Wednesday, 2016-01-06

siva_krishnanCan someone help me in resolving the below error while running s3compatibility test suite http://paste.openstack.org/show/483172/19:10
timburkesiva_krishnan: is your swift server running with SSL enabled? if not, try setting is_secure=no in the testing config file20:04
timburke(i don't know that i've actually tried running the tests with ssl enabled... i feel like it would probably have been difficult enough that i would remember it :P)20:06
siva_krishnan@timburke would try that out20:07
siva_krishnantim_bruke got it fixed. Thanks for the help. In my previous config I disabled it . this one I didn't notice. Still getting the SAXParseException mismatched tag20:13
siva_krishnanhttp://paste.openstack.org/show/483178/20:14
timburkehrm. can we get the content that was transmitted? like with tcpdump or something? or just dive into the boto source and log the body20:16
timburkei'm about to go grab lunch, but i should be around to help again later20:16
siva_krishnanok sure20:30
*** chsc has joined #swift322:51
kota_hello23:01
kota_anyone else here?23:01
chscyes, hello23:01
kota_chsc: o/23:02
siva_krishnano/23:02
kota_siva_krishnan: o/23:03
kota_should we wait someone before starting meeting?23:04
kota_i expects timburke joins23:04
timburkehi all23:08
timburkesorry, was talking to torgomatic about something...23:08
kota_\o/23:08
kota_never mind23:09
kota_ok let's get started23:09
kota_there is vacation weeks on the end of year, so i assume we don't have so much progress for the term.23:10
kota_and i might forget something for priority items so i wanna catch up recent updates at first.23:11
kota_someone?23:11
kota_oh, yeah happy new year swift3 guys :)23:12
chscHappy New Year :)23:12
chscOne thing to mention would be Bug #152257823:12
openstackbug 1522578 in Swift3 "Etag generation in swift3 does not match AWS S3" [Undecided,New] https://launchpad.net/bugs/152257823:12
siva_krishnanHappy New Year!23:12
kota_chsc: ok go ahead.23:12
chscThis was reported to us from an application that tried to validate uploads against AMZ and then swift323:12
chscBasically the calculation of the md5sum is different. Should be easy to fix in swift3, i just have not had the time to look into the code.23:13
chscOn a high level, it should be easy to adjust swift3 to match amz. I would look into this, unless someone objects.23:13
kota_look like multi uploads?23:14
chscyes23:14
timburkeno objections here. you'll need to watch out to make sure it works for both object/key HEAD and container/bucket GET23:15
*** m_kazuhiro has joined #swift323:15
chscok23:16
timburkemaybe add a piece of metadata on upload that has the S3-etag?23:16
timburkestill painful to get it into the container listing, though23:16
kota_exactly23:16
timburkebury it in the content-type, like we do swift_bytes? bleh23:17
timburke(and by "we" i mean "slo")23:17
chscok23:18
chscAnyway, that is all i had to bring up.23:19
kota_or extra HEAD to the manifest if it is slo.23:19
kota_not sure, tho23:19
kota_no s/extra HEAD/extra GET manifest./23:20
kota_ok23:20
kota_anything?23:20
kota_from my side,23:21
kota_we got some gate issues in the end of last year.23:21
kota_one came from os-client-config but it resolved in upstream.23:21
kota_right now, we don't have the issue anymore.23:22
timburkethanks kota_. glad to see our gates working again :)23:22
kota_and one more pylint became smarter and hit a pkg_resource issue (assignment-no-return)23:23
kota_i resolved it to change code to use with statement but it perhaps might be an issue again if pylint would be smarter again.23:23
kota_just notification, tho.23:24
kota_timburke:^^23:24
kota_timburke: thanks for working on them a lot, it was helpful for me :)23:24
timburkeglad to help23:24
kota_they are all gate issues i know23:25
kota_and as another topic23:26
kota_i have a concern to make ranged copy in multipart upload but it was resolved in today's Swift meeting23:27
kota_we can go current way on patch 25506923:28
patchbotkota_: https://review.openstack.org/#/c/255069/ - Support x-amz-copy-source-range for Upload Part - ...23:28
kota_congrats timburke^^23:28
kota_that's all from me about recent updates23:28
timburkethanks for bringing that up; i'm glad we had the discussion. i still need to address the other comment on my patch, though...23:28
timburkenot so recent, but what do we think about patch 248178?23:29
patchbottimburke: https://review.openstack.org/#/c/248178/ - Ignore Content-Type from client when deleting mult...23:29
kota_good catch23:30
kota_can i make sure that status23:30
kota_?23:30
kota_i might forget something right now.23:31
timburkeyeah, no problem. just wanted to keep it on your radar23:31
kota_ok, am going to look at this during this week.23:34
kota_you seems to have added great commen for my question.23:34
kota_could we move on to next agenda?23:36
kota_well moving on to next...23:37
kota_Drop python 2.6 support23:37
kota_as we know, openstack gating environment doesn't have py26, right?23:38
kota_but we still have some py26 docs in tox and packaging and it makes some constraints on coding, imo.23:39
kota_m_kazuhiro is now trying to follow the openstack movement and he wants to drop py26 envrionment from swift323:40
kota_but i have concerns someone is hurt for it (i don't think anyone exist) but, to make sure, i brought this in meeting.23:41
timburkeit seems reasonable. without a gate, it's hard to truly claim support; plus swift has necessarily dropped support anyway23:41
kota_mainly, to rackers and swiftstack?23:41
notmynameswiftstack doesn't care about py26 support23:42
timburkewe'll install py27 if it isn't already available (eg, on rhel)23:42
kota_timburke: yes, and i wanna update some test Swift environment to >=2.4.0 which doen't have py26 anymore.23:43
kota_notmyname: thx!23:43
kota_not test but requirements23:43
timburkei'm more nervous about that. i'd like to preserve compatibility with as old a version of Swift as we reasonably can23:44
kota_right23:45
timburke(i'm already a little nervous that we don't *actually* test against 2.1.0, the claimed minimum)23:45
kota_me too :)23:45
kota_one more question, does anyone want to make a release point before dropping py2.6?23:45
kota_if no one, just do in upstream and will make a release *after* that.23:46
kota_looks no objections...23:47
timburkeas i recall, there wasn't too much going on since 1.9. i suppose we *could*, similar to what was done for swiftclient, but i'm not sure there's a need to23:48
kota_i also guess no much progress since 1.923:49
kota_ok, if you all have some matters in packaging, please let me know23:50
kota_m_kazuhiro: do you have any issues or opinions?23:50
kota_for working on them?23:50
m_kazuhirokota_: nothing, but i want to talk to you after this meeting.23:52
openstackgerritTim Burke proposed openstack/swift3: py26 is no longer supported by Infra's CI  https://review.openstack.org/26170123:52
openstackgerritTim Burke proposed openstack/swift3: remove python 2.6 trove classifier  https://review.openstack.org/26084423:53
timburke(sorry, rebasing those to after the gate getting fixed)23:53
kota_timburke: never mind and good call i noticed that patch already exist in gerrit23:54
kota_timburke: thanks!23:55
kota_that's all for today.23:55
kota_anything?23:55
timburkenot from me23:56
kota_end meeting, thanks for working on swift3 and I'm looking forward to great progress on this year :)23:57
timburkethanks kota_!23:57
chscthank you23:57
*** chsc has quit IRC23:57
m_kazuhirothank you23:57

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!