*** openstack has joined #swift3 | 00:26 | |
*** lyrrad has quit IRC | 00:50 | |
openstackgerrit | Tim Burke proposed openstack/swift3: run functional tests with 's3v4' enabled https://review.openstack.org/272516 | 04:03 |
---|---|---|
timburke | kota_: looking at http://logs.openstack.org/78/285678/1/check/gate-swift3-tox-keystone/7e3256c/console.html#_2016-03-15_11_32_01_021, the 5 failures are all resolved by https://github.com/boto/boto/pull/3176, and the error by https://github.com/boto/boto/pull/3513 | 04:07 |
timburke | looking at the canonical request (http://logs.openstack.org/78/285678/1/check/gate-swift3-tox-keystone/7e3256c/console.html#_2016-03-15_11_32_01_019) -- basically, x-amz-meta-bar2 should sort *after* x-amz-meta-bar, not before, and the host header should be localhost:8080, not localhost:8080:8080 | 04:10 |
kota_ | timburke: awesome! | 04:46 |
kota_ | timburke: Do we have no way to enable v4 auth via boto args? | 04:48 |
kota_ | looking at patch 272516 | 04:49 |
patchbot | kota_: https://review.openstack.org/#/c/272516/ - swift3 - run functional tests with 's3v4' enabled | 04:49 |
kota_ | or might be useful via os.environ at TestCase init? | 04:52 |
kota_ | IMO, if swift3 can handle both (v2 and v4) at the same time, i think we can merge the tox env and run all functests via both or them. | 04:53 |
kota_ | though, we should update the keystone constraint before we do that. | 04:54 |
timburke | kota_: ...maybe? the only way i've done it locally is via ~/.boto config file. agreed though, for the purposes of testing in the gate, it'd be better to cover both auth options. although we may need to create a @skip_v4 decorator to skip the current failures (since those two PRs aren't merged yet) | 05:07 |
kota_ | timburke: \o/ | 05:08 |
timburke | (well, the config file and the env variable, via the keystone_s3v4 tox env) | 05:08 |
timburke | i also submitted https://bugs.launchpad.net/swift3/+bug/1557260 as part of all of this | 05:09 |
openstack | Launchpad bug 1557260 in Swift3 "Swift3 functests should transition to using boto3" [Undecided,New] | 05:09 |
timburke | just figuring that boto3 is more likely to receive fixes | 05:10 |
kota_ | timburke: i know, it's great. | 05:10 |
kota_ | thanks! | 05:13 |
timburke | ok, i'm up to 12 drafts comments (one of which still warrants follow-up; aside from that, lots of questions, no major issues) and a 70-ish line diff for my follow-up patch. i'm going to bed, will try to actually publish tomorrow | 05:43 |
kota_ | hmm... I got 67 errors when I ran 'tox -e keystone_s3v4' :/ | 07:51 |
kota_ | looks like all of errors are 403 forbidden so I cannot authorize via the script... I don't get anything yet. | 07:52 |
kota_ | maybe some constraints for running well seems still not enough. | 07:53 |
kota_ | calling, IRC meeting will be held in March 16th (today/tommorow in your time). | 09:48 |
*** acoles_ is now known as acoles | 09:48 | |
kota_ | it's still 21:00 UTC, please be careful for the time changed (PST->PDT) | 09:49 |
kota_ | If you have something to bring up, please add your agend to https://wiki.openstack.org/wiki/Meetings/swift3 | 09:50 |
*** openstackgerrit has quit IRC | 09:53 | |
*** openstackgerrit_ is now known as openstackgerrit | 09:53 | |
*** openstackgerrit has quit IRC | 09:53 | |
*** openstackgerrit_ has joined #swift3 | 09:54 | |
*** openstackgerrit_ is now known as openstackgerrit | 09:54 | |
*** openstackgerrit_ has joined #swift3 | 09:55 | |
*** openstackgerrit has quit IRC | 09:55 | |
*** openstackgerrit has joined #swift3 | 09:56 | |
kota_ | hmm... my local setup looks failed at keystone setup... | 11:01 |
*** openstackgerrit has quit IRC | 11:18 | |
*** openstackgerrit has joined #swift3 | 11:18 | |
kota_ | the diff from gerritt looks just versions of webob, wheel, oslo.messaging... | 11:29 |
*** georgem1 has joined #swift3 | 13:01 | |
kota_ | looks like pki_setup failing but I'm wondering why it does. old keystone and new keystone at gerrit work. only new keystone at my local failed... | 14:09 |
*** georgem1 has left #swift3 | 14:11 | |
kota_ | ugh...openstack command looks failed to get the tenant list /and create with 401 Unauthorized. | 14:31 |
kota_ | seems like more environ variable for that? | 14:31 |
kota_ | needed | 14:31 |
timburke | kota_: were you still using patchset 6 from patch 272516? try patchset 7; the diff is just https://review.openstack.org/#/c/272516/6..7/swift3/test/functional/conf/keystone.conf.in | 16:11 |
patchbot | timburke: https://review.openstack.org/#/c/272516/ - swift3 - run functional tests with 's3v4' enabled | 16:11 |
patchbot | timburke: patch 272516 - swift3 - run functional tests with 's3v4' enabled | 16:11 |
timburke | a few commits relevant to that 401: https://github.com/openstack/keystone/commit/37e9d6b (need to explicitly set admin_token), https://github.com/openstack/keystone/commit/5286b4a (admin_token_auth deprecated) | 16:20 |
timburke | https://github.com/openstack/keystone/commit/a14d761 (admin_token_auth undeprecated), https://github.com/openstack/keystone/commit/d446e15 (added `keystone-manage bootstrap` command) | 16:21 |
*** acoles is now known as acoles_ | 16:31 | |
*** lyrrad has joined #swift3 | 16:34 | |
*** acoles_ is now known as acoles | 16:47 | |
*** chsc has joined #swift3 | 17:46 | |
*** acoles is now known as acoles_ | 19:52 | |
*** acoles_ is now known as acoles | 20:30 | |
*** acoles is now known as acoles_ | 20:31 | |
*** mattoliverau has quit IRC | 20:59 | |
*** matt6434 has joined #swift3 | 20:59 | |
*** matt6434 is now known as mattoliverau | 21:00 | |
kota_ | timburke: that looks what I am looking for! | 21:02 |
kota_ | timburke: setup looks runnning well now, still waiting the result. | 21:02 |
*** acoles_ is now known as acoles | 21:02 | |
timburke | kota_: you may still get those 5 failures and 1 error. you'll want to either pip install boto from git, or just manually edit it in the tox env | 21:07 |
*** acoles is now known as acoles_ | 22:02 | |
kota_ | timburke: alright, 5 failures and 1 error I finally reached at you | 22:34 |
timburke | which means there's 60+ passes :-) | 22:46 |
timburke | also, the logging in the failures verifies that we're using sigv4 | 22:46 |
kota_ | oh | 23:01 |
kota_ | logging broken? | 23:01 |
kota_ | hello meeting | 23:01 |
kota_ | looks like timburke was here. | 23:05 |
timburke | hi kota_! | 23:05 |
kota_ | timburke: thank you so much to test/review the sigv4 patch! | 23:05 |
kota_ | timburke: I just in time sent an E-mail for Andrey including you in c.c. | 23:06 |
timburke | sure! apparently we've finally gotten enough demand for it from customers that i could make it a real priority | 23:06 |
timburke | thanks for that; i don't typically follow the mailing list much | 23:06 |
timburke | (i suppose i should fix that) | 23:07 |
kota_ | alright, sounds good. | 23:07 |
kota_ | yeah, but I also had fault that I've been pending a lot of patches (including yours!) :/ | 23:07 |
kota_ | so sorry... | 23:08 |
timburke | no worries on mine; i still need to get the tests passing. just haven't found time :( | 23:09 |
kota_ | yup, we should review a lot accross over projects (i mean, swift/swift3/swift-client, etc...) | 23:10 |
kota_ | recently, pyeclib also, storlets also on my side... | 23:10 |
timburke | wheee! | 23:10 |
kota_ | hehe, and we need to land the encryption at crypt branch by summit :) | 23:11 |
timburke | yeah, i've been trying to load that in my head, too. then i get distracted again | 23:11 |
kota_ | :) | 23:12 |
timburke | it'll be nice for swift3, though; we can get server-side-encryption moving! | 23:12 |
kota_ | yup | 23:12 |
kota_ | timburke: so anything updates else from you? | 23:14 |
timburke | not really. mainly doing reviews | 23:14 |
kota_ | ok | 23:14 |
kota_ | from myside, a few things. | 23:14 |
kota_ | I think sigv4 staff is already we know and it's in reviewing/talking via E-mail so nothing special here if noyone else is here. | 23:16 |
kota_ | s/noyone/no one/ | 23:16 |
kota_ | and one more | 23:16 |
kota_ | for dependency updates | 23:16 |
kota_ | while reviewing the sigv4 patch(and trying to solve environ issue), I've been realizing we need update some dependency/scripts/API's soon. | 23:17 |
timburke | i think i'm in favor of waiting for a mitaka release from Keystone before updating there. not sure how aggressive we want to be about upgrading swift, though | 23:18 |
kota_ | e.g. current gerrit functional gate is against to swift-kilo and older Keystone than kilo | 23:18 |
timburke | i thought it was a kilo keystone as well, just early kilo | 23:19 |
kota_ | yup, it fits my opinion and it make me to get headack sigv4 needs newer keystone. | 23:19 |
kota_ | and looks like pki_setup (included setup_keystone script) and v2 api deprecated | 23:20 |
kota_ | not look at in detail, just saw a few notes, conversations online. | 23:20 |
kota_ | v2 means keystone v2 api | 23:21 |
timburke | ah, thanks :) | 23:22 |
kota_ | timburke: and also boto you suggested :) | 23:22 |
timburke | speaking of setup_keystone, i think we'll probably want to add a `set -e` toward the top; it should make it more obvious when our setup has gone bad and keep us from wasting time running a bunch of tests that will fail | 23:22 |
kota_ | looks good | 23:24 |
kota_ | alright, for now, I don't have strong opinion for this, but still worried about the env migration so it seems good time to consider to bump them past mitaka release. | 23:27 |
kota_ | timburke: thanks for your feedback :) | 23:27 |
timburke | yeah, that sounds good to me | 23:27 |
kota_ | that's all for me. | 23:28 |
timburke | me too. which works well; i've got to pick up my daughter :-) | 23:28 |
kota_ | timburke: thanks for joining, my daughter is playing by me at sofa | 23:30 |
kota_ | :-) | 23:30 |
kota_ | timburke: have a good time with her :) | 23:30 |
timburke | you too! | 23:31 |
openstackgerrit | Tim Burke proposed openstack/swift3: Signature v4 follow-up https://review.openstack.org/293760 | 23:32 |
*** chsc has quit IRC | 23:32 | |
timburke | ok, review's posted :) now i'm actually leaving | 23:35 |
notmyname | give account = AUTH_foo and container = my_c and my swift hostname as foo.bar, I'd use http://foo.bar/v1/AUTH_foo/my_c to do swift API requests | 23:58 |
notmyname | what would I use for swift3 requests? | 23:58 |
notmyname | AUTH_foo.foo.bar/my_c? | 23:59 |
notmyname | my_c.AUTH_foo.foo.bar? | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!