*** _CrustY has quit IRC | 00:02 | |
*** douglascorrea has quit IRC | 05:14 | |
*** douglascorrea has joined #swift3 | 05:15 | |
*** douglascorrea has quit IRC | 05:15 | |
*** douglascorrea has joined #swift3 | 05:16 | |
*** douglascorrea has quit IRC | 07:02 | |
*** douglascorrea has joined #swift3 | 07:03 | |
*** douglascorrea has quit IRC | 07:07 | |
*** openstackgerrit has quit IRC | 08:48 | |
*** acoles_ is now known as acoles | 11:14 | |
*** acoles is now known as acoles_ | 18:21 | |
*** _CrustY has joined #swift3 | 19:29 | |
*** Taaronja has quit IRC | 19:57 | |
*** acoles_ is now known as acoles | 20:53 | |
*** nikivi has joined #swift3 | 21:18 | |
*** nikivi has quit IRC | 21:22 | |
*** acoles is now known as acoles_ | 22:13 | |
*** CrustY_ has joined #swift3 | 22:24 | |
*** _CrustY has quit IRC | 22:24 | |
*** nikivi has joined #swift3 | 22:30 | |
*** CrustY_ has quit IRC | 22:33 | |
*** _CrustY has joined #swift3 | 22:33 | |
*** nikivi has quit IRC | 22:36 | |
*** nikivi has joined #swift3 | 22:42 | |
*** nikivi has quit IRC | 22:45 | |
kota_ | meeting? | 23:08 |
---|---|---|
kota_ | sorry late 8 minutes | 23:08 |
kota_ | timburke: ? | 23:08 |
timburke | hi kota_! | 23:08 |
timburke | thanks for the ping :-) | 23:08 |
kota_ | :-) | 23:08 |
kota_ | today, I don't have a lot of things here so i expect this is short. | 23:09 |
kota_ | but the first | 23:09 |
kota_ | as well as swift irc meeting, I'd like to consider skip swift3 meeting in two weeks | 23:09 |
timburke | i'd be fine with that | 23:10 |
kota_ | ok, let's schedule the next one to Jan 4th. | 23:10 |
timburke | sounds good | 23:10 |
kota_ | timburke: In your time ;-) 5th in my time. | 23:11 |
kota_ | kk, let's move onto topics to discuss | 23:11 |
kota_ | timburke: do you have something here? | 23:11 |
kota_ | i think, ceph-s3 tests is in now non-voting gate | 23:11 |
timburke | i figure it's worth pointing out that we've got some new non-voting jobs | 23:12 |
timburke | yes, that :-) | 23:12 |
kota_ | and you rebased some patches off to resolve the failures for non-voting | 23:12 |
timburke | https://review.openstack.org/#/c/350223/ demonstrates it fairly well | 23:12 |
timburke | note that our known-failures are already a little out of date; eventlet had a release that included one of my fixes | 23:13 |
kota_ | https://review.openstack.org/#/c/350223/6/swift3/test/functional/conf/ceph-known-failures-keystone.yaml | 23:14 |
kota_ | which one (or more?) | 23:14 |
kota_ | expecting *_readable is for this fix and *_contentlength_empty is for the eventlet release. | 23:15 |
kota_ | I'm really happy to be able to see the tables -> http://logs.openstack.org/23/350223/6/check/gate-swift3-tox-s3tests_tempauth-ubuntu-xenial/7523702/console.html#_2016-12-12_23_56_12_894682 in the gerrit gate :-) | 23:17 |
kota_ | looking the patch... | 23:26 |
timburke | ugh. sorry to disappear; my computer decided to restart | 23:27 |
kota_ | no worries | 23:28 |
kota_ | timburke: one question | 23:28 |
kota_ | i'm looking at https://review.openstack.org/#/c/350223/6/swift3/test/unit/test_middleware.py@166 | 23:29 |
kota_ | and wondering how ceph-s3 test works? | 23:29 |
kota_ | does not it use boto for the testing request? | 23:29 |
timburke | yeah, the tables are nice. and since it built off of other logging improvements, we've got http://logs.openstack.org/23/350223/6/check/gate-swift3-tox-s3tests_tempauth-ubuntu-xenial/7523702/tox/ceph-s3-summary.log.txt | 23:29 |
timburke | ah, yeah, that! it does some monkey patching as i recall... | 23:30 |
timburke | lemme find it... | 23:30 |
timburke | https://github.com/ceph/s3-tests/blob/master/s3tests/functional/test_headers.py#L49-L87 i think? | 23:31 |
timburke | with the extra setup code below it | 23:32 |
kota_ | ok, I don't get completly but it looks sort of making the headers user-free. | 23:34 |
notmyname | the tables are cool! | 23:35 |
notmyname | I saw it on kota_'s twitter account first :-) | 23:35 |
kota_ | notmyname: :-) | 23:35 |
kota_ | notmyname: that's swiftstack's work! | 23:36 |
notmyname | :-) | 23:36 |
timburke | another nice thing is that the raw results are available in xunit format (iirc; xml at any rate) at http://logs.openstack.org/23/350223/6/check/gate-swift3-tox-s3tests_tempauth-ubuntu-xenial/7523702/tox/ceph-s3-details.xml for example | 23:37 |
kota_ | hmm | 23:43 |
kota_ | looking at patch 350223 | 23:43 |
kota_ | i have small idea to cleanup the code around string_to_sign and canonical string | 23:43 |
kota_ | due to the test case difference auth v2 is using string_to_sign for canonical string | 23:44 |
kota_ | but it's not so a big blocker and the priority should be fixing a bug first. | 23:44 |
kota_ | Done | 23:46 |
timburke | thanks! | 23:46 |
kota_ | timburke: anything else? | 23:48 |
timburke | further work could be done to get a v4 signature job going, but i'm kinda hesitant to go there while the keystone ceph test is taking so long | 23:48 |
timburke | https://review.openstack.org/#/c/406424/ may help? | 23:48 |
kota_ | oh yeah | 23:49 |
kota_ | that's priority one | 23:49 |
timburke | on that note, i suppose we ought to look into getting proper v3 responses from keystone at some point... looks like we'd need another controller in https://github.com/openstack/keystone/blob/master/keystone/contrib/s3/core.py, similar to what was done for the ec2 extension | 23:50 |
kota_ | ok | 23:51 |
timburke | but that can definitely be done later | 23:51 |
kota_ | ah, maybe do we need to land for v3 patch first? | 23:51 |
kota_ | https://review.openstack.org/#/c/384659/ | 23:52 |
timburke | yeah, probably a good idea | 23:53 |
kota_ | so.... should the order be v3 patch -> keystone extention -> one keystone request? | 23:53 |
timburke | it uncovered something interesting: keystone can currently serve v2 responses through their v3 pipeline with the s3 extension enabled | 23:54 |
kota_ | interesting | 23:54 |
timburke | idk, i think one keystone request could land in isolation -- once we have the keystone extension, we can update the code to handle the new response format | 23:54 |
kota_ | make sense, we need to clarify the response format, *which* keystone can return *what* format. | 23:55 |
*** openstackgerrit has joined #swift3 | 23:56 | |
openstackgerrit | Merged openstack/swift3: Properly strip headers for request signing https://review.openstack.org/350223 | 23:56 |
kota_ | thank you gerrit bot | 23:56 |
kota_ | kk, will look at | 23:57 |
kota_ | timburke: sorry it looks like i'm running out of time. I'll check the keystone response with looking at your code. | 23:58 |
kota_ | thanks for joining and working great jobs. | 23:58 |
timburke | thanks kota_! | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!