| timburke | kota_: so i'm thinking about forking tempauth into the swift3 repo to avoid needing to do this backport dance in the future... | 00:14 |
|---|---|---|
| *** f0lder has quit IRC | 00:31 | |
| notmyname | timburke: `tox -etempauth` | 00:36 |
| notmyname | timburke: ERROR: could not install deps (OSError: [Errno 28] No space left on device) | 00:37 |
| *** f0lder has joined #swift3 | 00:40 | |
| *** f0lder has quit IRC | 00:45 | |
| *** f0lder has joined #swift3 | 00:54 | |
| notmyname | some thoughts, no order | 00:58 |
| notmyname | tempauth shouldn't have anything to do with s3/aws auth | 00:58 |
| notmyname | nor should any other auth system | 00:58 |
| notmyname | swift/swift3 should have an auth pattern like keystone: one middleware to talk to the source of truth for users/authz, one middleware to do the mapping for the particular service things to the format that the other canonical talker-to-auth uses | 01:00 |
| notmyname | so swift3 should provide something that identifies an s3 request and puts that into a format that the auth middleware (tempauth, keystone, etc) understands | 01:01 |
| notmyname | maybe that's s3token, I don't know | 01:01 |
| notmyname | note ont eh disk full, I rebooted and cleaned up some stuff on the vm. looks like doing `tox -etempauth` requires about 1Gb of free space | 01:02 |
| notmyname | timburke: kota_: I've now looked at the swift tempauth changes needed for the swift3 patch. I've also run the tests and seen tests pass with the patch and test fail without | 01:11 |
| notmyname | so all that's good | 01:11 |
| notmyname | timburke: kota_: however, I'm not going to push +2/+A until we talk about the questions/thoughts above first. either tonight on IRC or tomorrow in office (with timburke) or in irc with everyone | 01:12 |
| kota_ | notmyname, timburke: I don't like to have just a fork of tempauth into swift3 repo. That causes to divide from the upstream tempauth and it would be hard to track what's difference each other. | 03:19 |
| kota_ | notmyname: however, perhaps simlier with notmyname's idea, it could be... better (? not sure) to clip up the s3 auth part from tempauth and create a new middleware like s3auth into swift3 repo. | 03:20 |
| kota_ | the demerit of that action seems current tempauth user (it should not be in production) will be affected as they should port (or duplicate) the auth info from tempauth section to the new s3auth section. | 03:22 |
| kota_ | on another perspective to the auth get into swift3, it could make us harder to make swift3 back to swift repo anyway :/ | 03:23 |
| *** amoralej|off is now known as amoralej | 09:13 | |
| *** openstackgerrit has joined #swift3 | 10:03 | |
| openstackgerrit | Rico Lin proposed openstack/swift3 master: [Fix gate]Update test requirement https://review.openstack.org/440283 | 10:03 |
| *** chsc has joined #swift3 | 16:39 | |
| *** chsc has quit IRC | 16:39 | |
| *** chsc has joined #swift3 | 16:39 | |
| *** DSLegends has quit IRC | 16:42 | |
| *** amoralej is now known as amoralej|off | 18:44 | |
| *** iGeni has joined #swift3 | 19:00 | |
| *** tdasilva has quit IRC | 19:46 | |
| *** tdasilva has joined #swift3 | 19:56 | |
| *** tdasilva has quit IRC | 23:05 | |
| *** tdasilva has joined #swift3 | 23:23 | |
| *** chsc has quit IRC | 23:36 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!