| johnsom | Neat, if quotas.in_use_load_balancer > 0: doesn't work in py35 where it's false in py27 when in_use_load_balancer is None | 00:11 |
|---|---|---|
| rm_work | errr | 00:27 |
| rm_work | so in py27 it's None and None > 0 evaluates to False? | 00:28 |
| rm_work | and py35 None > 0 doesn't evaluate to false (or at all)? | 00:28 |
| rm_work | seems correct | 00:29 |
| rm_work | that's an odd change though | 00:29 |
| johnsom | Yeah, you end up with this: TypeError: unorderable types: NoneType() > int() | 00:35 |
| *** ducttape_ has joined #openstack-lbaas | 00:42 | |
| *** electrichead is now known as rerobot | 00:45 | |
| rm_work | urgh, deciding how far to go with this network_id thing | 01:05 |
| rm_work | about to submit some more work on it | 01:05 |
| rm_work | if the user provides a subnet, should I fill the network_id for them during validation? | 01:06 |
| johnsom | Yeah, might as well to be consistent. | 01:07 |
| rm_work | is it gross to do that as part of a validation step? :/ | 01:07 |
| johnsom | Can't we just do it when we populate the other DB stuff? | 01:08 |
| rm_work | hmm | 01:09 |
| rm_work | well, I think we populate the VIP object in the API here | 01:09 |
| rm_work | and that's what I mean | 01:09 |
| rm_work | filling the vip object with the network_id during the subnet_id validation, right before we persist it | 01:09 |
| johnsom | Yep, that would be the place | 01:09 |
| rm_work | k... i'll submit and see what you think | 01:09 |
| johnsom | You can do the same on this py3 stuff when I'm done. I am sure you are going to say, why do this when you could have done that.... | 01:11 |
| *** xgerman has joined #openstack-lbaas | 01:24 | |
| rm_work | i'm maybe getting less picky these days T_T | 01:27 |
| rm_work | my days of bright eyed youth and optimism are gone now that I'm basically on my own with octavia over here and trying to get a ton of stuff knocked out, lol | 01:28 |
| *** ducttape_ has quit IRC | 01:39 | |
| *** Guest17022 has joined #openstack-lbaas | 01:42 | |
| *** Guest17022 has quit IRC | 01:56 | |
| rm_work | johnsom: https://review.openstack.org/#/c/429926/4/octavia/api/v1/controllers/load_balancer.py | 02:02 |
| rm_work | johnsom: on line 109 (existing code) | 02:02 |
| rm_work | does that not mean the API returns a *404* because the subnet is missing?! | 02:02 |
| rm_work | that seems not right | 02:02 |
| rm_work | but exceptions.NotFound is set to 404 | 02:03 |
| rm_work | I feel like it should be a 400... | 02:03 |
| rm_work | so maybe we need different exceptions for this | 02:03 |
| johnsom | Just a minute, recovering from a power outage | 02:04 |
| rm_work | ack | 02:05 |
| johnsom | rm_work Yeah, that is what is happening there.... Hmmm | 02:10 |
| rm_work | I'm ... changing it | 02:10 |
| rm_work | ugh sec | 02:10 |
| johnsom | I think that is what neutron does though | 02:10 |
| rm_work | so many test changes | 02:10 |
| rm_work | seriously? | 02:10 |
| rm_work | so we have to stick with that? | 02:10 |
| rm_work | I mean I guess that's ... ok | 02:10 |
| rm_work | I have it fixed | 02:11 |
| rm_work | but I can undo that | 02:11 |
| johnsom | I'm not sure I have a functional stack after the power hit, but you can try creating a port with a bogus subnet | 02:11 |
| rm_work | k | 02:11 |
| johnsom | Or, yeah, try nlbaas | 02:11 |
| rm_work | eugh | 02:13 |
| rm_work | port create does some sanity first | 02:13 |
| rm_work | with the client... | 02:13 |
| rm_work | need to do it directly | 02:14 |
| rm_work | uhh also it takes a network not a subnet? wut | 02:14 |
| johnsom | Yeah, I thought it might... | 02:14 |
| johnsom | fixed ip maybe? | 02:14 |
| rm_work | still doing sanity first T_T | 02:16 |
| rm_work | need to do a direct call | 02:16 |
| rm_work | which ... ugh k | 02:16 |
| johnsom | Yeah, client is giving back not found, but it's basically wrapping a get call that gets [] | 02:17 |
| rm_work | yep | 02:18 |
| rm_work | RESP: [400] Content-Type: application/json Content-Length: 250 X-Openstack-Request-Id: req-12fe0e50-d8aa-4625-8799-7b346d93e263 Date: Tue, 07 Feb 2017 02:18:39 GMT Connection: keep-alive | 02:19 |
| rm_work | RESP BODY: {"NeutronError": {"message": "Invalid input for operation: Failed to create port on network 7779b2a1-7754-473c-99b9-801be0394cda, because fixed_ips included invalid subnet a322b2cd-aaba-4392-ab3d-79f925914f52.", "type": "InvalidInput", "detail": ""}} | 02:19 |
| rm_work | sooo looks like 400 | 02:19 |
| johnsom | Yep | 02:19 |
| rm_work | so we're doing it wrong, fixing | 02:19 |
| johnsom | Thanks for testing it. At least we know we are compatible | 02:20 |
| rm_work | ok just about good | 02:29 |
| rm_work | you about done for the day? | 02:29 |
| rm_work | this turned into a more complex thing than I originally anticipated :/ | 02:29 |
| johnsom | Yeah, just wondering why the py35 tests claim the key needs a passphrase on a key that doesn't have one.... | 02:30 |
| rm_work | err | 02:30 |
| johnsom | Dinner is about ready though | 02:30 |
| rm_work | so there was some funky stuff around that | 02:30 |
| johnsom | I can push it up broken if you want a look | 02:30 |
| rm_work | sidenote - Did you know Musinex makes a version *with* pseudophedrine? I did not! but they do! the world is wonderous | 02:31 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Add network_id to VIP to match neutron-lbaas https://review.openstack.org/429926 | 02:31 |
| johnsom | HA, if you are in a state you can buy it in.... | 02:31 |
| rm_work | there's mine | 02:31 |
| rm_work | friend picked it up :P | 02:31 |
| rm_work | so i'm about to be pretty useless, lol | 02:31 |
| *** gongysh has joined #openstack-lbaas | 02:31 | |
| rm_work | caught convention plague I think | 02:32 |
| *** ducttape_ has joined #openstack-lbaas | 02:33 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Fix py3x gates https://review.openstack.org/424890 | 02:33 |
| rm_work | I can't seem to find the thing that validates that the user passed at least ONE OF the vip fields... | 02:33 |
| rm_work | do we not check that? | 02:34 |
| johnsom | Ok, so that actually boots a LB local for me, but py35 is bombing | 02:34 |
| rm_work | k i'll take a look when the tests run | 02:34 |
| rm_work | or can i repro easily with tox? | 02:34 |
| johnsom | Ah, yes, I am pretty sure it's there somewhere. | 02:34 |
| johnsom | tox -e py35 should do it | 02:34 |
| rm_work | oh, it's that early, k | 02:35 |
| johnsom | Ok, spousal unit calling. Catch you later | 02:35 |
| rm_work | kk | 02:35 |
| rm_work | johnsom: fixed it | 02:42 |
| rm_work | one sec | 02:42 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Fix py3x gates https://review.openstack.org/424890 | 02:45 |
| *** ducttape_ has quit IRC | 02:45 | |
| rm_work | done | 02:45 |
| rm_work | sorry that was some bad code to begin with in some places there T_T | 02:45 |
| *** ianychoi has quit IRC | 02:59 | |
| *** ianychoi has joined #openstack-lbaas | 02:59 | |
| *** ducttape_ has joined #openstack-lbaas | 03:02 | |
| rm_work | wut pep8 fail | 03:07 |
| * rm_work sighs | 03:07 | |
| rm_work | ok so, found a bunch of real craziness | 03:14 |
| rm_work | that I am sure is probably mostly my fault | 03:14 |
| rm_work | cleaning up all the py2/py3 craziness and removing the dependency on six | 03:14 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Fix py3x gates https://review.openstack.org/424890 | 03:19 |
| rm_work | why did I do so much dumb | 03:19 |
| *** ducttape_ has quit IRC | 03:37 | |
| *** links has joined #openstack-lbaas | 03:41 | |
| *** ducttape_ has joined #openstack-lbaas | 03:47 | |
| *** ducttape_ has quit IRC | 04:03 | |
| *** ducttape_ has joined #openstack-lbaas | 04:05 | |
| *** ducttape_ has quit IRC | 04:05 | |
| *** ducttape_ has joined #openstack-lbaas | 04:06 | |
| *** fnaval_ has joined #openstack-lbaas | 04:06 | |
| *** fnaval has joined #openstack-lbaas | 04:13 | |
| *** fnaval_ has quit IRC | 04:15 | |
| *** armax has quit IRC | 04:21 | |
| rm_work | oh hmm nm this is a problem i guess | 04:25 |
| rm_work | I tested that but i guess it is somewhat weird | 04:26 |
| *** ducttape_ has quit IRC | 04:30 | |
| johnsom | rm_work FYI, I want to talk about this bug at the meeting this week: https://bugs.launchpad.net/octavia/+bug/1659488 | 04:35 |
| openstack | Launchpad bug 1659488 in octavia "Octavia is not handling VIPs on the same subnet as the lb-mgmt-net" [Medium,Triaged] | 04:35 |
| johnsom | You might be interested in it | 04:35 |
| rm_work | hmm mayhaps | 04:35 |
| rm_work | i think i had some changes around that | 04:35 |
| johnsom | sanfern Is thinking about working on it | 04:35 |
| rm_work | also, i realized I DID need some of the code i simplified | 04:35 |
| rm_work | so i'm working on reverting just the necessary bits | 04:35 |
| johnsom | Grin | 04:36 |
| *** ducttape_ has joined #openstack-lbaas | 04:37 | |
| johnsom | I am still getting my head around the best way to deal with these py27/py35 issues. More to learn.... | 04:40 |
| *** ducttape_ has quit IRC | 04:44 | |
| *** sanfern has joined #openstack-lbaas | 04:44 | |
| *** csomerville has quit IRC | 05:00 | |
| *** csomerville has joined #openstack-lbaas | 05:00 | |
| rm_work | yeah the problem is i think we're not TESTING with the right object types... | 05:04 |
| johnsom | Yeah, I did get a load balancer to create with that code. | 05:10 |
| rm_work | ugh well it SHOULDN'T | 05:10 |
| johnsom | I didn't run the full test suite, but at least it created | 05:10 |
| rm_work | because it's actually wrong for 3.5 | 05:11 |
| rm_work | http://logs.openstack.org/90/424890/14/check/gate-octavia-v1-dsvm-py3x-scenario-ubuntu-xenial-nv/0faa560/logs/screen-o-cw.txt.gz#_2017-02-07_03_46_42_685 | 05:11 |
| rm_work | I thought i tested .encode on 3.5 but it isn't working the way i expected | 05:11 |
| johnsom | Hmmm, I thought I had fixed that error at one point, but ... | 05:12 |
| rm_work | well i broke it | 05:12 |
| rm_work | trying to simplify | 05:12 |
| johnsom | Ok. Well, I am about to sign off for the night. I will take a look again in the morning. | 05:13 |
| *** anilvenkata has joined #openstack-lbaas | 05:13 | |
| rm_work | kk | 05:13 |
| *** fnaval has quit IRC | 05:38 | |
| *** fnaval has joined #openstack-lbaas | 05:39 | |
| *** fnaval has quit IRC | 05:44 | |
| *** itsuugo has joined #openstack-lbaas | 05:45 | |
| *** sanfern has quit IRC | 05:46 | |
| *** sanfern has joined #openstack-lbaas | 05:51 | |
| *** links has quit IRC | 05:51 | |
| *** rcernin has joined #openstack-lbaas | 06:00 | |
| *** rcernin has quit IRC | 06:03 | |
| *** rcernin has joined #openstack-lbaas | 06:04 | |
| *** links has joined #openstack-lbaas | 06:07 | |
| *** sanfern has quit IRC | 06:13 | |
| *** sanfern has joined #openstack-lbaas | 06:14 | |
| *** itsuugo has quit IRC | 06:20 | |
| *** links has quit IRC | 06:22 | |
| *** itsuugo has joined #openstack-lbaas | 06:35 | |
| *** links has joined #openstack-lbaas | 06:38 | |
| *** ipsecguy has joined #openstack-lbaas | 06:49 | |
| *** ipsecguy_ has quit IRC | 06:50 | |
| *** itsuugo has quit IRC | 06:55 | |
| *** sanfern has quit IRC | 06:55 | |
| rm_work | startMarkers = dict(map(lambda x: (x[1], x[0]), | 06:57 |
| rm_work | enumerate(map(lambda x: x[0], markers)))) | 06:57 |
| rm_work | WHAT | 06:57 |
| rm_work | markers is a tuple of two strings... | 06:58 |
| rm_work | carlos why did you do this | 06:58 |
| rm_work | it literally outputs: {'str1': 0} and always will | 06:59 |
| *** links has quit IRC | 07:07 | |
| *** eezhova has joined #openstack-lbaas | 07:11 | |
| *** links has joined #openstack-lbaas | 07:19 | |
| *** itsuugo has joined #openstack-lbaas | 07:21 | |
| korean101 | hi guys. | 07:23 |
| korean101 | how can i use "insert_headers"? | 07:24 |
| korean101 | https://docs.openstack.org/developer/octavia/api/octaviaapi.html#create-listener | 07:24 |
| korean101 | i need "X-Forwarded-For" option | 07:24 |
| korean101 | use this command? # neutron lbaas-loadbalancer-create | 07:24 |
| korean101 | oh no... | 07:25 |
| korean101 | mis typing | 07:25 |
| korean101 | neutron lbaas-listener-create | 07:25 |
| *** links has quit IRC | 07:30 | |
| *** tesseract has joined #openstack-lbaas | 07:31 | |
| rm_work | johnsom: oh my god i think i might be out of it and i am pretty sure i have hacked this all to shit so uhhhh if you need to revert my patches from today, you can probably do that. BUT, I am hopeful about this one >_> | 07:32 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Fix py3x gates https://review.openstack.org/424890 | 07:32 |
| *** itsuugo has quit IRC | 07:32 | |
| *** yamamoto has quit IRC | 07:38 | |
| *** links has joined #openstack-lbaas | 07:43 | |
| *** pcaruana has joined #openstack-lbaas | 07:46 | |
| *** pcaruana has quit IRC | 07:48 | |
| *** sanfern has joined #openstack-lbaas | 07:50 | |
| *** cody-somerville has quit IRC | 07:52 | |
| *** pcaruana has joined #openstack-lbaas | 07:55 | |
| *** kobis has joined #openstack-lbaas | 08:04 | |
| *** yamamoto has joined #openstack-lbaas | 08:32 | |
| rm_work | anywho there's some percentage chance that i actually made things better... | 08:52 |
| rm_work | had to switch all the test data to be bytes since we changed the open() to use rb mode | 08:52 |
| rm_work | so the tests were exactly wrong | 08:52 |
| nmagnezi | rm_work, still up? :) | 09:18 |
| rm_work | yeah waiting to see the result of this run | 09:19 |
| nmagnezi | rm_work, nice you see that you moved to me timezone :D | 09:19 |
| rm_work | lol | 09:20 |
| rm_work | ooo got further | 09:20 |
| nmagnezi | rm_work, i'm looking for some patches to review, anything with a high prio that I can help with? | 09:20 |
| rm_work | if we could get the API stuff working it'd be good... | 09:20 |
| rm_work | all the various v2 api alignment patches | 09:20 |
| nmagnezi | aye | 09:21 |
| nmagnezi | a silly question possibly, but still | 09:21 |
| rm_work | i just did this today: https://review.openstack.org/#/c/429926/ | 09:21 |
| nmagnezi | how can I test this code? say I deploy a devstack node | 09:21 |
| nmagnezi | and I want to see this in action | 09:21 |
| rm_work | hmmm | 09:21 |
| rm_work | i think they might all be in a chain | 09:21 |
| rm_work | i was actually looking at how to do that myself | 09:21 |
| rm_work | if so, just checkout the last review in the chain... but i don't know if they're all set up that way | 09:22 |
| M3m0r3x | xgerman, just testet the anti-aff. feature and the filter is not using different az - any ideas how to solve this? | 09:22 |
| nmagnezi | assuming we will have the chain in place, how can I send an API request directly? (curl?..) | 09:22 |
| rm_work | uhh | 09:23 |
| rm_work | yeah | 09:23 |
| nmagnezi | rm_work, shall I send an initial commit to a new repo named python-octaviaclient? :D | 09:24 |
| rm_work | i was actually going to basically get the curl command by doing 'neutron lbaas-loadbalancer-create --debug' and then just copy the command from there | 09:24 |
| *** Alex_Staf has joined #openstack-lbaas | 09:24 | |
| rm_work | and direct it to octavia instead | 09:24 |
| rm_work | oh, great | 09:25 |
| rm_work | johnsom: we ran into that project-id bug in the py35 tests | 09:25 |
| rm_work | johnsom: http://logs.openstack.org/90/424890/15/check/gate-octavia-v1-dsvm-py3x-scenario-ubuntu-xenial-nv/9919eaa/logs/screen-o-cw.txt.gz#_2017-02-07_07_59_50_388 | 09:25 |
| rm_work | that's the one i reported a while back | 09:25 |
| rm_work | pretty sure | 09:25 |
| *** gongysh has quit IRC | 09:29 | |
| *** strigazi1AFK is now known as strigazi | 09:29 | |
| *** gongysh has joined #openstack-lbaas | 09:29 | |
| *** gongysh has quit IRC | 09:30 | |
| *** yamamoto has quit IRC | 09:30 | |
| rm_work | maybe | 09:32 |
| rm_work | will check when I wake up | 09:33 |
| rm_work | uhh, hello bot? | 09:33 |
| *** yamamoto has joined #openstack-lbaas | 09:33 | |
| *** yamamoto has quit IRC | 09:39 | |
| nmagnezi | rm_work, ack. so I will try to use the same method you mentioned (--debug and aim to Octavia) right after I finish with https://review.openstack.org/#/c/429398/ . Michael asked my to check a kind of a different approach here with the way we feed haproxy with config files | 09:41 |
| nmagnezi | s/my/me | 09:42 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Fix py3x gates https://review.openstack.org/424890 | 09:42 |
| rm_work | ah yeah | 09:42 |
| rm_work | kk | 09:42 |
| nmagnezi | rm_work, ^^ how is that progressing? | 09:42 |
| rm_work | i think we're almost there | 09:42 |
| rm_work | though i may have really mucked things about today | 09:42 |
| nmagnezi | rm_work, i wanted to help with this but I saw you guys are very active about it so i didn't want to duplicate the effort | 09:43 |
| rm_work | not thinking straight, lots of cold meds :P | 09:43 |
| rm_work | yeah we'll see | 09:43 |
| rm_work | i'm hopeful about this run | 09:43 |
| rm_work | but will have to check it in the morning | 09:43 |
| * nmagnezi crossing fingers | 09:43 | |
| nmagnezi | rm_work, http://blogs.mathworks.com/developer/files/runALL_THE_TESTS.jpg | 09:44 |
| nmagnezi | rm_work, good night :) | 09:44 |
| rm_work | night | 09:44 |
| *** Dave___ is now known as Dave | 09:52 | |
| *** jschwarz has quit IRC | 10:01 | |
| *** jschwarz has joined #openstack-lbaas | 10:01 | |
| openstackgerrit | Zhixin Li proposed openstack/neutron-lbaas master: Don not create loadbalancer if no eligible lbaasv2 agent https://review.openstack.org/430141 | 10:02 |
| *** jsheeren has joined #openstack-lbaas | 10:28 | |
| *** jsheeren has quit IRC | 10:31 | |
| *** sanfern has quit IRC | 11:19 | |
| *** ducttape_ has joined #openstack-lbaas | 11:20 | |
| *** ducttape_ has quit IRC | 11:24 | |
| *** RichB has joined #openstack-lbaas | 11:38 | |
| *** jschwarz has quit IRC | 12:00 | |
| *** links has quit IRC | 12:48 | |
| *** Alex_Staf has quit IRC | 12:50 | |
| *** catintheroof has joined #openstack-lbaas | 12:50 | |
| *** ducttape_ has joined #openstack-lbaas | 12:51 | |
| *** ducttape_ has quit IRC | 12:55 | |
| *** ducttape_ has joined #openstack-lbaas | 13:08 | |
| *** sanfern has joined #openstack-lbaas | 13:11 | |
| *** chlong has joined #openstack-lbaas | 13:14 | |
| *** dileepr has quit IRC | 13:19 | |
| *** cody-somerville has joined #openstack-lbaas | 13:32 | |
| openstackgerrit | Abed Abu dbai proposed openstack/octavia master: ACTIVE-ACTIVE - tmp commit not for release https://review.openstack.org/429369 | 13:36 |
| *** amotoki has joined #openstack-lbaas | 13:44 | |
| *** ducttape_ has quit IRC | 13:44 | |
| *** matt-borland has joined #openstack-lbaas | 14:18 | |
| *** ducttape_ has joined #openstack-lbaas | 14:49 | |
| *** fnaval has joined #openstack-lbaas | 14:57 | |
| *** cody-somerville has quit IRC | 15:01 | |
| *** anilvenkata has quit IRC | 15:30 | |
| *** armax has joined #openstack-lbaas | 15:34 | |
| *** armax has quit IRC | 15:34 | |
| *** armax has joined #openstack-lbaas | 15:35 | |
| *** RichB has quit IRC | 15:46 | |
| *** fishbone has quit IRC | 15:56 | |
| openstackgerrit | Shashank Kumar Shankar proposed openstack/octavia master: Introduce Octavia v2 API for pools https://review.openstack.org/405922 | 15:57 |
| rm_work | johnsom: hmmm no errors in o-cw, looks good? but http://logs.openstack.org/90/424890/17/check/gate-octavia-v1-dsvm-py3x-scenario-ubuntu-xenial-nv/1ab3d98/console.html#_2017-02-07_10_26_50_627839 | 16:09 |
| rm_work | not sure what happened | 16:09 |
| johnsom | Just sat down to take a look | 16:09 |
| johnsom | rm_work I can take a pass at this | 16:11 |
| rm_work | johnsom: so umm | 16:11 |
| rm_work | johnsom: my changes NEEEED review | 16:12 |
| rm_work | I was pretty out of it | 16:12 |
| rm_work | but it seems to have mostly worked :P | 16:12 |
| johnsom | Grin | 16:12 |
| johnsom | Yeah | 16:12 |
| rm_work | I almost threw the whole class out and pulled in https://pypi.python.org/pypi/pem | 16:13 |
| rm_work | but it ... is actually more thorough than pem <_< | 16:14 |
| *** rcernin has quit IRC | 16:16 | |
| johnsom | Yeah, and you would have had one of those fun global requirements fights | 16:16 |
| *** pcaruana has quit IRC | 16:17 | |
| rm_work | yes :/ | 16:17 |
| *** fishbone has joined #openstack-lbaas | 16:22 | |
| openstackgerrit | Merged openstack/neutron-lbaas master: Turn on q-lbaasv2 service by default https://review.openstack.org/409803 | 16:33 |
| *** openstackgerrit has quit IRC | 16:35 | |
| johnsom | You guys are interesting.... | 16:36 |
| johnsom | We don't "enable_service in the settings file in octavia | 16:36 |
| xgerman | ha | 16:36 |
| *** cody-somerville has joined #openstack-lbaas | 16:36 | |
| *** cody-somerville has quit IRC | 16:36 | |
| *** cody-somerville has joined #openstack-lbaas | 16:36 | |
| johnsom | That patch just doesn't make any sense to me, but ok.... | 16:37 |
| *** csomerville has quit IRC | 16:38 | |
| xgerman | yeah, there are a few other patches we can consider in LBaaS | 16:41 |
| johnsom | Yeah, tread really lightly there | 16:41 |
| johnsom | We need to decide on the freeze over there | 16:41 |
| rm_work | ah, xgerman didn't wait for you johnsom :P | 16:41 |
| xgerman | well, most of them are the bot | 16:42 |
| xgerman | we should approve those to get stable brnaches IHMO | 16:42 |
| johnsom | Yeah, we have to wait on at least one of those as the URL for upper constraints doesn't work yet | 16:42 |
| xgerman | ok | 16:42 |
| johnsom | Yeah, still broken | 16:43 |
| johnsom | We are waiting for some of the projects to finish their stable cut last time I checked | 16:44 |
| johnsom | I haven't checked my e-mail today yet | 16:44 |
| *** openstackgerrit has joined #openstack-lbaas | 16:46 | |
| openstackgerrit | Michael Johnson proposed openstack/octavia master: Fix py3x gates https://review.openstack.org/424890 | 16:46 |
| rm_work | johnsom: find something? | 16:46 |
| rm_work | lol bad test? | 16:47 |
| johnsom | Yeah, I understand one of the errors, but not the other, so I pushed the fix for the error that I understood | 16:47 |
| johnsom | Yeah, not sure why py27 passed on that, but clearly a bad test | 16:47 |
| rm_work | huh. yeah | 16:48 |
| *** sanfern has quit IRC | 16:49 | |
| *** eezhova has quit IRC | 16:50 | |
| *** eezhova has joined #openstack-lbaas | 16:59 | |
| rm_work | wait i only saw one | 17:01 |
| rm_work | johnsom: what did you think of just *removing* that check to fix that bug? >_> | 17:02 |
| rm_work | also why did THAT not happen on py27... | 17:02 |
| *** kobis1 has joined #openstack-lbaas | 17:09 | |
| *** kobis has quit IRC | 17:12 | |
| johnsom | I don't like removing the checks, but I also know we need to get our project ID story straight still | 17:13 |
| rm_work | if we require one to delete, we should require one to create | 17:13 |
| rm_work | oh hmm i wonder if the test is just bugged and doesn't pass it correctly on the delete (but did on the create?) | 17:14 |
| rm_work | didn't think of that last night | 17:14 |
| *** kobis1 has quit IRC | 17:14 | |
| *** eezhova has quit IRC | 17:23 | |
| rm_work | johnsom: http://logs.openstack.org/90/424890/18/check/gate-octavia-v1-dsvm-py3x-scenario-ubuntu-xenial-nv/a932caa/console.html#_2017-02-07_17_21_19_317223 | 17:24 |
| rm_work | so it sees even request counts >_> 125 each | 17:24 |
| rm_work | why does it think it failed | 17:24 |
| johnsom | Yeah, that is the part I didn't understand | 17:25 |
| rm_work | OH | 17:26 |
| rm_work | I GET IT | 17:26 |
| rm_work | qur0q98hrc8mfijw | 17:26 |
| rm_work | k | 17:26 |
| rm_work | string issue again | 17:26 |
| rm_work | if member not in list(counters): | 17:27 |
| rm_work | I bet one is bytes and one is str | 17:27 |
| rm_work | i'm just not sure which is which | 17:27 |
| johnsom | Well, that test was still wrong | 17:27 |
| rm_work | k | 17:27 |
| rm_work | but can you run the scenario locally and figure out which one is bytes so we can convert? | 17:27 |
| rm_work | i'm try to read through and see if it's obvious | 17:28 |
| johnsom | Sure, just restarted so I need to restack my py35 stack | 17:28 |
| rm_work | k | 17:29 |
| rm_work | yeah it's hard to tell | 17:29 |
| rm_work | but I KNOW that's it | 17:29 |
| rm_work | oh derp | 17:29 |
| rm_work | i see it right there in the output | 17:29 |
| rm_work | Member b'server2_0' saw 125 | 17:30 |
| rm_work | lol | 17:30 |
| rm_work | ['server1_0', 'server1_1'] -> [b'server1_0', b'server1_1'] | 17:30 |
| rm_work | and the same in the test | 17:30 |
| rm_work | i can do it unless you have pending changes | 17:31 |
| johnsom | No go ahead, but there are two tests to fix, same check | 17:31 |
| johnsom | just grep for the check method in that directory | 17:31 |
| rm_work | yeah i did | 17:31 |
| rm_work | uhh i'm going to be defensive in addition | 17:32 |
| rm_work | actually i'll JUST be defensive, easier | 17:35 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Fix py3x gates https://review.openstack.org/424890 | 17:36 |
| rm_work | members = map(lambda x: six.b(x) if type(x) == six.text_type else x, members) | 17:37 |
| rm_work | hmm or should i do it the other way around | 17:38 |
| rm_work | is {}.keys() not portable? | 17:40 |
| johnsom | Oye, don't know | 17:40 |
| rm_work | seems clearer to me than list({}) to get keys | 17:40 |
| *** eezhova has joined #openstack-lbaas | 17:40 | |
| rm_work | ah well, not going to futz with it too much more | 17:40 |
| rm_work | uhh i think i see another bug | 17:42 |
| rm_work | scenario/base.py line 705 | 17:42 |
| rm_work | counters = self._send_requests(self.vip_ip, ["server1", "server2"]) | 17:42 |
| *** eezhova has quit IRC | 17:42 | |
| rm_work | wtf is that | 17:43 |
| rm_work | def _send_requests(self, vip_ip, path='') | 17:43 |
| rm_work | so "path" is ["server1", "server2"] ??? | 17:43 |
| rm_work | server = urllib2.urlopen("http://{0}/{1}".format(vip_ip, path), None, 2).read() | 17:43 |
| johnsom | Yeah, I fixed one of those as well, must have missed that one | 17:43 |
| rm_work | so is it supposed to iterate? | 17:44 |
| rm_work | what is this even doing | 17:45 |
| rm_work | i think that's going to get hit by the bytes issue too | 17:45 |
| rm_work | if member == 'server1' | 17:45 |
| rm_work | i seriously am not sure what this is supposed to be doing | 17:46 |
| *** eezhova has joined #openstack-lbaas | 17:49 | |
| rm_work | uhh does py35 even have urllib2 | 17:49 |
| rm_work | oh nm i see, we pull it from six | 17:50 |
| rm_work | oh, literally nothing uses _traffic_validation_after_stopping_server | 17:51 |
| rm_work | k. | 17:51 |
| rm_work | that explains it | 17:51 |
| johnsom | You will note, I reworked a bunch of this stuff for the octavia v2 api tests. Though I suspect the same py3 gremlins are in that patch as I didn't take that into account while working on it. | 17:52 |
| rm_work | k | 17:52 |
| rm_work | we'll get to that when we get to it i guess :P | 17:53 |
| rm_work | but i'm feeling good about this run again | 17:53 |
| johnsom | Cool. Yeah, looking at the base.py, I kind of want to restructure a bunch of it. | 17:54 |
| openstackgerrit | Merged openstack/neutron-lbaas master: Update reno for stable/ocata https://review.openstack.org/428751 | 17:54 |
| johnsom | It could be much simpler | 17:54 |
| rm_work | i've found a lot of stuff that could be simplified | 17:54 |
| rm_work | but only so much time, right? :P | 17:54 |
| johnsom | Yep | 17:54 |
| johnsom | Ok, I will pull that down and test local too | 17:55 |
| johnsom | Ugh, the topic changed | 17:57 |
| rm_work | ? | 18:00 |
| johnsom | On the py35 patch. | 18:00 |
| rm_work | ah | 18:00 |
| johnsom | There is a community goal for that, so it is supposed to be tagged special | 18:01 |
| rm_work | they changed it? :/ | 18:01 |
| johnsom | https://governance.openstack.org/tc/goals/pike/python35.html | 18:01 |
| *** tesseract has quit IRC | 18:01 | |
| rm_work | still says goal-python35 ? | 18:01 |
| johnsom | I changed it back | 18:01 |
| rm_work | oh. | 18:01 |
| rm_work | woo my ipv6 actually *works*, crazy | 18:02 |
| * rm_work watches the test run | 18:03 | |
| rm_work | seriously, where can I submit a change for "... this takes 10 - 15 minutes" to make it more accurate (30 - 45 minutes) | 18:05 |
| openstackgerrit | Sindhu Devale proposed openstack/octavia master: Introduce Octavia v2 API for Health Monitor https://review.openstack.org/406256 | 18:05 |
| rm_work | found it | 18:06 |
| rm_work | - echo "... this takes 10 - 15 minutes (logs in logs/devstacklog.txt.gz)" | 18:09 |
| rm_work | + echo "... this can take up to 45 minutes (logs in logs/devstacklog.txt.gz)" | 18:09 |
| rm_work | I wonder if that'd be considered trolling | 18:09 |
| rm_work | WHAT, failed already <_< | 18:10 |
| rm_work | https://review.openstack.org/#/c/430393/ | 18:11 |
| johnsom | Ok, locally I get errors that I understand why they happened | 18:13 |
| rm_work | oh damnit i knew that | 18:13 |
| rm_work | i forgot to do the binary safety check the other way | 18:13 |
| rm_work | wait no, it SHOULD have worked? | 18:15 |
| rm_work | `members` should be bytes | 18:15 |
| rm_work | `member` is coming from the counters list, and from the output we see it is also bytes | 18:15 |
| rm_work | OHHHHH ffff | 18:16 |
| rm_work | my baf | 18:16 |
| rm_work | *my bad | 18:16 |
| rm_work | forgot to list() the map | 18:17 |
| rm_work | so it emptied | 18:17 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Fix py3x gates https://review.openstack.org/424890 | 18:17 |
| rm_work | ^^ | 18:17 |
| rm_work | fixed | 18:17 |
| rm_work | lol having the bugfix in there constantly changes the topic, awesome >_> | 18:18 |
| rm_work | fixed it back | 18:18 |
| rm_work | johnsom: had you found MORE issues locally? or just that one | 18:19 |
| johnsom | Probably that one, just restarted my local test with your code | 18:19 |
| rm_work | k | 18:21 |
| rm_work | grab my fix and retest? | 18:22 |
| *** faizy has joined #openstack-lbaas | 18:22 | |
| johnsom | yep, running | 18:22 |
| johnsom | {1} octavia.tests.tempest.v1.scenario.test_load_balancer_tree_minimal.TestLoadBalancerTreeMinimal.test_load_balancer_tree_minimal [265.994918s] ... ok | 18:24 |
| rm_work | sweet. | 18:24 |
| rm_work | i forgot mapping makes a one-time generator | 18:25 |
| rm_work | T_T | 18:25 |
| johnsom | Both passed lcoal | 18:26 |
| johnsom | scenario: commands succeeded | 18:27 |
| johnsom | congratulations :) | 18:27 |
| rm_work | sweet | 18:27 |
| rm_work | so that COULD be it? | 18:27 |
| rm_work | py35 support? :P | 18:27 |
| rm_work | xgerman: WTB review :P | 18:28 |
| johnsom | Now, when I run tox under py3 I get a bunch of out-of-order imports due to concurrent futures. | 18:32 |
| johnsom | I was going to look into that too | 18:32 |
| johnsom | Though, currently our gate only runs pep8 under py2 | 18:32 |
| rm_work | hmm | 18:32 |
| johnsom | functional bombs too | 18:34 |
| johnsom | sigh | 18:34 |
| rm_work | gluh. | 18:37 |
| rm_work | ok testing | 18:37 |
| johnsom | We can do that in a followup though right? | 18:37 |
| rm_work | I guess so? | 18:39 |
| rm_work | if we don't forget lol | 18:39 |
| rm_work | should we make a NV job for functional-py35? | 18:39 |
| johnsom | Yes, working on that too | 18:39 |
| rm_work | kk | 18:39 |
| *** eezhova has quit IRC | 18:40 | |
| *** ducttape_ has quit IRC | 18:41 | |
| johnsom | I also want to put the tests for project ID back, so I'm going to poke at that for a few minutes | 18:42 |
| rm_work | k | 18:42 |
| rm_work | hmmm wtf, something is killing the API in the functional tests? | 18:49 |
| johnsom | Yeah | 18:49 |
| rm_work | everything starts getting 403s | 18:49 |
| johnsom | I hate debugging that too, it's slow | 18:49 |
| johnsom | There is a 500 in there for good measure | 18:49 |
| rm_work | AHHHH | 18:50 |
| rm_work | LOL | 18:50 |
| rm_work | test_create_over_quota does its thing | 18:50 |
| rm_work | and doesn't clean up | 18:50 |
| johnsom | Ah, because there is no project_id... | 18:51 |
| rm_work | hah | 18:51 |
| rm_work | it's a project-id problem | 18:51 |
| rm_work | no, there is | 18:51 |
| johnsom | So, scenario passed with the project ID stuff back in there | 18:51 |
| rm_work | py35 ACTUALLY WORKS for project-id | 18:51 |
| rm_work | py27 just seems to not | 18:51 |
| rm_work | this test is wrong | 18:51 |
| rm_work | got it | 18:52 |
| rm_work | yeah ok | 18:52 |
| rm_work | fixed | 18:52 |
| rm_work | had to add: self.addCleanup(self.check_quota_met_true_mock.stop | 18:53 |
| rm_work | i am super curious why these things are passing in py27 | 18:53 |
| rm_work | can I update it on this patch? | 18:53 |
| rm_work | or do you want to merge this | 18:53 |
| johnsom | No, go ahead | 18:53 |
| rm_work | k | 18:53 |
| rm_work | looking at the 500 as well | 18:53 |
| rm_work | probably a b thing | 18:54 |
| johnsom | It's that supposed to stop the mock when it goes out of scope? | 18:54 |
| rm_work | yes | 18:54 |
| johnsom | I will look at your change, curious. That was my code, so ... | 18:54 |
| johnsom | Curious to learn... | 18:55 |
| rm_work | ANY TIME you do a start() on a mock | 18:55 |
| rm_work | please add the stop to cleanup | 18:55 |
| rm_work | i might just go hunting for those... | 18:55 |
| rm_work | yeah same in more places | 18:55 |
| rm_work | almost done | 19:00 |
| rm_work | kk, running pep8 and functional-py27 again | 19:01 |
| rm_work | and then we should be good | 19:01 |
| rm_work | seems all passing | 19:02 |
| openstackgerrit | Nakul Dahiwade proposed openstack/octavia master: Align Octavia API to n-lbaasv2 for L7Policy https://review.openstack.org/406328 | 19:02 |
| johnsom | Good passes on the old patch, so making progress! | 19:04 |
| openstackgerrit | Adam Harwell proposed openstack/octavia master: Fix py3x gates and functional https://review.openstack.org/424890 | 19:06 |
| rm_work | kk | 19:06 |
| rm_work | johnsom: should be good :) | 19:07 |
| johnsom | Ok, thanks | 19:07 |
| *** ducttape_ has joined #openstack-lbaas | 19:10 | |
| *** faizy has quit IRC | 19:16 | |
| xgerman | rm_work link for what I should review… | 19:17 |
| rm_work | ^^ most recent | 19:17 |
| *** eezhova has joined #openstack-lbaas | 19:22 | |
| rm_work | i just went through and re-reviewed and it seems good to me | 19:24 |
| rm_work | actually not as horrific as i thought it would be last night, lol | 19:24 |
| johnsom | Yeah, so far good local. Just want to look at the project ID change | 19:24 |
| rm_work | kk | 19:24 |
| rm_work | after lunch I probably need to disappear back into downstream-land for a little bit | 19:25 |
| johnsom | Yeah, no problem. Thanks for your help here | 19:25 |
| *** dileepr has joined #openstack-lbaas | 19:27 | |
| *** ducttape_ has quit IRC | 19:31 | |
| rm_work | i guess I'll poke you again later about https://review.openstack.org/#/c/429926/ and https://review.openstack.org/#/c/424816/ | 19:34 |
| *** ducttape_ has joined #openstack-lbaas | 19:54 | |
| rm_work | erg so why can't we use https://github.com/openstack/nova-lxd again? | 20:01 |
| * rm_work veers wildly offtopic | 20:01 | |
| johnsom | rm_work It's on my list of stuff to look at again for pike. First issue we had was config_drive was non-functional in nova-lxd. That has been fixed. | 20:03 |
| johnsom | Last time someone tried nova-lxd out, the lb-mgmt-net interface was not coming up in the container on boot. | 20:03 |
| rm_work | ah. | 20:04 |
| johnsom | It's probably about time to give it another spin | 20:04 |
| johnsom | rm_work FYI: https://bugs.launchpad.net/nova-lxd/+bug/1632814 | 20:06 |
| openstack | Launchpad bug 1632814 in nova-lxd "Build container - no networks connected" [Undecided,New] | 20:06 |
| johnsom | Though I'm not sure people are looking at the bugs.... | 20:06 |
| rm_work | k | 20:06 |
| openstackgerrit | Sindhu Devale proposed openstack/octavia master: Introduce Octavia v2 API for Health Monitor https://review.openstack.org/406256 | 20:09 |
| openstackgerrit | Nakul Dahiwade proposed openstack/octavia master: Align Octavia API to n-lbaasv2 for L7Policy https://review.openstack.org/406328 | 20:11 |
| *** ducttape_ has quit IRC | 20:14 | |
| *** ducttape_ has joined #openstack-lbaas | 20:14 | |
| rm_work | johnsom: on https://review.openstack.org/#/c/429539/ isn't that a copy/pasted file? do we fix things in it? | 20:16 |
| johnsom | It is, but it's broken. pep8 under python3 complains | 20:17 |
| rm_work | :/ | 20:17 |
| rm_work | k | 20:17 |
| johnsom | Also note, two lines above we have customized it | 20:17 |
| johnsom | Oh, and that was a change I made to that file... Trying to give a clear error when people try to use alembic without our tool | 20:18 |
| rm_work | ah >_< | 20:18 |
| *** cody-somerville has quit IRC | 20:31 | |
| *** cody-somerville has joined #openstack-lbaas | 20:31 | |
| *** cody-somerville has quit IRC | 20:31 | |
| *** cody-somerville has joined #openstack-lbaas | 20:31 | |
| *** ducttape_ has quit IRC | 20:31 | |
| *** ducttape_ has joined #openstack-lbaas | 20:43 | |
| *** amotoki has quit IRC | 20:51 | |
| rm_work | johnsom: so, if you magically have some time when you aren't really busy, let me know, would like to chat re: AAP | 20:51 |
| rm_work | but, right now I guess I'll grab lunch | 20:54 |
| johnsom | Ok, just sat down to eat lunch, will ping after | 20:54 |
| rm_work | lol | 20:54 |
| rm_work | good timing | 20:54 |
| *** ducttape_ has quit IRC | 20:59 | |
| *** ducttape_ has joined #openstack-lbaas | 20:59 | |
| *** ducttape_ has quit IRC | 21:01 | |
| openstackgerrit | Merged openstack/octavia master: Remove white space between print () https://review.openstack.org/429539 | 21:04 |
| *** harlowja has quit IRC | 21:09 | |
| xgerman | johnsom so for that proxy — some error messages are pretty specific right now (e.g. loadbalancerv2.TLSDefaultContainerNotSpecified) so I would need to scrape the error message in the proxy to ttanslate to that exception or just use generic errors | 21:19 |
| xgerman | I feel generic errors are better since we are planning to localize? | 21:19 |
| johnsom | So, these are exceptions generated in the lbaas API? | 21:27 |
| johnsom | I think we need to make sure we send back the same HTTP codes | 21:28 |
| johnsom | It seems like this is still a pass through situation | 21:28 |
| xgerman | that, true, but what’s the code for TLSDfaultContainer not specified | 21:29 |
| xgerman | likely a 400 catch all | 21:29 |
| xgerman | so I would need to scrape the body | 21:29 |
| *** eezhova has quit IRC | 21:35 | |
| *** harlowja has joined #openstack-lbaas | 21:35 | |
| *** yamamoto has joined #openstack-lbaas | 21:43 | |
| *** catintheroof has quit IRC | 21:48 | |
| *** catintheroof has joined #openstack-lbaas | 21:49 | |
| blogan | ndahiwade: did that fix your issue? what i showed you earlier? | 21:51 |
| ndahiwade | blogan, yes it did :) | 21:53 |
| blogan | ndahiwade: awesome! | 21:53 |
| blogan | sshank: you were mentioning in the hangout you were seeing issues? | 21:53 |
| *** catintheroof has quit IRC | 21:54 | |
| ndahiwade | so blogan, now we have to initiate all controllers in the LBaasController class? | 21:54 |
| *** matt-borland has quit IRC | 21:54 | |
| ndahiwade | so that it hits /v2.0/lbaas/url for the functional tests | 21:54 |
| blogan | ndahiwade: actually, you should do it in the LBaaSCotnroller class AND the V2Controller class | 21:55 |
| blogan | so it its both /v2.0/lbaas/resource and /v2.0/resource | 21:55 |
| ndahiwade | ahh okay thanks:) | 21:55 |
| *** ipsecguy has quit IRC | 21:55 | |
| blogan | ndahiwade: ah just noticed, the BaseV2Controller is having those controllers added to, when it should just be the V2Controller | 21:58 |
| openstackgerrit | Sindhu Devale proposed openstack/octavia master: DB migration to add ID column to HM https://review.openstack.org/430451 | 21:58 |
| johnsom | rm_work available when you are | 21:58 |
| openstackgerrit | Sindhu Devale proposed openstack/octavia master: [Super WIP] DB migration to add ID column to HM https://review.openstack.org/430451 | 21:59 |
| sshank | blogan, Still working on it. I wasn't removing the root tags in the response as well. | 21:59 |
| *** tesseract has joined #openstack-lbaas | 21:59 | |
| *** tesseract- has joined #openstack-lbaas | 22:00 | |
| rm_work | johnsom: cool | 22:02 |
| rm_work | johnsom: just got free | 22:02 |
| rm_work | johnsom: don't suppose you have zoom? :P | 22:02 |
| *** ducttape_ has joined #openstack-lbaas | 22:02 | |
| johnsom | zoom? Like Mazda zoom-zoom? | 22:03 |
| rm_work | heh | 22:03 |
| ndahiwade | blogan, Okay,will change that:) | 22:03 |
| openstackgerrit | Brandon Logan proposed openstack/octavia master: Clean up base v2 controllers https://review.openstack.org/430455 | 22:04 |
| *** yamamoto has quit IRC | 22:07 | |
| *** tesseract has quit IRC | 22:07 | |
| *** tesseract- has quit IRC | 22:07 | |
| openstackgerrit | Brandon Logan proposed openstack/octavia master: Clean up base v2 controllers https://review.openstack.org/430455 | 22:08 |
| *** tesseract has joined #openstack-lbaas | 22:08 | |
| openstackgerrit | Brandon Logan proposed openstack/octavia master: Clean up base v2 controllers https://review.openstack.org/430455 | 22:10 |
| *** yamamoto has joined #openstack-lbaas | 22:12 | |
| *** ducttape_ has quit IRC | 22:13 | |
| blogan | johnsom: https://github.com/openstack/neutron-lbaas/commit/4455759f4506a43d70f811a32ee60b13af6afd8d | 22:16 |
| blogan | i had to see for myself | 22:16 |
| blogan | i see the point, its just giong to be confusing to end users that they can't get an IP in a range they expect | 22:17 |
| *** tesseract is now known as tesseract-RH | 22:26 | |
| *** tesseract-RH is now known as tesseract | 22:31 | |
| *** ducttape_ has joined #openstack-lbaas | 22:32 | |
| *** ducttape_ has quit IRC | 22:56 | |
| *** tesseract has quit IRC | 23:00 | |
| *** tesseract has joined #openstack-lbaas | 23:00 | |
| rm_work | blogan: you disagreeing with that? :P | 23:01 |
| *** tesseract-RH has joined #openstack-lbaas | 23:01 | |
| rm_work | blogan: was that prompted by my octavia version? | 23:01 |
| *** tesseract has quit IRC | 23:01 | |
| *** tesseract-RH has quit IRC | 23:01 | |
| blogan | rm_work: yeah | 23:08 |
| blogan | rm_work: and im indifferent twoards it, bc its a feature i wanted in the beginning as well but then i realized the lbaas service would either have to choose the subnet that the vip comes from OR allow neutron to decide, which means it can choose to do many vips but we can only show one | 23:10 |
| rm_work | right | 23:11 |
| rm_work | in my case, I *will* be using it to provision FLIPs | 23:11 |
| rm_work | which are network-only anyway | 23:11 |
| rm_work | blogan: I note that you didn't actually COMMENT on my patch | 23:16 |
| blogan | rm_work: i have to admit, i didn't even look at your patch | 23:18 |
| * blogan looks now | 23:19 | |
| rm_work | lol what | 23:20 |
| rm_work | my patch caused you to look into n-lbaas code but not actually AT my patch? rofl | 23:20 |
| blogan | rm_work: lol well really the conversation about your comment spurred me to look at the vip_network_id attribute bc i did nto nkow that got added in | 23:21 |
| rm_work | blogan: BTW I couldn't actually find the code that validates the user passes enough info as part of the VIP object... do you know where it actually looks to make sure subnet is present? | 23:21 |
| blogan | rm_work: like where it connects to neutron to say it really exists? | 23:21 |
| rm_work | no | 23:22 |
| rm_work | I mean like | 23:22 |
| rm_work | for Octavia to reject a request that doesn't have enough info to actually make a vip | 23:22 |
| rm_work | like, it will reject if you pass a blank vip object, won't it? O_o | 23:22 |
| rm_work | or did we just ... not do a validation layer yet | 23:22 |
| blogan | ah, that'd be in the type or in the controller...preferrably the type..but in your case you want either vip.network_id or vip.subnet_id, so you can't make either one of them mandatory | 23:23 |
| blogan | so you'll have to resort to doing that in the controller layer, or validation layer | 23:23 |
| rm_work | yeah but does it do anything ALREADY? I don't see it | 23:24 |
| rm_work | it is mandatory already... the "vip" object at least... | 23:24 |
| rm_work | but inside that, you could pass a combo of information (or none?) that wouldn't be enough | 23:24 |
| rm_work | right? | 23:24 |
| blogan | doesn't look like it ever did it with the vip type, let me check the controller | 23:24 |
| rm_work | it already takes "subnet_id" OR "port_id / IP" | 23:24 |
| rm_work | i didn't see it T_T | 23:24 |
| blogan | yeah but ideally the subnet_id would say mandatory too but it doesnt...so i'm thinking its in the controller | 23:25 |
| blogan | or its just not there :) | 23:25 |
| rm_work | yep, verified, nothing | 23:25 |
| blogan | https://review.openstack.org/#/c/429926/5/octavia/api/v1/controllers/load_balancer.py | 23:25 |
| rm_work | just did one | 23:25 |
| blogan | yeah that would not fail it | 23:25 |
| blogan | so yeah, no vip validation before! | 23:25 |
| blogan | well i mean | 23:25 |
| blogan | you can pass in an emtpy vip | 23:25 |
| rm_work | yep | 23:26 |
| rm_work | so i guess I don't care as part of my patch either | 23:26 |
| rm_work | but i handled the case where you pass both | 23:26 |
| rm_work | so that's about the best I can do I think >_> | 23:26 |
| rm_work | unless I start ADDING further validation | 23:26 |
| rm_work | I *can* | 23:27 |
| rm_work | must have at least *one of* `(port_id and ip_address) or subnet_id or network_id` | 23:27 |
| rm_work | easy enough to add | 23:28 |
| blogan | i think just port_id is what neutron-lbaas did | 23:32 |
| blogan | i think it just chose the first fixed_ip when that happeend | 23:33 |
| *** ducttape_ has joined #openstack-lbaas | 23:40 | |
| *** ipsecguy has joined #openstack-lbaas | 23:43 | |
| *** ducttape_ has quit IRC | 23:50 | |
| *** ducttape_ has joined #openstack-lbaas | 23:51 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
