| *** catintheroof has quit IRC | 00:03 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Remove deprecated signing_dir config setting https://review.openstack.org/473630 | 00:04 |
|---|---|---|
| *** sticker has joined #openstack-lbaas | 00:05 | |
| *** catintheroof has joined #openstack-lbaas | 00:11 | |
| *** catintheroof has quit IRC | 00:25 | |
| rm_work | johnsom: ugh ok... so what did you do to fix that | 00:39 |
| rm_work | i will notice it on my next stack | 00:39 |
| *** armax has quit IRC | 00:50 | |
| *** armax has joined #openstack-lbaas | 00:53 | |
| *** JudeC has joined #openstack-lbaas | 00:54 | |
| openstackgerrit | Jude Cross proposed openstack/octavia master: Add filtering and field selection to API https://review.openstack.org/469275 | 01:00 |
| JudeC | not sure why those gates failed :/ | 01:01 |
| *** JudeC has quit IRC | 01:02 | |
| *** aojea has joined #openstack-lbaas | 01:12 | |
| *** dayou has joined #openstack-lbaas | 01:37 | |
| *** aojea has quit IRC | 02:15 | |
| *** yamamoto has joined #openstack-lbaas | 02:25 | |
| *** chlong has quit IRC | 02:42 | |
| *** cody-somerville has quit IRC | 03:20 | |
| *** gcheresh has joined #openstack-lbaas | 03:26 | |
| *** gans has joined #openstack-lbaas | 03:33 | |
| *** armax has quit IRC | 03:50 | |
| *** gcheresh has quit IRC | 03:51 | |
| *** links has joined #openstack-lbaas | 04:32 | |
| *** aojea has joined #openstack-lbaas | 04:45 | |
| *** aojea has quit IRC | 04:45 | |
| *** gans819 has joined #openstack-lbaas | 04:46 | |
| *** gans819 has quit IRC | 04:47 | |
| *** gans has quit IRC | 04:48 | |
| *** aojea has joined #openstack-lbaas | 04:50 | |
| *** aojea has quit IRC | 04:59 | |
| *** aojea has joined #openstack-lbaas | 05:00 | |
| *** aojea has quit IRC | 05:05 | |
| *** gans819 has joined #openstack-lbaas | 05:18 | |
| *** gcheresh has joined #openstack-lbaas | 05:21 | |
| *** rstarmer has joined #openstack-lbaas | 05:26 | |
| *** amotoki_away is now known as amotoki | 05:27 | |
| *** gans819 has quit IRC | 05:30 | |
| *** gans819 has joined #openstack-lbaas | 05:31 | |
| *** gans819 has quit IRC | 05:37 | |
| *** sticker_ has joined #openstack-lbaas | 06:03 | |
| *** rcernin has joined #openstack-lbaas | 06:05 | |
| *** sticker has quit IRC | 06:06 | |
| *** pcaruana has joined #openstack-lbaas | 06:06 | |
| *** gans has joined #openstack-lbaas | 06:08 | |
| *** fnaval has joined #openstack-lbaas | 06:10 | |
| *** fnaval has quit IRC | 06:13 | |
| *** sticker_ is now known as sticker | 06:16 | |
| *** tesseract has joined #openstack-lbaas | 06:41 | |
| *** isantosp has quit IRC | 07:02 | |
| openstackgerrit | Deepak Mourya proposed openstack/octavia master: Remove pbr warnerrors in favor of sphinx check https://review.openstack.org/473714 | 07:22 |
| nmagnezi | rm_work, you do self code reviews? :D https://review.openstack.org/#/c/435612/1/octavia/controller/worker/tasks/network_tasks.py | 07:52 |
| rm_work | maybe :P | 07:54 |
| *** yamamoto has quit IRC | 08:30 | |
| *** yamamoto has joined #openstack-lbaas | 08:35 | |
| *** rstarmer has quit IRC | 08:49 | |
| *** rstarmer has joined #openstack-lbaas | 09:13 | |
| *** rstarmer has quit IRC | 09:17 | |
| *** yamamoto has quit IRC | 09:19 | |
| *** yamamoto has joined #openstack-lbaas | 09:36 | |
| *** yamamoto has quit IRC | 09:37 | |
| *** rstarmer has joined #openstack-lbaas | 09:44 | |
| *** yamamoto has joined #openstack-lbaas | 09:54 | |
| *** yamamoto has quit IRC | 10:23 | |
| *** gans819 has joined #openstack-lbaas | 10:31 | |
| *** gans has quit IRC | 10:34 | |
| *** sanfern has joined #openstack-lbaas | 10:40 | |
| *** sanfern has quit IRC | 10:49 | |
| *** yamamoto has joined #openstack-lbaas | 10:57 | |
| *** sticker_ has joined #openstack-lbaas | 11:07 | |
| *** sticker__ has joined #openstack-lbaas | 11:08 | |
| *** gans819 has quit IRC | 11:11 | |
| *** sticker has quit IRC | 11:11 | |
| *** sticker_ has quit IRC | 11:12 | |
| *** atoth has joined #openstack-lbaas | 11:24 | |
| *** sticker__ is now known as sticker | 11:26 | |
| *** yamamoto has quit IRC | 11:31 | |
| *** sanfern has joined #openstack-lbaas | 11:45 | |
| *** yamamoto has joined #openstack-lbaas | 11:46 | |
| *** yamamoto has quit IRC | 11:50 | |
| *** sanfern has quit IRC | 11:55 | |
| *** aojea has joined #openstack-lbaas | 11:56 | |
| *** sanfern has joined #openstack-lbaas | 11:56 | |
| *** chlong has joined #openstack-lbaas | 11:57 | |
| *** yamamoto has joined #openstack-lbaas | 12:16 | |
| *** openstackgerrit has quit IRC | 12:18 | |
| *** yamamoto has quit IRC | 12:20 | |
| *** leitan has joined #openstack-lbaas | 12:26 | |
| *** catintheroof has joined #openstack-lbaas | 12:33 | |
| *** yamamoto has joined #openstack-lbaas | 13:01 | |
| *** leitan has quit IRC | 13:05 | |
| *** cpuga has joined #openstack-lbaas | 13:06 | |
| *** yamamoto has quit IRC | 13:06 | |
| *** cpuga has quit IRC | 13:11 | |
| *** aojea has quit IRC | 13:11 | |
| *** aojea has joined #openstack-lbaas | 13:22 | |
| *** isantosp has joined #openstack-lbaas | 13:27 | |
| *** links has quit IRC | 13:33 | |
| *** reedip_ has joined #openstack-lbaas | 13:40 | |
| *** leitan has joined #openstack-lbaas | 14:04 | |
| *** armax has joined #openstack-lbaas | 14:08 | |
| *** openstackgerrit has joined #openstack-lbaas | 14:11 | |
| openstackgerrit | Merged openstack/neutron-lbaas master: Replace assertEqual([], items) with assertEmpty(items) https://review.openstack.org/473333 | 14:11 |
| *** leitan has quit IRC | 14:13 | |
| *** leitan has joined #openstack-lbaas | 14:15 | |
| *** leitan_ has joined #openstack-lbaas | 14:17 | |
| *** leitan has quit IRC | 14:17 | |
| *** yamamoto has joined #openstack-lbaas | 14:26 | |
| *** gcheresh has quit IRC | 14:30 | |
| *** yamamoto has quit IRC | 14:32 | |
| *** aojea has quit IRC | 14:32 | |
| *** Guest6511 is now known as med_ | 14:32 | |
| *** med_ has quit IRC | 14:32 | |
| *** med_ has joined #openstack-lbaas | 14:32 | |
| *** belharar has joined #openstack-lbaas | 14:34 | |
| *** cpuga has joined #openstack-lbaas | 14:38 | |
| *** cpuga_ has joined #openstack-lbaas | 14:39 | |
| *** aojea has joined #openstack-lbaas | 14:42 | |
| *** cpuga has quit IRC | 14:43 | |
| *** belharar has quit IRC | 14:52 | |
| *** armax has quit IRC | 14:57 | |
| *** armax has joined #openstack-lbaas | 14:59 | |
| *** tesseract has quit IRC | 15:10 | |
| *** kobis has joined #openstack-lbaas | 15:17 | |
| *** kobis has quit IRC | 15:25 | |
| *** leitan_ has quit IRC | 15:32 | |
| *** aojea has quit IRC | 15:41 | |
| *** rstarmer has quit IRC | 15:42 | |
| *** cody-somerville has joined #openstack-lbaas | 15:43 | |
| *** reedip_ has quit IRC | 15:54 | |
| *** leitan has joined #openstack-lbaas | 15:57 | |
| *** aojea has joined #openstack-lbaas | 16:00 | |
| *** aojea has quit IRC | 16:08 | |
| *** blogan has joined #openstack-lbaas | 16:09 | |
| *** reedip_ has joined #openstack-lbaas | 16:13 | |
| johnsom | FYI: https://review.openstack.org/#/c/467327/ | 16:14 |
| johnsom | OSA enabling alternate haproxy templates. | 16:14 |
| *** csomerville has joined #openstack-lbaas | 16:21 | |
| *** cody-somerville has quit IRC | 16:24 | |
| *** aojea has joined #openstack-lbaas | 16:28 | |
| *** armax has quit IRC | 16:31 | |
| *** aojea has quit IRC | 16:31 | |
| *** armax has joined #openstack-lbaas | 16:33 | |
| *** reedip_ has quit IRC | 16:38 | |
| *** rcernin has quit IRC | 16:41 | |
| *** aojea has joined #openstack-lbaas | 16:53 | |
| *** cpuga_ has quit IRC | 16:57 | |
| *** cpuga has joined #openstack-lbaas | 16:58 | |
| *** csomerville has quit IRC | 17:00 | |
| *** amotoki is now known as amotoki_away | 17:01 | |
| *** leitan has quit IRC | 17:06 | |
| *** ianychoi has joined #openstack-lbaas | 17:15 | |
| *** JudeC has joined #openstack-lbaas | 17:24 | |
| *** kobis has joined #openstack-lbaas | 17:24 | |
| *** sshank has joined #openstack-lbaas | 17:28 | |
| JudeC | johnsom: In regard to your comments on https://review.openstack.org/#/c/469275/8 im not sure how to rectify the issue of filtering on elements like vip_subnet_id, vip_network_id, vip_port_id, etc. I could do a hybrid where it will use the sqlalchemy filter_by() on any known elements and only bubble up the controller filtering on elements that need to go through _conver_db_to_type() first. Im not sure. | 17:29 |
| *** todin has quit IRC | 17:31 | |
| *** leitan has joined #openstack-lbaas | 17:37 | |
| *** kobis has quit IRC | 17:49 | |
| *** gcheresh has joined #openstack-lbaas | 17:49 | |
| rm_work | johnsom: do you know what needs to change in the haproxy configs to optimize for more than one vCPU? | 17:55 |
| rm_work | if any | 17:55 |
| rm_work | i don't know if it's locked to one core by default or if it will automatically expand to use more? | 17:55 |
| johnsom | It is locked to one core by default | 17:57 |
| johnsom | So, there are a couple of things here | 17:57 |
| johnsom | 1. Our flavor is 1 vCPU | 17:57 |
| johnsom | 2. When you enable nProc >1 in haproxy it has a bunch of "restrictions" or things that don't work any longer. | 17:58 |
| johnsom | a. Restarts had an issue | 17:58 |
| johnsom | b. peer sync had some issues | 17:58 |
| *** kobis has joined #openstack-lbaas | 17:58 | |
| johnsom | c. Those may or may not have been resolved. It will take some time studying the docs to find out I think. | 17:59 |
| johnsom | 3. We will natively use more than one core as each listener is a seperate process | 17:59 |
| johnsom | 4. With more than one vCPU you run into the reload issue where some new connections might get TCP RST during the reload unless you are running the 1.8 development version of haproxy (releases early next year I think) | 18:00 |
| johnsom | So, there are some top of mind thoughts.... | 18:00 |
| rm_work | yeah | 18:02 |
| rm_work | if we're less worried about 4 | 18:02 |
| rm_work | and for 2a you mean VM restarts? cause... lol | 18:03 |
| rm_work | 1, our smallest cloud flavor is 4vCPU | 18:03 |
| johnsom | 2a haproxy process restarts | 18:04 |
| *** kobis has quit IRC | 18:06 | |
| *** kobis has joined #openstack-lbaas | 18:09 | |
| *** kobis has quit IRC | 18:11 | |
| johnsom | It has been in the back of my mind to go back, read through the config guide and management guide to see what the current gotchas are. I suspect some have been resolved in the newer versions of HAProxy, but I just haven't had the time to go dig through all that again | 18:11 |
| johnsom | JudeC I think we should be able to generate a filter set that is passed into sqlalchemy based on the API input, such that the query results come back filtered. | 18:14 |
| johnsom | Things like the VIP table, we should be able to create a special repo call that does a DB join on those tables. | 18:15 |
| johnsom | http://docs.sqlalchemy.org/en/latest/orm/query.html#sqlalchemy.orm.query.Query.join | 18:16 |
| JudeC | Gracias. | 18:16 |
| johnsom | My worry is around the filter method our repo is using. Does it have everything we need or are we going to be required to switch to the other sqlalchemy filter method. Which, ugh, some work | 18:17 |
| JudeC | Im not sure, I will spin my wheels at this see where I end up. Let me tell you doing the filter on the controller made things super easy but I agree with you on the performance hit we could take by doing that :P | 18:19 |
| johnsom | Well, the performance is one thing, but the post pagination filtering is the show stopper IMO. Pulling back a sorted list, with results on pages 2+ means the user would get zero results with a filter that may actually match. | 18:20 |
| rm_work | yeah filtering has to happen first | 18:22 |
| rm_work | before paging | 18:22 |
| rm_work | i thought that's where it was | 18:22 |
| johnsom | Not in the current version of the filtering patch | 18:22 |
| rm_work | :( | 18:23 |
| rm_work | so what i was recommending before | 18:24 |
| rm_work | is you add a method to each *model* that knows how to filter its objects and then call that from the context-hook | 18:25 |
| rm_work | since it is technically possible to get to the models from there | 18:25 |
| *** rcernin has joined #openstack-lbaas | 18:26 | |
| *** ipsecguy_ has joined #openstack-lbaas | 18:29 | |
| *** ipsecguy has quit IRC | 18:31 | |
| *** sshank has quit IRC | 18:42 | |
| *** aojea has quit IRC | 18:55 | |
| *** aojea has joined #openstack-lbaas | 19:03 | |
| *** csomerville has joined #openstack-lbaas | 19:39 | |
| *** armax has quit IRC | 19:40 | |
| *** leitan has quit IRC | 20:07 | |
| *** diltram_ has quit IRC | 20:07 | |
| *** diltram has joined #openstack-lbaas | 20:14 | |
| *** leitan has joined #openstack-lbaas | 20:17 | |
| *** wasmum has joined #openstack-lbaas | 20:21 | |
| wasmum | is there anyplace for a full se of doc on how to get octavia running in a mitaka stack? Stack is a dual controller, dual neutron env using vrrp and dvr on the compute nodes. We have the haproxy lb working but needing and wanting to move to octavia. Either I am not finding the docs or not following them correctly? was using this, which is limited - https://docs.openstack.org/admin-guide/networking-config-agents.html | 20:24 |
| wasmum | watched the hands on lab video but it didn't cover the installation and configuration | 20:24 |
| wasmum | just a pointer would be great | 20:24 |
| *** leitan has quit IRC | 20:28 | |
| johnsom | wasmum Our docs are here: https://docs.openstack.org/developer/octavia/ | 20:28 |
| johnsom | This guide gives an overview of the installation process: https://docs.openstack.org/developer/octavia/guides/dev-quick-start.html | 20:29 |
| wasmum | johnsom: much thanks | 20:29 |
| johnsom | It is not a step by step, but at least a good overview. | 20:29 |
| johnsom | This is the script we use to setup octavia for our testing gates: https://github.com/openstack/octavia/blob/master/devstack/plugin.sh | 20:29 |
| johnsom | It can also be a useful reference | 20:29 |
| johnsom | Note, you can also run a newer version of Octavia with a Mitaka OpenStack. | 20:30 |
| *** gcheresh has quit IRC | 20:33 | |
| rm_work | wasmum: I recommend running as new as possible, we are still fixing a LOT of bugs per cycle | 20:33 |
| rm_work | wasmum: i am deploying *master* against our liberty cloud successfully | 20:34 |
| rm_work | but if you're uncomfortable with that much bleeding-edge, i'd use AT LEAST ocata, if not target pike once it's released | 20:34 |
| rm_work | or target pike1/2 | 20:34 |
| wasmum | rm_work: great i'll give it a go, thanks! | 20:35 |
| johnsom | Yeah, Mitaka was the first version people started shipping/using, but really the newer versions are much better | 20:35 |
| johnsom | Ugh, who on earth things that this is useful documentation: | 20:36 |
| johnsom | :param dict target: As much information about the object being operated | 20:36 |
| johnsom | on as possible. | 20:36 |
| *** leitan has joined #openstack-lbaas | 20:40 | |
| rm_work | lol | 20:42 |
| rm_work | i dunno -- who does think that? :P | 20:42 |
| rm_work | git blame | 20:42 |
| johnsom | "It's a nice object with many fields, let me tell you more about it...." | 20:43 |
| *** ianychoi has quit IRC | 20:46 | |
| *** cpuga has quit IRC | 20:49 | |
| *** cpuga has joined #openstack-lbaas | 20:50 | |
| johnsom | Here is another one: | 20:58 |
| johnsom | :param target: dictionary representing the object of the action | 20:58 |
| johnsom | for object creation this should be a dictionary representing the | 20:58 |
| johnsom | location of the object e.g. ``{'project_id': context.project_id}`` | 20:58 |
| * johnsom Scratches head | 20:58 | |
| *** cpuga has quit IRC | 21:08 | |
| *** cpuga_ has joined #openstack-lbaas | 21:08 | |
| openstackgerrit | Mohammed Salih Puthenpurayil proposed openstack/neutron-lbaas master: Fixes module not found error in NetScaler LBaaSV2 driver https://review.openstack.org/473988 | 21:17 |
| *** aojea has quit IRC | 21:24 | |
| *** aojea has joined #openstack-lbaas | 21:24 | |
| *** leitan has quit IRC | 21:25 | |
| *** cpuga has joined #openstack-lbaas | 21:27 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Add RBAC enforcement to Octavia v2 API https://review.openstack.org/472872 | 21:28 |
| johnsom | Ignore these for today ^^^ they are still WIP, just moving some code over to a test host | 21:29 |
| *** cpuga_ has quit IRC | 21:30 | |
| johnsom | stack@devstackpy27-2:~/devstack$ openstack loadbalancer create --name lb2 --vip-subnet-id private-subnet --project 653c55322f2a414786f12e8e86d536e8 | 21:39 |
| johnsom | Unauthorized (HTTP 401) (Request-ID: req-9ebe8697-26fe-4ac2-85e2-87644dcee07e) | 21:39 |
| *** cpuga has quit IRC | 21:39 | |
| johnsom | Well, at some basic level I seem to have it working. | 21:39 |
| *** leitan has joined #openstack-lbaas | 21:45 | |
| *** csomerville has quit IRC | 21:48 | |
| *** rcernin has quit IRC | 21:54 | |
| *** csomerville has joined #openstack-lbaas | 22:06 | |
| *** sshank has joined #openstack-lbaas | 22:18 | |
| *** csomerville has quit IRC | 22:20 | |
| *** cpuga has joined #openstack-lbaas | 22:27 | |
| rm_work | heh | 22:28 |
| rm_work | step one: make no one authorized to do anything | 22:28 |
| *** cpuga_ has joined #openstack-lbaas | 22:29 | |
| rm_work | step two: completely secure system! | 22:29 |
| rm_work | step three: party | 22:29 |
| johnsom | Yeah, that is kind of close to where I am. I can create stuff for my own project, but admin can't create in other projects. | 22:29 |
| *** cpuga has quit IRC | 22:31 | |
| JudeC | I ran into another snag when figuring out how to do this on the db side. I now have the filtering working on elements in other tables but now I see that stuff like provider and flavor are not modeled. | 22:34 |
| JudeC | I am not sure if I should be looking for another way to figure those out or if they should be added to the model. | 22:35 |
| johnsom | Ah, well, my vote would be to just add them to the model. they will need to be there eventually | 22:36 |
| JudeC | kk just checking ty :) | 22:38 |
| *** csomerville has joined #openstack-lbaas | 22:39 | |
| *** catintheroof has quit IRC | 22:47 | |
| *** blogan_ has joined #openstack-lbaas | 22:58 | |
| *** blogan has quit IRC | 22:58 | |
| rm_work | johnsom: i recommended we don't deal with them yet (no point trying to do a bunch of work to filter on something that's a static value) | 23:10 |
| rm_work | and when someone actually *adds them* and they do something, we can make sure filtering works (but it should just "start working") | 23:10 |
| johnsom | Maybe I mis-understood his question, I assumed it was easier to just add those fields to the model. | 23:11 |
| *** sshank has quit IRC | 23:18 | |
| rm_work | eh it seems weird to add those in the commit for filtering, rather than the commit for ... flavors or provider support lol | 23:19 |
| johnsom | Yeah, I just figured it was an issue since we added them to the WSME types, etc | 23:20 |
| *** leitan has quit IRC | 23:23 | |
| *** csomerville has quit IRC | 23:26 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Add RBAC enforcement to Octavia v2 API https://review.openstack.org/472872 | 23:50 |
| *** blogan_ has quit IRC | 23:52 | |
| *** sanfern has quit IRC | 23:52 | |
| *** sanfern has joined #openstack-lbaas | 23:54 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!