| *** https_GK1wmSU has joined #openstack-outreachy | 00:31 | |
| *** https_GK1wmSU has left #openstack-outreachy | 00:33 | |
| *** lwanderley has quit IRC | 00:49 | |
| *** lwanderley has joined #openstack-outreachy | 00:53 | |
| *** lwanderley has quit IRC | 01:18 | |
| *** rmascena has quit IRC | 01:58 | |
| *** rossella_s has quit IRC | 02:19 | |
| *** rossella_s has joined #openstack-outreachy | 02:21 | |
| *** https_GK1wmSU has joined #openstack-outreachy | 02:56 | |
| *** https_GK1wmSU has left #openstack-outreachy | 02:59 | |
| *** leanderthal has joined #openstack-outreachy | 10:46 | |
| *** leanderthal is now known as leanderthal|afk | 10:48 | |
| *** lwanderley has joined #openstack-outreachy | 11:59 | |
| *** raildo has joined #openstack-outreachy | 12:10 | |
| *** lwanderley has quit IRC | 12:30 | |
| *** rmascena has joined #openstack-outreachy | 15:40 | |
| *** raildo has quit IRC | 15:42 | |
| *** rossella_s has quit IRC | 15:48 | |
| *** rossella_s has joined #openstack-outreachy | 15:50 | |
| *** lwanderley has joined #openstack-outreachy | 16:57 | |
| *** lwanderley has quit IRC | 17:31 | |
| *** lwanderley has joined #openstack-outreachy | 17:31 | |
| lwanderley | ping raildo rodrigods | 17:33 |
|---|---|---|
| rodrigods | hey lwanderley | 17:33 |
| rodrigods | :) | 17:33 |
| lwanderley | hi :) | 17:33 |
| rodrigods | around rmascena? | 17:33 |
| rodrigods | lwanderley, first patch merged! \o/ | 17:33 |
| lwanderley | yay, yes! | 17:33 |
| rodrigods | i took some time last week to ping the QA ppl | 17:34 |
| lwanderley | thanks | 17:34 |
| rodrigods | let's wait a bit for rmascena | 17:34 |
| lwanderley | ok | 17:34 |
| rodrigods | let's get started lwanderley | 17:36 |
| rodrigods | rmascena can catch up later | 17:36 |
| rodrigods | lwanderley, i remember that you were going to try to fix the tempest failures in the LDAP env | 17:37 |
| rodrigods | correct? | 17:37 |
| lwanderley | yep | 17:37 |
| rodrigods | (i just forgot to add the AI in the meeting notes) | 17:37 |
| lwanderley | and i have some news on that | 17:37 |
| lwanderley | i need some help to decide what to do | 17:37 |
| rodrigods | cool | 17:37 |
| rodrigods | go ahead :) | 17:37 |
| lwanderley | The problem is not the LDAP domain specifically, it happens because our scenario has domain-specific drivers enabled. Because these option is enabled, the test_list_users test creates another domainto hold its domain enabled user, this new domain is different from the one that contains all the services and the not enabled users (the default domain) | 17:38 |
| lwanderley | . | 17:38 |
| lwanderley | here #link https://github.com/openstack/tempest/blob/master/tempest/api/identity/admin/v3/test_list_users.py#L40 | 17:38 |
| lwanderley | When the test uses keystone api endpoints to retrieve its users (and groups) they only return objects from the default domain, that causes the tests to fail. | 17:39 |
| rodrigods | "another domainto hold its domain enabled user" couldn't understand that | 17:39 |
| rmascena | sorry, I'm here now | 17:40 |
| lwanderley | i don't understand exactly how it happens but when the domain specific driver are enabled that line on the test creates a new domain | 17:41 |
| lwanderley | hi rmascena | 17:41 |
| lwanderley | then it adds a user to that domain and the api does not retrieve it | 17:42 |
| rodrigods | lwanderley, only when the specific driver is enabled? | 17:43 |
| rodrigods | thought that would always create another domain | 17:43 |
| lwanderley | yes, you are right sorry | 17:44 |
| lwanderley | but still what happens is that the api does not retrieve it | 17:44 |
| lwanderley | there's a note about this on this keystone configuration doc | 17:45 |
| lwanderley | #link https://github.com/openstack/keystone/blob/master/doc/source/configuration.rst | 17:45 |
| lwanderley | Note | 17:46 |
| lwanderley | It is important to notice that by enabling either of the domain-specific configuration methods, the operations of listing all users and listing all groups are not supported, those calls will need either a domain filter to be specified or usage of a domain scoped token. | 17:46 |
| rodrigods | cool, got it | 17:46 |
| rodrigods | yeah, that's what i suspected | 17:46 |
| rodrigods | so basically, what you need to do is to specify the domain in list_users | 17:46 |
| rmascena | exactly, on this line: https://github.com/openstack/tempest/blob/master/tempest/api/identity/admin/v3/test_list_users.py#L92 | 17:47 |
| rmascena | you have to say in which domain you're listing that users | 17:47 |
| lwanderley | ok | 17:47 |
| rmascena | since we now have more than 1 domain, this asserts will fail | 17:47 |
| rmascena | propbably the same thing for groups | 17:47 |
| lwanderley | what happens in this specific test is that it tests if the 2 users created on the test setup are inside the list, but each user belongs to a different domain | 17:48 |
| lwanderley | so using one domain id would still fail the assert | 17:49 |
| rmascena | so, you can do a request for each domain and assert that with this result | 17:50 |
| rodrigods | ok, you need to modify the test than | 17:51 |
| rodrigods | basically, what you need is a feature flag | 17:51 |
| rodrigods | feature flags are settings used to modify the test behavior depending in the environment | 17:51 |
| rodrigods | lwanderley, example: https://review.openstack.org/#/c/428292/ | 17:51 |
| rmascena | lwanderley, makes sense for you? | 17:51 |
| rodrigods | lwanderley, see: this portion of the tests is only executed if the feature flag is enabled: https://review.openstack.org/#/c/427345/11/tempest/api/identity/admin/v3/test_roles.py | 17:52 |
| rodrigods | you basically need to do something similar to that | 17:52 |
| lwanderley | ok, i think i get it | 17:53 |
| rodrigods | cool | 17:53 |
| rodrigods | let us know if you have any doubts | 17:53 |
| lwanderley | just one thing | 17:53 |
| rodrigods | but, basically: add a setting in tempest for the LDAP feature flag | 17:53 |
| rodrigods | then enable this flag only in the specific job that configures the LDAP plugin | 17:54 |
| rodrigods | you can even modify your LDAP plugin to enable this whenever the "ldap" service is enabled | 17:54 |
| rodrigods | got it? | 17:54 |
| lwanderley | yes, i got it | 17:55 |
| lwanderley | i will have another test case that tests basically the same thing but only runs when ldap is enabled while the current test case should be prevented to run | 17:55 |
| lwanderley | when ldap is enabled? | 17:55 |
| rodrigods | you don't need an extra test case | 17:56 |
| lwanderley | yeah, i see it now just a condition | 17:56 |
| rodrigods | just add a condition in the existing ones to match the behavior when it is enabled | 17:56 |
| rodrigods | lwanderley, exactly | 17:57 |
| lwanderley | cool | 17:57 |
| rmascena | great | 17:57 |
| rodrigods | i'll let you figure out the patches dependencies yourself, lwanderley | 17:58 |
| lwanderley | ok, no problem | 17:59 |
| rodrigods | since you might have a chicken/egg problem there :) | 17:59 |
| rodrigods | anything else? | 17:59 |
| lwanderley | not from me | 17:59 |
| rmascena | me neither | 17:59 |
| rodrigods | cool, so see you next week | 17:59 |
| rodrigods | let us know if you have any questions | 18:00 |
| rodrigods | good progress so far, lwanderley | 18:00 |
| rodrigods | o/ | 18:00 |
| rmascena | lwanderley, yeap, if you don't have so much progress on this, please ping us asap | 18:00 |
| lwanderley | ok, thanks! | 18:00 |
| lwanderley | rmascena rodrigods have a nice week | 18:02 |
| rodrigods | thanks, you too! | 18:02 |
| *** lwanderley has quit IRC | 18:05 | |
| *** rmascena has quit IRC | 22:00 | |
| -openstackstatus- NOTICE: osic nodes have been removed from nodepool due to a problem with the mirror host beginning around 22:20 UTC. please recheck any jobs with failures installing packages. | 23:47 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!