rm_work | is xgerman back today? | 16:24 |
| xgerman | rm_work last day | 16:34 |
| rm_work | xgerman: err... ?? | 16:36 |
| rm_work | last day of vacation? or... *last day*? | 16:36 |
| *** leitan has joined #openstack-lbaas | 16:37 | |
| rm_work | leitan: from your earlier question, for us, we deploy by just making an image, pushing it to glance, and packaging up a venv with octavia installed | 16:39 |
| rm_work | leitan: it's very simple. because, we already have the rest of the cloud running, so we just point Octavia at it in config and it works | 16:39 |
| leitan | rm_work: ok, but if i want to do it outside a virtual env, living with the rest of the components | 16:40 |
| leitan | is there any place so i can read a step by step install | 16:40 |
| leitan | ? | 16:40 |
| rm_work | I mean | 16:40 |
| leitan | cause everything i can find | 16:40 |
| leitan | is for devstack | 16:40 |
| rm_work | you would just install it | 16:40 |
| rm_work | yeah, you can look at what devstack does | 16:41 |
| leitan | if i do pip install octavia on the controller nodes | 16:41 |
| rm_work | basically, you just need to "pip install octavia" | 16:41 |
| rm_work | and create a network for it to use for mgmt | 16:41 |
| rm_work | and have an admin account for it to auth with neutron and nova | 16:41 |
| leitan | rm_work: about the network, can it be a floating ip network ? | 16:42 |
| leitan | that its a flat net | 16:42 |
| rm_work | no | 16:42 |
| rm_work | it needs to be a network that can create/bind ports | 16:42 |
| rm_work | not associate flips | 16:42 |
| leitan | ok, so if the external net is a flat one | 16:42 |
| leitan | it doest work either | 16:43 |
| rm_work | err | 16:43 |
| leitan | i needs to be a flat external_router=false | 16:43 |
| rm_work | I am not sure... | 16:43 |
| rm_work | so, I have a driver that I wrote, that does FLIP binding for the VIP instead of a port | 16:43 |
| rm_work | but | 16:43 |
| rm_work | it is not in the main tree yet | 16:43 |
| rm_work | maybe it is like what you need? | 16:43 |
| rm_work | also it allows doing management over a default net | 16:44 |
| rm_work | instead of a specific management network | 16:44 |
| leitan | no, im talking about the management network | 16:44 |
| leitan | for the vip | 16:44 |
| rm_work | https://review.openstack.org/#/c/435612/ | 16:44 |
| leitan | i can later asociate a floating ip to it | 16:44 |
| rm_work | ok | 16:44 |
| leitan | so is not an issue | 16:44 |
| rm_work | the management network needs to be a normal neutron network | 16:44 |
| leitan | rm_work: normal means a flat ? accesible from the physical nodes ? | 16:45 |
| leitan | BRB | 16:45 |
| rm_work | I don't know what a flat network means | 16:45 |
| rm_work | it needs to be the kind of network you create with 'neutron net-create' | 16:45 |
| rm_work | johnsom: maybe you can help me with terminology? I'm bad at neutron | 16:46 |
| johnsom | It doesn't have to be flat, it can be routable, but the simplest configuration is flat | 16:46 |
| openstackgerrit | Merged openstack/octavia master: Fix doc generation for Python3 https://review.openstack.org/457224 | 16:52 |
| *** amotoki has joined #openstack-lbaas | 16:54 | |
| openstackgerrit | Merged openstack/octavia master: Fix imports in v2 listener types https://review.openstack.org/454177 | 17:01 |
| *** gcheresh_ has quit IRC | 17:07 | |
| *** chlong has quit IRC | 17:09 | |
| *** harlowja_ has quit IRC | 17:13 | |
| *** harlowja has joined #openstack-lbaas | 17:15 | |
| *** harlowja has quit IRC | 17:17 | |
| *** harlowja has joined #openstack-lbaas | 17:17 | |
| *** krypto has quit IRC | 17:25 | |
| *** chlong has joined #openstack-lbaas | 17:25 | |
| *** aojea has joined #openstack-lbaas | 17:38 | |
| xgerman | Last day of vacation | 17:39 |
| *** eandersson has joined #openstack-lbaas | 17:40 | |
| *** catintheroof has quit IRC | 17:47 | |
| leitan | johnsom: ok, so the external network that i already have configured on neutron will do as the management, but if i need to assign a floating ip to the vip will this break up the routing inside the amphora vm | 17:48 |
| leitan | ? | 17:48 |
| johnsom | leitan Let me give an overview of networking for octavia just to make sure we are both talking about the same thing. | 17:48 |
| leitan | rm_work: when i pip install octavia, the i have to configure all the endpoints and that stuff like the ansible playbook does, or its not necesary ? and can work using just the neutron APi and shoot amphorae vms using the driver ? | 17:49 |
| leitan | im not clear about the if the "rest mode" its an option, or its mandatory | 17:49 |
| johnsom | Octavia currently uses service VMs we call amphora. The amphora have an agent running inside them that provides monitoring and management. | 17:49 |
| rm_work | yeah so you will just configure octavia.conf with an admin account and your keystone endpoint | 17:50 |
| rm_work | and it will pick up nova/neutron endpoints from keystone | 17:50 |
| johnsom | The controller processes talk to the amphora agent over the lb-mgmt-network and the amphora send health and stats back to the controllers over this network. That is all it is used for, it is a private network simply for the controllers to talk with the amphora agent. This network is configured at amphora boot time. | 17:50 |
| johnsom | Then, once the amphora is assigned to a load balancer, we hot plug the VIP network into the amphora and add another interface to the amphora. This interface is isolated in a network namespace inside the amphora for security. This is the VIP that customer traffic arrives on. | 17:52 |
| johnsom | When the user adds a member to the load balancer we will hot-plug those required networks into the amphroa inside the network namespace. | 17:52 |
| johnsom | Typically the lb-mgmt-network is not reachable from other networks and it has no need to be routable publicly. It is a private management network. | 17:53 |
| johnsom | leitan Under ocata (I think that was the release you are using) the API is through neutron. API requests go to neutron, then neutron-lbaas, then octavia driver for neutron-lbaas, and then to the octavia API. In this case, you do not need to register the octavia API endpoint with keystone. In fact, we don't recommend it. | 17:55 |
| johnsom | With Pike this will change. | 17:55 |
| johnsom | Optionally | 17:55 |
| leitan | johnsom: great so it will be doing nova interface-add to the amphorae basically | 17:55 |
| leitan | johnsom: im using Mitaka | 17:56 |
| leitan | johnsom: so i need to also bring up the octavia api, i configure the octavia.conf with the database keystone and credentials, sync the db, bring up the api | 17:56 |
| leitan | if i not register the API into a keystone endpoint how neutron knows were the octavia api is ? | 17:57 |
| leitan | or it will capture messages from the queue ? | 17:57 |
| johnsom | leitan Under /etc/neutron/... There is a configuration file for neutron lbaas the has an "[octavia]" section where you specify the octavia API endpoint | 18:00 |
| johnsom | It's called "base_url" I think | 18:00 |
| leitan | johnsom: ooook because i dindt find that in any guide, can you pin point me to an actual "complete" docs, because the lbaasv2 part on the neutron doc is not very octavia friendly | 18:07 |
| leitan | and the etherpads i found it just to build the amphorae image | 18:07 |
| johnsom | https://docs.openstack.org/developer/octavia/ | 18:07 |
| leitan | nothin related to the octaviaAPI/service configuration | 18:07 |
| johnsom | Specifically the Operator quick start guide: https://docs.openstack.org/developer/octavia/guides/dev-quick-start.html | 18:08 |
| johnsom | I'm not sure what etherpads you are speaking of. | 18:08 |
| leitan | johnsom: some i got from the IRC time ago | 18:09 |
| leitan | johnsom: the operator guide will do | 18:09 |
| leitan | ill start from there | 18:09 |
| leitan | and ping you guys if i cant understand something or if its not working | 18:09 |
| leitan | ill probably build a docker image | 18:09 |
| leitan | with the octavia api | 18:09 |
| leitan | so if you guys are interested i can share it later | 18:10 |
| rm_work | yeah, that is what we do | 18:10 |
| leitan | rm_work: is this image public ? | 18:10 |
| rm_work | there is also support for octavia in kolla | 18:10 |
| rm_work | hmm, I don't think so :/ | 18:10 |
| leitan | rm_work: ill take a look at kolla then, maybe it can save me some time | 18:10 |
| rm_work | kolla is for k8s | 18:10 |
| rm_work | but it has some templates for things | 18:11 |
| rm_work | that may be useful | 18:11 |
| rm_work | also openstack-ansible has octavia support I believe | 18:11 |
| reedip_ | rm_work, johnsom : do we need to verify if IP provided to a pool-member should be valid on the subnet its connected to ? Or would that be done automatically by raising an error if it isnt | 18:11 |
| johnsom | leitan OpenStack Ansible also puts all of the octavia processes in containers | 18:11 |
| reedip_ | rm_work, johnsom : reference: https://bugs.launchpad.net/octavia/+bug/1683824 | 18:11 |
| openstack | Launchpad bug 1683824 in octavia "[neutron lbaasv2] ip and subnet is dismatched when creating pool-member" [Undecided,In progress] - Assigned to QingchuanHao (haoqingchuan-28) | 18:11 |
| rm_work | so similarly, there should be some useful templates there | 18:11 |
| rm_work | reedip_: we don't verify, it's up to the user to specify a correct subnet | 18:11 |
| johnsom | leitran Also, you can review the devstack plugin.sh as it has all of the steps we use to configure octavia for devstack: https://github.com/openstack/octavia/blob/master/devstack/plugin.sh | 18:12 |
| rm_work | or to not specify any subnet and we assume it is routable publically | 18:12 |
| leitan | rm_work: yes, i was looking at ansible-octavia at first | 18:12 |
| rm_work | (via the VIP network) | 18:12 |
| leitan | rm_work: this will do https://github.com/openstack/kolla/tree/master/docker/octavia | 18:12 |
| reedip_ | rm_work: thats the bug , the developer/user may want the check to exist in the API (??) | 18:12 |
| rm_work | leitan: cool, yeah | 18:12 |
| johnsom | reedip_ We do not restrict member IPs to the subnet specified as there are use cases where that IP is routable from the subnet specified. | 18:13 |
| leitan | johnsom rm_work thanks for the support guys, ill start from there | 18:13 |
| rm_work | leitan: FYI there is no difference between any of the dockerfiles, other than some weird thing in the api one to kick the db? | 18:13 |
| rm_work | leitan: but IMO just use the same one for every service | 18:13 |
| *** aojea has quit IRC | 18:13 | |
| rm_work | there is no reason to differentiate, they all use exactly the same config and data and such | 18:13 |
| leitan | rm_work: ill build an AIO from that ones | 18:13 |
| rm_work | we use the base | 18:13 |
| reedip_ | johnsom : I gues maybe someone , like you , can explain it a bit more on the bug directly ?? :) | 18:13 |
| *** aojea has joined #openstack-lbaas | 18:14 | |
| johnsom | reedip_ Yeah, I have to close these like every six months because people don't think of the routable case (though some folks use it) | 18:14 |
| reedip_ | johnsom : probably means this sort of use case needs to be documented in the docs, right ? | 18:14 |
| reedip_ | just my thought | 18:15 |
| *** voelzmo has joined #openstack-lbaas | 18:15 | |
| johnsom | reedip_ Sigh, yes. I am working on docs! You can help with reviews.... https://review.openstack.org/455755 | 18:16 |
| johnsom | I'm not down to the members yet. Starting pools now | 18:16 |
| reedip_ | ok .. let me read them :) | 18:16 |
| johnsom | You can click on the API-REF gate to read them rendered | 18:17 |
| rm_work | I am planning to review Listeners doc today | 18:17 |
| rm_work | I actually opened it and started, but I had to stop to talk to someone | 18:17 |
| *** aojea has quit IRC | 18:18 | |
| *** chlong has quit IRC | 18:20 | |
| reedip_ | rm_work : lol , this actually happens everytime a big doc review comes to me :D | 18:22 |
| *** gcheresh_ has joined #openstack-lbaas | 18:25 | |
| johnsom | I'm breaking it down by sections..... Sorry they are a bit big | 18:26 |
| *** catinthe_ has joined #openstack-lbaas | 18:30 | |
| *** amotoki has quit IRC | 18:30 | |
| leitan | rm_work: i know its fun to document, hope i didnt break up your doc mojo :P | 18:34 |
| *** aojea has joined #openstack-lbaas | 18:34 | |
| *** aojea_ has joined #openstack-lbaas | 18:35 | |
| *** aojea has quit IRC | 18:38 | |
| *** catinthe_ has quit IRC | 18:39 | |
| *** aojea_ has quit IRC | 18:41 | |
| *** JudeC has joined #openstack-lbaas | 18:41 | |
| *** aojea_ has joined #openstack-lbaas | 18:44 | |
| *** chlong has joined #openstack-lbaas | 18:45 | |
| *** aojea has joined #openstack-lbaas | 18:47 | |
| *** aojea_ has quit IRC | 18:48 | |
| *** aojea_ has joined #openstack-lbaas | 18:50 | |
| *** aojea has quit IRC | 18:53 | |
| rm_work | hah i was not referring to you actually, your issue was relatively quick | 18:54 |
| rm_work | I was on hold with American Airlines, who couldn't understand how to process my credit card and thought because I booked in JPY not USD I must have booked via a travel agent, even though it was their website forcing me to use JPY... | 18:55 |
| *** JudeC has quit IRC | 19:00 | |
| *** kobis has quit IRC | 19:07 | |
| *** kobis has joined #openstack-lbaas | 19:09 | |
| leitan | rm_work: missing form validation engineer on the staff :P | 19:16 |
| leitan | johnsom: i should probably go with octavia 3.0.3-beta.1 since im using mitaka ? or if i use the master tag should go fine ? | 19:18 |
| johnsom | leitan We have not released a 3.0.3 version of octavia yet.... We are currently working on the 1.0.0 release | 19:19 |
| leitan | im talking about the kolla tag | 19:19 |
| leitan | thought that maybe was synched with the actuall version | 19:20 |
| johnsom | leitan We create stable branches for the releases, so stable/mitaka (which is being EOL'd as we speak) | 19:20 |
| johnsom | https://docs.openstack.org/releasenotes/octavia/ | 19:20 |
| johnsom | Mitaka was 0.8.0 for us | 19:21 |
| johnsom | I'm not very familiar with kolla, but each project has it's own version number chain. | 19:21 |
| johnsom | Also not, for the most part you can use newer versions of Octavia than the rest of your stack should you want new features. You just need to test it to make sure. | 19:22 |
| leitan | ill will take a look at that | 19:22 |
| johnsom | Once mitaka is fully EOL, they will create a tag for mitaka-eol that is the last release of the project. | 19:23 |
| johnsom | For that cycle | 19:23 |
| *** reedip_ has quit IRC | 19:26 | |
| rm_work | whoops, HM bug | 19:26 |
| rm_work | sec | 19:26 |
| *** JudeC has joined #openstack-lbaas | 19:28 | |
| rm_work | going to push up the fix as part of paging/sorting, then backport it to HM | 19:32 |
| rm_work | because it's a big enough issue... | 19:32 |
| rm_work | running tests now | 19:33 |
| leitan | johnsom: ill test the master tag on dockerhub, and downgrade if i find some phantom behaviour | 19:34 |
| leitan | thanks | 19:34 |
| rm_work | yeah, we are planning to run Octavia from master, even on our liberty cloud | 19:36 |
| rm_work | leitan: ^^ | 19:36 |
| rm_work | it is working fine | 19:36 |
| leitan | rm_work: good to hear, ill take that path then, fearless :P | 19:37 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Implement sorting and pagination for octavia https://review.openstack.org/382147 | 19:38 |
| *** voelzmo has quit IRC | 19:44 | |
| *** dmk0202 has joined #openstack-lbaas | 19:47 | |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Remove singleton pattern from API controllers https://review.openstack.org/457813 | 19:48 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Introduce Octavia v2 API for Health Monitor https://review.openstack.org/406256 | 19:48 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Implement sorting and pagination for octavia https://review.openstack.org/382147 | 19:49 |
| rm_work | ok | 19:49 |
| rm_work | everything is nice and fixed and rebased | 19:49 |
| rm_work | johnsom: need to re-vote on HM | 19:49 |
| johnsom | Ok | 19:50 |
| johnsom | I think I need to make a change to the quota patch too for the singleton issue | 19:50 |
| rm_work | yeah | 19:52 |
| rm_work | update it per my template | 19:52 |
| rm_work | and rebase on top of that change | 19:52 |
| rm_work | I wasted like 20 hours of my life on that stupid singleton thing | 19:54 |
| rm_work | once i realized, it took like 5 minutes to fix >_< | 19:56 |
| openstackgerrit | Nir Magnezi proposed openstack/octavia master: Auto detect haproxy user_group https://review.openstack.org/429398 | 19:57 |
| *** voelzmo has joined #openstack-lbaas | 19:57 | |
| johnsom | Octavia meeting starting soon on #openstack-meeting-alt | 19:59 |
| *** catintheroof has joined #openstack-lbaas | 19:59 | |
| *** bread_ has quit IRC | 20:00 | |
| *** voelzmo has quit IRC | 20:02 | |
| rm_work | JudeC: join #openstack-meeting-alt :) | 20:03 |
| *** cody-somerville has quit IRC | 20:08 | |
| *** cody-somerville has joined #openstack-lbaas | 20:09 | |
| *** cody-somerville has quit IRC | 20:09 | |
| *** cody-somerville has joined #openstack-lbaas | 20:09 | |
| *** aojea has joined #openstack-lbaas | 20:18 | |
| *** aojea_ has quit IRC | 20:21 | |
| leitan | im trying something funny i dont know if if should be considered as a bug but, if you bring up the api, without synching the DB i get a fully output of the failed sql query and debug is set to false in the API, is this the epected response or should be a more "hide/friendly error" | 20:22 |
| *** aojea_ has joined #openstack-lbaas | 20:25 | |
| *** aojea has quit IRC | 20:28 | |
| *** gcheresh_ has quit IRC | 20:43 | |
| *** catintheroof has quit IRC | 20:55 | |
| *** catintheroof has joined #openstack-lbaas | 20:55 | |
| *** leitan has quit IRC | 20:59 | |
| *** catintheroof has quit IRC | 21:00 | |
| *** catintheroof has joined #openstack-lbaas | 21:01 | |
| johnsom | It should be a better error | 21:02 |
| *** kobis has quit IRC | 21:03 | |
| *** aojea_ has quit IRC | 21:06 | |
| *** catintheroof has quit IRC | 21:24 | |
| rm_work | hmm, interesting that we exposed that directly | 21:31 |
| rm_work | I guess we just didn't consider | 21:31 |
| rm_work | we should really fix those | 21:31 |
| rm_work | I tried to do a little around kinda genericising the exceptions | 21:31 |
| *** catintheroof has joined #openstack-lbaas | 21:34 | |
| rm_work | nmagnezi: one last set of changes | 21:45 |
| rm_work | for py3 | 21:45 |
| * nmagnezi listens | 21:45 | |
| rm_work | line 119 | 21:46 |
| nmagnezi | rm_work, listener.py? | 21:46 |
| rm_work | one sec | 21:46 |
| rm_work | posted | 21:47 |
| *** jerrygb has quit IRC | 21:47 | |
| nmagnezi | rm_work, I'm not sure i'm following those comments.. | 21:53 |
| nmagnezi | rm_work, i'm not saying there is anything wrong with them | 21:53 |
| nmagnezi | rm_work, i just don't understand :\ | 21:53 |
| nmagnezi | rm_work, will look more closely.. | 21:53 |
| rm_work | we got rid of the binary mode stuff | 22:04 |
| rm_work | so it shouldn't be looking for a bytes write anymore | 22:05 |
| rm_work | right? | 22:05 |
| rm_work | http://logs.openstack.org/98/429398/36/check/gate-octavia-tox-functional-py35-ubuntu-xenial/b42566a/testr_results.html.gz | 22:05 |
| rm_work | JudeC: http://logs.openstack.org/20/456420/9/check/gate-neutron-lbaasv2-dsvm-py3x-scenario-ubuntu-xenial-nv/8a69425/console.html#_2017-04-19_03_27_07_739803 | 22:07 |
| *** chlong has quit IRC | 22:07 | |
| rm_work | JudeC: I THINK that is because of something that I fixed already... | 22:07 |
| rm_work | I wonder if you need to rebase that | 22:07 |
| rm_work | hmm | 22:07 |
| *** dmk0202 has quit IRC | 22:17 | |
| JudeC | rm_work I can try a rebase right now if you would like | 22:19 |
| *** gongysh has joined #openstack-lbaas | 22:38 | |
| *** gongysh has quit IRC | 22:40 | |
| *** fnaval has quit IRC | 22:46 | |
| *** jerrygb has joined #openstack-lbaas | 22:49 | |
| *** reedip_ has joined #openstack-lbaas | 22:52 | |
| *** jerrygb has quit IRC | 22:53 | |
| rm_work | eh, it had a problem anyway JudeC | 22:57 |
| rm_work | might need to test it in devstack yourself and see what's going on | 22:57 |
| JudeC | rm_work ok I will test it right now | 22:58 |
| rm_work | try removing the join timeout | 23:07 |
| rm_work | it never joins | 23:07 |
| *** vlaza has quit IRC | 23:29 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!