| *** openstackgerrit_ has joined #openstack-shade | 06:23 | |
| *** openstackgerrit_ has quit IRC | 06:25 | |
| *** openstackgerrit_ has joined #openstack-shade | 06:26 | |
| *** openstackgerrit_ has quit IRC | 06:27 | |
| *** openstackgerrit_ has joined #openstack-shade | 06:34 | |
| *** openstackgerrit_ has quit IRC | 06:35 | |
| *** yfried has joined #openstack-shade | 06:55 | |
| *** yolanda has joined #openstack-shade | 06:58 | |
| *** openstackgerrit_ has joined #openstack-shade | 06:59 | |
| *** openstackgerrit_ has quit IRC | 07:00 | |
| *** gouthamr has joined #openstack-shade | 07:43 | |
| *** abregman has joined #openstack-shade | 07:51 | |
| *** gouthamr has quit IRC | 08:00 | |
| *** cdent has joined #openstack-shade | 10:20 | |
| *** cdent has quit IRC | 10:33 | |
| *** cdent has joined #openstack-shade | 10:35 | |
| *** cdent has quit IRC | 11:40 | |
| *** abregman is now known as abregman|mtg | 12:14 | |
| *** abregman_ has joined #openstack-shade | 12:22 | |
| *** abregman_ has quit IRC | 12:25 | |
| *** abregman|mtg has quit IRC | 12:25 | |
| *** abregman has joined #openstack-shade | 12:25 | |
| *** abregman is now known as abregman|mtg | 12:39 | |
| *** cdent has joined #openstack-shade | 12:49 | |
| *** cdent_ has joined #openstack-shade | 12:51 | |
| *** cdent has quit IRC | 12:53 | |
| *** cdent_ is now known as cdent | 12:53 | |
| *** abregman|mtg has quit IRC | 13:53 | |
| *** abregman has joined #openstack-shade | 13:54 | |
| *** abregman is now known as abregman|mtg | 14:08 | |
| *** yolanda has quit IRC | 14:11 | |
| *** yolanda has joined #openstack-shade | 14:13 | |
| *** yolanda has quit IRC | 14:15 | |
| *** yolanda has joined #openstack-shade | 14:15 | |
| *** purplerbot has quit IRC | 14:18 | |
| *** abregman|mtg is now known as abregman | 14:44 | |
| *** purplerbot has joined #openstack-shade | 15:02 | |
| -openstackstatus- NOTICE: The Gerrit service on review.openstack.org is restarting now to address acute performance issues, and will be back online momentarily. | 15:07 | |
| *** cdent has quit IRC | 15:08 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Fix _neutron_available_floating_ips filtering https://review.openstack.org/409532 | 15:11 |
|---|---|---|
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Make available_floating_ips use normalized keys https://review.openstack.org/409824 | 15:11 |
| mordred | Shrews: ^^ I updated https://review.openstack.org/409532 - yucky typo | 15:11 |
| mordred | Shrews: it's a lovely typo isn't it? | 15:16 |
| Shrews | yup | 15:25 |
| *** abregman has quit IRC | 15:45 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add functional test to show floating ip reuse https://review.openstack.org/409842 | 15:47 |
| mordred | Shrews: also - look! I wrote tests :) | 15:47 |
| Shrews | what the heck is wrong with you? | 15:48 |
| Shrews | tests? geez | 15:48 |
| Shrews | it's a Festivus miracle | 15:48 |
| *** yfried has quit IRC | 15:49 | |
| Shrews | mordred: in test_reuse_auto_ip, what's the purpose for assertLessEqual checks both before AND after the create_server? | 15:56 |
| mordred | Shrews: create_server should not create a new fip in that case | 15:56 |
| mordred | Shrews: but we don't know if the cloud has fips at all - so we'll either start with 0 or 1 fips, and we darn-sure should not end up with 2 fips | 15:57 |
| mordred | we could probably be cleverer and save the len(fips) before the create and then assertEqual after it | 15:57 |
| Shrews | except you don't guarantee that. it could be 0 before the create and 1 after, and you'd still pass | 15:57 |
| mordred | yah | 15:57 |
| Shrews | we should probably do the cleverer thing, and maybe add a comment there | 15:58 |
| mordred | fix coming | 15:58 |
| mordred | yah | 15:58 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add functional test to show floating ip reuse https://review.openstack.org/409842 | 15:59 |
| Shrews | mordred: but... hrm | 15:59 |
| mordred | Shrews: hows that | 15:59 |
| Shrews | couldn't other tests interfere with that number? | 16:00 |
| Shrews | a race on list_floating_ips | 16:00 |
| Shrews | oh, _setup_networks might cover that | 16:02 |
| Shrews | yeah, nm | 16:02 |
| Shrews | i think we might need to filter on net in list_floating_ips | 16:04 |
| mordred | Shrews: well, we also run with concurrency=1 | 16:05 |
| Shrews | we do that still? | 16:05 |
| mordred | yah | 16:05 |
| Shrews | oh | 16:05 |
| mordred | but I can poke at filtering on network too | 16:05 |
| Shrews | probably should filter. one less thing to fix if we decide to change concurrency. | 16:07 |
| Shrews | i do like PS2 though. much clearer | 16:08 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add functional test to show floating ip reuse https://review.openstack.org/409842 | 16:14 |
| mordred | Shrews: cool. that should be better then ^^ | 16:15 |
| Shrews | lgtm | 16:16 |
| openstackgerrit | Merged openstack-infra/shade: Stop telling users to check logs https://review.openstack.org/409290 | 16:32 |
| *** cdent has joined #openstack-shade | 16:39 | |
| *** gouthamr has joined #openstack-shade | 17:08 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add test to trap for missing services https://review.openstack.org/409869 | 17:12 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add functional test to show floating ip reuse https://review.openstack.org/409842 | 17:46 |
| *** gouthamr has quit IRC | 17:50 | |
| *** gouthamr has joined #openstack-shade | 17:52 | |
| *** gouthamr has quit IRC | 17:52 | |
| *** openstack has joined #openstack-shade | 17:57 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add functional test to show floating ip reuse https://review.openstack.org/409842 | 18:50 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add functional test to show floating ip reuse https://review.openstack.org/409842 | 19:32 |
| Shrews | mordred: why are the gate-shade-dsvm-functional-libs-nv jobs consistently failing in your patch series? | 19:39 |
| mordred | Shrews: job config I think | 19:42 |
| Shrews | mordred: also, https://review.openstack.org/385194 | 19:42 |
| mordred | (should get fixed by the magnum patches) | 19:42 |
| Shrews | you remove python-glaceclient from requirements.txt, yet you still import it | 19:43 |
| Shrews | do i need more coffee? | 19:43 |
| Shrews | you move the import in openstackcloud.py | 19:43 |
| mordred | Shrews: it's imported on demand in def glance_client ... but it's not needed if you don't want a glanceclient.Client object | 19:44 |
| Shrews | ah, i see. we don't use it | 19:45 |
| Shrews | so why still offer it? | 19:45 |
| mordred | because we were silly and didn't mark it private? | 19:46 |
| *** gouthamr has joined #openstack-shade | 19:46 | |
| Shrews | but i think we explicitly say to not depend on those (somewhere) | 19:46 |
| mordred | yah. we do | 19:46 |
| mordred | Shrews: although I cannot find that place right now | 19:47 |
| Shrews | nor can i | 19:48 |
| Shrews | which makes me sad | 19:48 |
| mordred | honestly, if you're using cloud.glance_client something is going wrong | 19:48 |
| Shrews | maybe i imagined the thing | 19:49 |
| mordred | maybe it was in my shade talk? | 19:49 |
| Shrews | mordred: like i would have listened to that | 19:49 |
| Shrews | mordred: another response_status_code bug in https://review.openstack.org/385193 | 19:52 |
| mordred | Shrews: I see it | 19:53 |
| mordred | Shrews: I'm going to wait til you're at the end of that stack before I push up the fix to the first patch (try to reduce the patch-bomb) | 20:02 |
| mordred | thanks for looking at that, btw | 20:02 |
| *** cdent has quit IRC | 20:06 | |
| Shrews | mordred: done | 20:10 |
| Shrews | mordred: only one additional -1 | 20:10 |
| mordred | \o/ | 20:11 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Convert test_create_image_put_v2 to requests_mock https://review.openstack.org/407670 | 20:13 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Honor image_api_version when doing version discovery https://review.openstack.org/409270 | 20:13 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Move image tests from caching to image test file https://review.openstack.org/407668 | 20:13 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Remove caching config from test_image https://review.openstack.org/407669 | 20:13 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Remove a few glance client mocks we missed https://review.openstack.org/385191 | 20:13 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Change fixtures to use https https://review.openstack.org/409275 | 20:13 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Change image update to REST https://review.openstack.org/385193 | 20:13 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Remove glanceclient and warlock from shade https://review.openstack.org/385194 | 20:13 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Replace swift capabilities call with REST https://review.openstack.org/408865 | 20:13 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Change register_uri to use the per-method calls https://review.openstack.org/407767 | 20:13 |
| mordred | Shrews: merry christmas! | 20:13 |
| Shrews | mordred: +2'd. feel free to get more eyes or self approve | 20:14 |
| mordred | Shrews: thanks! | 20:16 |
| mordred | Shrews: on the fip funcitonal test ... turns out I can't do the search-by-router thing | 20:17 |
| mordred | Shrews: because a floating ip doesn't have a router until it is attached to something | 20:17 |
| Shrews | mordred: how about by net? | 20:17 |
| mordred | so the first floating ip that's created is just a floating ip | 20:17 |
| mordred | nope - not that either | 20:17 |
| mordred | because they all come from the same public network | 20:17 |
| Shrews | ah, right | 20:17 |
| Shrews | well poo | 20:17 |
| mordred | yah. we'll just have to cross that bridge when we come to it | 20:18 |
| mordred | I _think_ we'd wind up needing to create a project for each test | 20:19 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add functional test to show floating ip reuse https://review.openstack.org/409842 | 20:20 |
| mordred | Shrews: there's that one | 20:20 |
| mordred | rcarrillocruz: if you're bored (now that you're back from time off_: https://review.openstack.org/#/q/status:open+project:openstack-infra/shade+branch:master+topic:switch-to-rest :) | 20:21 |
| rcarrillocruz | sure, i'll have a look | 20:22 |
| *** gouthamr has quit IRC | 20:48 | |
| mordred | Shrews: honestly - because of how racy these can be - I'm starting to think we're not going to be able to do this test safely unless we do project-per-test | 20:53 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add test to trap for missing services https://review.openstack.org/409869 | 21:38 |
| morgan | mordred: never ask a question you don't actually want the answer to | 22:16 |
| morgan | mordred: and "how racy"... is NEVER a good question ;) | 22:16 |
| morgan | ^_^ | 22:16 |
| mordred | :) | 22:16 |
| morgan | mordred: woot almost all packed | 22:16 |
| morgan | internet set for the new place | 22:16 |
| mordred | \o/ | 22:16 |
| morgan | moving on saturday! | 22:16 |
| clarkb | in the snowpocalypse! | 22:17 |
| clarkb | (I am secretly excited and hope that the forecast turns out to be accurate and we get like 10" of snow this week | 22:17 |
| morgan | clarkb: damn.... i hadn't seen the forecast | 22:18 |
| morgan | clarkb: dear god... 10" in PDX... | 22:18 |
| morgan | people could barely handle "light dusting with some ice here and there" | 22:18 |
| clarkb | morgan: my current forecast for wednesday is 5-8" then an inch here and there through sunday | 22:18 |
| morgan | well I *may* not be moving then | 22:18 |
| morgan | clarkb: yeah i see 5-8" on wed | 22:19 |
| morgan | wow. | 22:19 |
| clarkb | of course its portland so you can't really know till its over | 22:20 |
| morgan | right | 22:21 |
| morgan | we will see | 22:21 |
| morgan | hopefully I can get out and moved on saturday still ;) | 22:21 |
| mordred | morgan: SO - I've got a maybe fun (maybe not fun) thing for you ... | 22:45 |
| mordred | morgan: https://review.openstack.org/#/c/409842/ this morning I wrote that ... to exercise floating ip reuse and non-reuse cases. unfortunately, fip reuse is almost impossible to get right | 22:46 |
| mordred | morgan: which makes me think that each of those tests should actually have self.operator_cloud create a per-test project inside of which the rest of the test should run - so that we can be a bit more brazen and know that there are _no_ resources being created by or leftover from other tests | 22:47 |
| morgan | i think you're tight. | 22:48 |
| mordred | morgan: I'm not 100% on the whole assign-roles-to-users bit or whatever it is I'd need to do there - but thought it might be a nice mix of something you know and something new to learn | 22:48 |
| morgan | yep | 22:48 |
| morgan | happy to help there. | 22:48 |
| mordred | \o/ | 22:48 |
| morgan | now that mostly packed can focus on such things as getting those things working | 22:49 |
| morgan | and ksa contextmanager thing | 22:49 |
| morgan | and the MFA code in Keystone | 22:49 |
| * morgan is looking at that review actually right now ^^ | 22:50 | |
| morgan | yeah i think it should be pretty easy to do per-test project | 22:51 |
| mordred | in fact - if that works out sanely - one could imagine making that a general concept in the functional tests- all tests get their own project ... execpt of course for the tests that test project/domain stuff themselves | 22:51 |
| mordred | yay being packed! | 22:51 |
| morgan | I'm plannin on just making a .setup_project_for_test | 22:51 |
| morgan | that is part of setUp (well a fixture) | 22:52 |
| morgan | so we can simply apply it *everywhere* and encode the logic for roles and what not | 22:52 |
| morgan | i am also thinking of just using keystone inherited roles | 22:52 |
| morgan | so we apply it to the domain as inherited | 22:52 |
| morgan | so it's a one-shot set of roles | 22:52 |
| morgan | and all projects get them. | 22:52 |
| morgan | (aka easy to maintain as role magic becomes more prolific) | 22:53 |
| morgan | some massaging in ._setup_networks is going to be needed | 22:56 |
| morgan | but I think it's ultimately doable | 22:56 |
| mordred | woot | 23:00 |
| morgan | it really shouldn't be a ton of code to clean that bit up | 23:02 |
| *** rcarrillocruz has quit IRC | 23:05 | |
| mordred | nope. it'll be one of those times where thinking happens a bunch and then there's like 4 lines of code | 23:17 |
| mordred | morgan: the trickiest part, tbh, is likely going to be creating a new OpenStackCloud instance with the project id interjected on top of the normal auth info | 23:17 |
| mordred | morgan: but even _that_ won't be hard | 23:17 |
| mordred | however, even just thinking about it, it makes me think having shade have a method which is "give me a cloud just like this one except have it be scoped to this project instead" | 23:18 |
| mordred | would potentially bea thing that people could find useful | 23:18 |
| morgan | that is what I'm planning to do via a fixture | 23:18 |
| morgan | take cloud X <(what we have today)> and apply scope Y to it | 23:18 |
| mordred | ++ | 23:18 |
| morgan | and we can add the scope mechanism as an explicit fixture that overrides what we have | 23:19 |
| morgan | so we apply it as needed | 23:19 |
| morgan | vs "duplicate" cloud. | 23:19 |
| morgan | no need to do that. | 23:19 |
| morgan | it also could be used for explicit parts of a test that way | 23:19 |
| mordred | I believe you | 23:19 |
| morgan | instead of the "whole test" | 23:20 |
| morgan | since fixtures can be done as context managers | 23:20 |
| mordred | (this may be the reason I thought _you_ might be a good human to do this one :) ) | 23:20 |
| *** rcarrillocruz has joined #openstack-shade | 23:20 | |
| morgan | heh | 23:21 |
| morgan | it might be kindof up my alley | 23:21 |
| openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add test to trap for missing services https://review.openstack.org/409869 | 23:35 |
| *** rcarrillocruz has quit IRC | 23:38 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!