| *** mikecmpbll has quit IRC | 02:00 | |
| *** mikecmpbll has joined #openstack-swift | 02:05 | |
| *** mikecmpbll has quit IRC | 02:21 | |
| *** vr42 has joined #openstack-swift | 02:29 | |
| *** gyee has quit IRC | 02:35 | |
| *** rchurch_ has joined #openstack-swift | 02:35 | |
| *** rchurch has quit IRC | 02:36 | |
| *** vr42 has quit IRC | 02:38 | |
| openstackgerrit | Merged openstack/swift master: Python3: fix test/unit/obj/test_replicator.py https://review.openstack.org/633645 | 02:54 |
|---|---|---|
| *** psachin has joined #openstack-swift | 03:04 | |
| *** gkadam__ has joined #openstack-swift | 04:17 | |
| *** e0ne has joined #openstack-swift | 07:33 | |
| *** pcaruana has joined #openstack-swift | 07:45 | |
| *** hseipp has joined #openstack-swift | 07:46 | |
| *** ccamacho has joined #openstack-swift | 07:58 | |
| *** tkajinam has quit IRC | 08:06 | |
| *** mikecmpbll has joined #openstack-swift | 08:28 | |
| *** mikecmpbll has quit IRC | 08:58 | |
| *** mikecmpbll has joined #openstack-swift | 09:11 | |
| *** mikecmpb_ has joined #openstack-swift | 09:29 | |
| *** mikecmpbll has quit IRC | 09:30 | |
| *** hseipp has quit IRC | 09:47 | |
| *** hseipp has joined #openstack-swift | 09:59 | |
| *** ccamacho has quit IRC | 11:05 | |
| *** ccamacho has joined #openstack-swift | 11:12 | |
| *** cschwede has quit IRC | 12:15 | |
| *** cschwede has joined #openstack-swift | 12:23 | |
| *** ChanServ sets mode: +v cschwede | 12:23 | |
| *** e0ne has quit IRC | 12:41 | |
| *** gkadam__ has quit IRC | 12:58 | |
| *** vintbra has joined #openstack-swift | 13:14 | |
| *** e0ne has joined #openstack-swift | 13:42 | |
| *** mvkr has quit IRC | 13:52 | |
| *** psachin has quit IRC | 14:09 | |
| *** mvkr has joined #openstack-swift | 14:20 | |
| *** frickler has joined #openstack-swift | 14:54 | |
| tdasilva | good morning | 14:54 |
| rledisez | good morning | 14:56 |
| tdasilva | rledisez: hope you had a good trip home | 15:00 |
| tdasilva | no trains and cars this time ?? | 15:01 |
| rledisez | we were close to miss our flight connections, but we made it :) | 15:01 |
| tdasilva | nice! | 15:03 |
| rledisez | tdasilva: about p 527405, do you think it would be reasonable to directly patch here: https://github.com/openstack/swift/blob/master/swift/common/utils.py#L177 | 15:39 |
| rledisez | something like SWIFT_CONF_FILE = os.path.join(os.env.get('SWIFT_DIR', '/etc/swift'), 'swift.conf') | 15:39 |
| patchbot | https://review.openstack.org/#/c/527405/ - swift - environment variable to set swift_dir - 2 patch sets | 15:40 |
| rledisez | it would majke it more generic (eg: so I can start a reconstructor with a different swift.conf) | 15:40 |
| tdasilva | rledisez: yeah, i think so. was just looking back at that patch and I think timburke's idea is also good | 15:45 |
| rledisez | tdasilva: you're right, it seems a good idea | 15:50 |
| *** gyee has joined #openstack-swift | 16:02 | |
| *** e0ne has quit IRC | 16:05 | |
| *** ccamacho has quit IRC | 16:19 | |
| *** ccamacho has joined #openstack-swift | 16:20 | |
| timburke | env vars are cool, too | 16:23 |
| timburke | now i'm reminded of https://review.openstack.org/#/c/393952/ ... | 16:26 |
| patchbot | patch 393952 - swift - Allow custom swift configuration directory - 5 patch sets | 16:26 |
| *** openstackgerrit has quit IRC | 16:35 | |
| *** ccamacho has quit IRC | 16:41 | |
| *** openstackgerrit has joined #openstack-swift | 16:49 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Fix flakey symlink setup/teardown https://review.openstack.org/634976 | 16:49 |
| *** mvkr has quit IRC | 16:53 | |
| tdasilva | alecuyer, rledisez: this is the doc I was talking about: https://github.com/openstack/swift/blob/master/REVIEW_GUIDELINES.rst | 17:06 |
| rledisez | tdasilva: thx | 17:13 |
| *** hseipp has quit IRC | 17:25 | |
| clayg | timburke: oh nice p 634976 is for the flakey 202 vs 200 thing in the functests on p 629056 | 17:27 |
| patchbot | https://review.openstack.org/#/c/634976/ - swift - Fix flakey symlink setup/teardown - 1 patch set | 17:28 |
| patchbot | https://review.openstack.org/#/c/629056/ - swift - Rebuild frags for unmounted disks - 6 patch sets | 17:28 |
| timburke | clayg: so. many. flakey tests | 17:28 |
| clayg | ok, well - hopefully p 634797 is pretty uncontroversial | 17:29 |
| patchbot | https://review.openstack.org/#/c/634797/ - swift - Cleanup EC and SSYNC frag index parameters - 3 patch sets | 17:29 |
| clayg | timburke: if you wanna point out some of my stupid on the rebuild change maybe we can discuss it tomorrow in the meeting with an eye towards landing this week-ish (if it's ready) | 17:30 |
| clayg | timburke: I'd be happy to look at some test flakey patches - or anything else that's a higher priority? | 17:30 |
| timburke | iirc, yesterday 13/15 of my gerrit comments were "recheck" so... flakey tests sure feel like a high priority to me right now | 17:31 |
| *** mikecmpb_ has quit IRC | 17:31 | |
| clayg | ok, i'll start with the symlink one - do you have a list? | 17:32 |
| timburke | p 634456 should be working its way through the gate now, though -- i think that one was causing me the most pain | 17:32 |
| patchbot | https://review.openstack.org/#/c/634456/ - swift - func tests: Be willing to retry PUTs - 3 patch sets | 17:32 |
| timburke | p 634362, p 634449, p 634976 should all fix other failures i've seen recently | 17:33 |
| patchbot | https://review.openstack.org/#/c/634362/ - swift - Fix flakey func test setup - 1 patch set | 17:33 |
| patchbot | https://review.openstack.org/#/c/634449/ - swift - Fix flakey func test teardown - 1 patch set | 17:33 |
| patchbot | https://review.openstack.org/#/c/634976/ - swift - Fix flakey symlink setup/teardown - 1 patch set | 17:33 |
| timburke | after that... i know i've had people complaining to me about some things recently; see p 621055 (tripped by minio-go and various aws-sdk-* clients) and p 634342 | 17:36 |
| patchbot | https://review.openstack.org/#/c/621055/ - swift - s3api: Look for more indications of aws-chunked up... - 4 patch sets | 17:36 |
| patchbot | https://review.openstack.org/#/c/634342/ - swift - sharding: Make replicator logging less scary - 1 patch set | 17:36 |
| timburke | meanwhile, i think i might take a stab at addressing root causes for https://bugs.launchpad.net/swift/+bug/1813725 https://bugs.launchpad.net/swift/+bug/1758710 and https://bugs.launchpad.net/swift/+bug/1806468 | 17:38 |
| openstack | Launchpad bug 1813725 in OpenStack Object Storage (swift) "Object encryption key truncated sometimes when used with Barbican" [Undecided,New] | 17:38 |
| openstack | Launchpad bug 1758710 in OpenStack Object Storage (swift) "get internal error on download multi-part file" [Undecided,New] - Assigned to Bulat Gaifullin (bulat.gaifullin) | 17:38 |
| openstack | Launchpad bug 1806468 in OpenStack Object Storage (swift) "internal error in symlink proxy" [Undecided,New] | 17:38 |
| timburke | https://memegenerator.net/img/instances/84069856/cant-get-a-500-if-your-proxy-server-refuses-to-start.jpg | 17:41 |
| notmyname | good morning | 17:50 |
| DHE | well technically I'd get a 503 | 18:00 |
| *** mikecmpbll has joined #openstack-swift | 18:00 | |
| zaitcev | Man | 18:31 |
| zaitcev | The worst part of py3 port is external libraries | 18:31 |
| timburke | zaitcev: which now? | 18:32 |
| zaitcev | The port of the bulk middleware is stalled solid because I cannot figure out how to make the tarball module work in py3 like we need it (with binary filenames). | 18:32 |
| zaitcev | But that was a while ago. | 18:32 |
| zaitcev | Now, a test for proxy MIMEputter uses email module, which absolutely cannot work in py3 with binary data. Unlike bulk, I probably can inherit some classes in it and override its behavior. Also, it's merely a test. Worst come to worst, I can just hand-roll a full-on MIME parser without affecting the actual server code. | 18:34 |
| openstackgerrit | Merged openstack/swift master: func tests: Be willing to retry PUTs https://review.openstack.org/634456 | 18:34 |
| zaitcev | I just could not understand how they document that they handle binary attachments with a straight face and then this | 18:35 |
| zaitcev | Apparently they are always base64 | 18:35 |
| zaitcev | Had to examine the source to see that they just always do this: https://github.com/python/cpython/blob/3.7/Lib/email/message.py#L316 | 18:36 |
| zaitcev | Then I searched for a way to convert the "surrogatescape" unicode back into binary | 18:37 |
| timburke | ...which you can... it's just a bit of a pain | 18:37 |
| timburke | but yeah... the stdlib's a bit of a piece of work... you reminded me of https://bugs.python.org/issue33973 and https://bugs.python.org/issue34698 | 18:37 |
| zaitcev | timburke, do you remember where Sam dealt with surrogates? I cannot figure it out and I'm too snobbish to google through stackoverflow... | 18:38 |
| timburke | actually, i think it was me: https://github.com/openstack/swift/blob/2.20.0/swift/common/swob.py#L302-L305 -- so, native_str.encode('utf8', errors='surrogateescape') | 18:40 |
| timburke | but maybe you were thinking of https://github.com/openstack/swift/blob/2.20.0/swift/common/constraints.py#L373-L388 in check_utf8()? | 18:42 |
| *** ccamacho has joined #openstack-swift | 18:44 | |
| zaitcev | wait, that can't be right | 18:46 |
| zaitcev | The original binary is not UTF-8, it's our EC fragments | 18:47 |
| zaitcev | But thanks | 18:47 |
| timburke | well, what encoding is the stdlib using? | 18:47 |
| zaitcev | self._payload = payload.decode('ascii', 'surrogateescape') | 18:47 |
| timburke | whoa! *ascii* with surrogate escape?? how does that even work? | 18:48 |
| zaitcev | well, anything that's 0x80 to 0x100 is escaped, I guess. | 18:48 |
| timburke | looks like | 18:48 |
| zaitcev | But I need to stuff it back in order to compare with the original | 18:49 |
| timburke | seems like .encode('ascii', 'surrogateescape') should still work | 18:49 |
| zaitcev | Okay, it's clearly easier to override email.Message's methods | 18:49 |
| *** itlinux has joined #openstack-swift | 18:49 | |
| zaitcev | timburke: although u'A\udcff'.encode('ascii','surrogateescape') does work! | 18:51 |
| timburke | yeah, looks like it's reversible: http://paste.openstack.org/show/744589/ | 18:52 |
| *** pcaruana has quit IRC | 19:27 | |
| openstackgerrit | Merged openstack/swift master: Move KMIP conf validation out of _get_root_secret https://review.openstack.org/589673 | 19:30 |
| openstackgerrit | Merged openstack/swift master: Fix flakey symlink setup/teardown https://review.openstack.org/634976 | 19:41 |
| *** kei-ichi has quit IRC | 19:48 | |
| *** e0ne has joined #openstack-swift | 20:03 | |
| *** mvkr has joined #openstack-swift | 20:09 | |
| *** vintbra has quit IRC | 20:42 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Include some pipeline validation during proxy-server start-up https://review.openstack.org/635040 | 21:12 |
| timburke | that...probably deserves an UpgradeImpact... | 21:12 |
| *** itlinux has quit IRC | 21:29 | |
| *** e0ne has quit IRC | 21:33 | |
| zaitcev | https://www.meetup.com/GDG-Cloud-Austin/events/258700734/ gRPC you say | 21:42 |
| *** tkajinam has joined #openstack-swift | 22:55 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!
