| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2017-01-11 | |||
| 16:45:09 | briancurtin | reedip_outofmemo: openstacksdk hasn’t explicitly returned dictionaries ever, though some time ago it did return objects that can act like dictionaries. compute hasn’t done that for probably 8 months | |
| 16:45:11 | reedip_outofmemo | briancurtin : like Network client | |
| 16:45:32 | briancurtin | reedip_outofmemo: I’m not sure, i don’t work on OSC. i think OSC only uses openstacksdk for the network related commands though, not for compute and volume | |
| 16:46:04 | reedip_outofmemo | stevemar : any suggestions related to above query ? | |
| 16:46:17 | dtroyer | reedip: briancurtin is correct re compute and volume quotas | |
| 16:46:41 | reedip_outofmemo | dtroyer : ok, so sdk is used only for Networks?????? | |
| 16:46:54 | dtroyer | the root problem is that OSC was not ready for the 0.9.11 changes in the SDK because our functioanl tests were producing false positives for a time | |
| 16:47:00 | dtroyer | reedip: yes | |
| 16:47:03 | reedip_outofmemo | briancurtin : one more query | |
| 16:47:09 | briancurtin | reedip_outofmemo: sure | |
| 16:47:17 | reedip_outofmemo | dtroyer :okay, then I will fix the bug in OSC and not SDK | |
| 16:47:27 | reedip_outofmemo | briancurtin : this is related to quota set | |
| 16:47:41 | dtroyer | OSc needs to work with sdk >= 0.9.10 for a while yet | |
| 16:47:59 | reedip_outofmemo | briancurtin : in openstack SDK , you updated the proxy2 | |
| 16:49:00 | reedip_outofmemo | briancurtin : https://github.com/openstack/python-openstacksdk/blob/master/openstack/proxy2.py#L64 | |
| 16:49:12 | briancurtin | correct, there is a proxy2 | |
| 16:49:28 | reedip_outofmemo | briancurtin : Neutron Quota set does not need the ID which is being passed here in the PUT request | |
| 16:50:56 | reedip_outofmemo | Neutron Quota API do not need the tenant/project ID as they have an endpoint in the quota resource itself for the tenant ID. For example. if Tenant ID is ABCDE , then Neutron has the quota endpoint for tenant as http://10.0.0.1:23423/v2/quotas/ABCDE.json | |
| 16:50:58 | briancurtin | so when building the request don’t make it require an ID. you might need to override the create method of the Quota resource and pass requires_id=False to the _prepare_request method https://github.com/openstack/python-openstacksdk/blob/master/openstack/resource2.py#L502 | |
| 16:51:16 | reedip_outofmemo | briancurtin : Yup , I am trying to do that | |
| 16:51:41 | reedip_outofmemo | briancurtin : Just wanted to update you. There might be one more change, I will put up the patch tomorrow | |
| 16:51:47 | briancurtin | reedip_outofmemo: ok cool | |
| 16:52:02 | reedip_outofmemo | briancurtin : thanks , hope this resolves soon :) | |
| 16:52:35 | reedip_outofmemo | dtroyer : sorry for bugging you , but if OSC needs to work with SDK >=0.9.10, then should I just put a type check ? | |
| 16:53:35 | reedip_outofmemo | dtroyer: that would be simpler and can be removed once everything comes in line? Currently network quota is returned as an object, but earlier it was returned as a dict ( or something similar to a dict) | |
| 16:55:58 | dtroyer | reedip: there is currently a bit of compatibility code in the networking commands for this reason, do the same thing in the quota command and leave a note to remove it later | |
| 16:56:42 | reedip_outofmemo | dtroyer: aye aye ! | |
| 17:00:24 | dtroyer | dhellmann, stevemar: I'd like to do a cliff release this week yet, likely last before freeze. I'm OK with https://review.openstack.org/412742, chime in when you get a minute if we should include it or not... | |
| 17:09:18 | stevemar | dtroyer: we can probably release without that patch, it's more for making sure our gates aren't busted | |
| 17:09:50 | stevemar | dtroyer: i've also released ksa for us, but getting blocked for the u-c bump | |
| 17:10:02 | stevemar | https://review.openstack.org/#/c/418608/ | |
| 17:36:24 | openstackgerrit | Reedip proposed openstack/python-openstackclient: Fix quota show output https://review.openstack.org/419067 | |
| 17:38:17 | reedip_outofmemo | stevemar , briancurtin : ^^ | |
| 17:38:26 | openstackgerrit | Shashank Kumar Shankar proposed openstack/python-openstackclient: Introduce Neutron DHCP agent commands to OSC. https://review.openstack.org/387611 | |
| 17:44:42 | openstackgerrit | Shashank Kumar Shankar proposed openstack/python-openstackclient: Introduce Neutron DHCP agent commands to OSC. https://review.openstack.org/387611 | |
| 18:30:03 | ankur-gupta-f | dtroyer: here? | |
| 18:30:55 | dtroyer | ankur-gupta-f: yo | |
| 18:31:06 | ankur-gupta-f | Client functional tests :D | |
| 18:31:14 | ankur-gupta-f | so we tried the regexes | |
| 18:31:16 | ankur-gupta-f | now JSON | |
| 18:31:28 | ankur-gupta-f | but for existing patches can we leave regexes and follow up with JSON | |
| 18:32:05 | dtroyer | yes. it may still evolve more, I don't want you guys to chase too much of a moving target, OSC does that all the time anyway | |
| 18:32:18 | dtroyer | ther are a couple of things I'd like to make sure get in theough, | |
| 18:32:25 | dtroyer | using addCleanup() is one of those | |
| 18:32:49 | dtroyer | that is more than a cosmentic/developer preference thing | |
| 18:33:05 | ankur-gupta-f | for everything created. | |
| 18:33:32 | ankur-gupta-f | I do like the change that every test does the creation, <list, show, etc..> deletion | |
| 18:34:24 | ankur-gupta-f | Okay thats all. Thanks | |
| 18:40:56 | dtroyer | stevemar: I ran the -efunctional-tips locally, no tests ran…do we need to include the ostestr command in that section? | |
| 18:52:59 | cdent | elmiko: that link you tweeted a while ago to radanalytics wants auth? | |
| 18:56:22 | openstackgerrit | Dean Troyer proposed openstack/python-openstackclient: add support for running tests with master dependencies https://review.openstack.org/419032 | |
| 19:00:06 | stevemar | dtroyer: thats a good question, the [fucntional] target doesn't have any ostestr stuff | |
| 19:00:53 | stevemar | dtroyer: do you know why we have to add ostestr to -tips but not the other? | |
| 19:01:13 | openstackgerrit | Merged openstack/python-openstackclient: Fix filter error in os volume list https://review.openstack.org/416852 | |
| 19:01:24 | dtroyer | I just added it to the command list. I think tox inherits from [env] if there isn't a command= | |
| 19:01:55 | thingee | dtroyer seems to be working now | |
| 19:12:53 | thingee | sorry I missed that | |
| 19:18:55 | dtroyer | heh, I did too until I actually tried it :) | |
| 19:20:30 | stevemar | dtroyer: ahh, but since there is a command= there, it doesn't inherit from env | |
| 19:29:17 | openstackgerrit | Diana Clarke proposed openstack/python-openstackclient: WIP - Add nova microversion 2.40 support https://review.openstack.org/419108 | |
| 19:40:17 | openstackgerrit | Merged openstack/cliff: Add Constraints support https://review.openstack.org/412742 | |
| 19:49:58 | openstackgerrit | Ankur proposed openstack/python-openstackclient: Add meter rule to OSC https://review.openstack.org/378058 | |
| 20:27:44 | stevemar | dtroyer: have you been keeping track of what networking commands broke and are fixed in osc? | |
| 20:39:26 | dtroyer | I don't have an exhaustive list… AFAIK only the qos rule type is left though? | |
| 20:39:37 | dtroyer | well, I haven't gone looking for the remaining skips yet | |
| 20:40:21 | dtroyer | erg, there are still 4 | |
| 20:41:19 | dtroyer | stevemar: do you recall how many tests were being skipped before this mess started? | |
| 20:41:40 | stevemar | only 4 (they were conditional on a networking extension) | |
| 20:41:58 | stevemar | we had no skipped tests otherwise | |
| 20:42:04 | dtroyer | ok then, your 418650 has them all then | |
| 20:46:17 | openstackgerrit | Shashank Kumar Shankar proposed openstack/python-openstackclient: Introduce Neutron DHCP agent commands to OSC. https://review.openstack.org/387611 | |
| 20:47:08 | openstackgerrit | Steve Martinelli proposed openstack/python-openstackclient: unskip network qos rule type functional test https://review.openstack.org/419129 | |
| 20:47:27 | stevemar | dtroyer: ^ i think that one worked, just a bad test | |
| 20:48:17 | dtroyer | I think it is more than a bad test, I get a 404 when I try the command | |
| 20:53:43 | dtroyer | #$^%&%$#$!@#$%^*%&$#@#$%^*&^&%^%$@#^$&%&^*&%^$%#$@#!@#$^%&^ | |
| 20:54:01 | dtroyer | Why did we let an extension command into the main repo?????????????????? | |
| 20:55:51 | dtroyer | so nevermind my 404, I don't have the (*&^%$#$%^&*( extension enabled | |
| 20:56:32 | dtroyer | we gotta handle that better… and by better I mean at all | |
| 20:56:45 | dtroyer | Is ther any way to make the tests discover and auto-skip this? | |
| 20:57:38 | stevemar | dtroyer: i thought there was a discover and skip mechanism | |
| 20:57:46 | stevemar | i guess not, since it didn't skip for you | |
| 20:58:33 | stevemar | dtroyer: that test and command was created by rich back in the early networking days, before we had a fleshed out plan for extensions and had the neutron team on board | |
| 21:00:09 | dtroyer | hmmm… git sez it was committed on Aug 12 | |
| 21:00:27 | dtroyer | by https://github.com/ralonsoh | |
| 21:01:06 | dtroyer | well, that was an auto-complete | |
| 21:01:17 | dtroyer | ralonsoh, may or may not be the same person | |
| 21:02:09 | stevemar | dtroyer: i was referring to https://github.com/openstack/python-openstackclient/commit/4f23a77de04bfdfafd4bf8f16b0365df7663e9e5 | |
| 21:02:51 | stevemar | dtroyer: skippage is here: https://github.com/openstack/python-openstackclient/blob/2026dee17e7b6b8d738345fd27dc905efb6561f1/openstackclient/tests/functional/network/v2/test_network_segment.py#L54-L66 | |
| 21:02:54 | dtroyer | oh, I thought we were talking about qos rule type still | |
| 21:05:43 | stevemar | dtroyer: i thought we were tlaking about skipped tests -- 418650 unskips all the skipped ones, except those network segment ones | |
| 21:06:05 | stevemar | i have no idea what qos rule type is | |
| 21:06:24 | dtroyer | that's the one test that failed in your 418650 test with sdk 0.9.12 | |
| 21:06:35 | dtroyer | I can't tell (yet) if it ever worked | |
| 21:07:02 | dtroyer | have to restart my devstack with q-qos running, and that takes a long time these days | |
| 21:07:30 | stevemar | dtroyer: wel https://review.openstack.org/#/c/419129/ fixes it, i think | |
| 21:07:31 | dtroyer | I believe this is the last blocker for sdk 0.9.12 | |
| 21:07:33 | stevemar | let me run it for you | |
| 21:08:11 | dtroyer | so is that a fix or just not testing for the missing piece? | |
| 21:08:30 | dtroyer | ie, is something still broken since minimum_bandwidth was expected to be ther at one time? | |
| 21:12:19 | stevemar | dtroyer: i'm wondering that myself, looking at the neutron APIs while I'm stacking with qos enabled | |