| *** Turingtoast has joined #swift3 | 00:08 | |
| *** Turingtoast has quit IRC | 07:29 | |
| *** Turingtoast has joined #swift3 | 08:12 | |
| *** Turingtoast has quit IRC | 11:02 | |
| *** _CrustY has joined #swift3 | 12:58 | |
| *** CrustY_ has quit IRC | 12:58 | |
| *** _CrustY has quit IRC | 13:12 | |
| *** _CrustY has joined #swift3 | 13:13 | |
| *** Turingtoast has joined #swift3 | 13:32 | |
| *** Turingtoast has quit IRC | 14:21 | |
| *** Turingtoast has joined #swift3 | 15:30 | |
| *** Turingtoast has quit IRC | 15:31 | |
| *** Turingtoast has joined #swift3 | 16:26 | |
| *** Turingtoast has quit IRC | 16:33 | |
| *** iGeni has quit IRC | 17:25 | |
| *** cfarquhar has joined #swift3 | 17:35 | |
| *** notmyname has quit IRC | 17:43 | |
| *** notmyname has joined #swift3 | 17:44 | |
| *** chsc has joined #swift3 | 19:14 | |
| *** chsc has quit IRC | 19:14 | |
| *** chsc has joined #swift3 | 19:14 | |
| cfarquhar | hello. Does anyone here have swift3 working with ocata or pike? I have things working perfectly in Mitaka, but have not successfully moved past the issue described at https://ask.openstack.org/en/question/106557/swift3s3-api-errors-when-authenticating-with-ec2-keys/. | 19:28 |
|---|---|---|
| timburke | cfarquhar: how are you getting swift3? would you be comfortable running an untagged sha from master, or should i be bugging kota_ to do a release? | 19:37 |
| timburke | pike keystone *should* be able to work, but you'd need a recent-enough version of swift3 | 19:38 |
| timburke | (really, i should be bugging kota_ to do a release *anyway*... but there's enough other stuff going on that it hasn't been a high priority for me) | 19:39 |
| cfarquhar | timburke: I'm using the tip of master in both my ocata and pike attempts | 19:42 |
| cfarquhar | I get to the point that S3Token sends EC2 credentials to keystone, and gets a token back (I think) from keystone, but something is going wrong after that. | 19:43 |
| timburke | got proxy logs i could see? | 19:44 |
| timburke | also, sorry for not sorting out the gate on https://review.openstack.org/#/c/473939/ -- i really need to fix that! | 19:44 |
| cfarquhar | no worries! it's been on my to-do list to investigate but I've had more pressing things :\ | 19:46 |
| cfarquhar | http://paste.openstack.org/show/616785/ | 19:46 |
| timburke | it's a stupid version conflict, fairly certain it's resulting from our keystone gate not using upper-constraints and trying to run mitaka keystone | 19:51 |
| timburke | but back to your current problem -- looks like swift could contact keystone just fine, and knew enough to get the project it should use... is s3_acl enabled? | 19:52 |
| cfarquhar | no it's set to False | 19:54 |
| timburke | cool, just wanted to check since it uses a different access model | 19:55 |
| cfarquhar | If it helps, I'm getting user_token = None and user_token_valid = False here: https://github.com/openstack/keystonemiddleware/blob/4.16.0/keystonemiddleware/auth_token/__init__.py#L635, which is giving me that "Deferred reject downstream" | 19:55 |
| cfarquhar | (I'm assuming line #7 in my paste is where things begin to go wrong) | 19:56 |
| timburke | oh! right! i think it's to do with the pipeline ordering change following https://github.com/openstack/swift3/commit/6ffcc29 | 20:00 |
| cfarquhar | nice! testing now. | 20:02 |
| cfarquhar | thank you! | 20:04 |
| cfarquhar | that works! | 20:04 |
| timburke | perfect! i'm glad it works :-) | 20:04 |
| timburke | fwiw, the trouble was that authtoken strips out the headers that commit populates, which is reasonable behavior on their part. the bit in the commit message about old pipelines continuing to work assumed v2 responses, where we'd also populate the auth token -- when handling v3 responses, s3token never even sees a token id, so it *can't* populate it | 20:05 |
| cfarquhar | makes sense. thanks for the details. I briefly considered pipeline issues, but mostly in the context of unrelated modules we add. I'll submit a PR to reorder it in the README. | 20:12 |
| timburke | ...and i'll work on getting the gate untuck so we can actually land it :-) | 20:16 |
| *** openstackgerrit has joined #swift3 | 20:38 | |
| openstackgerrit | Charles Farquhar proposed openstack/swift3 master: Update pipeline ordering in README.md https://review.openstack.org/488197 | 20:38 |
| openstackgerrit | Tim Burke proposed openstack/swift3 master: Add support for upper-constraints.txt https://review.openstack.org/488200 | 20:49 |
| *** ChadTaljaardt has joined #swift3 | 21:00 | |
| *** cfarquhar has quit IRC | 21:47 | |
| *** Turingtoast has joined #swift3 | 21:50 | |
| *** openstack has joined #swift3 | 22:46 | |
| *** chsc_ has quit IRC | 23:27 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!