Thursday, 2019-08-08

« Wednesday, 2019-08-07 Index Friday, 2019-08-09 »
*** tdasilva has quit IRC00:03
*** tdasilva has joined #openstack-swift00:03
*** ChanServ sets mode: +v tdasilva00:03
*** zaitcev_ has quit IRC00:41
*** threestrands has joined #openstack-swift00:46
*** gyee has quit IRC00:49
*** zaitcev_ has joined #openstack-swift00:54
*** ChanServ sets mode: +v zaitcev_00:54
*** tdasilva has quit IRC01:04
*** tdasilva has joined #openstack-swift01:05
*** ChanServ sets mode: +v tdasilva01:05
*** diablo_rojo has joined #openstack-swift01:30
*** diablo_rojo has quit IRC01:48
*** tdasilva has quit IRC02:06
*** tdasilva has joined #openstack-swift02:06
*** ChanServ sets mode: +v tdasilva02:06
*** henriqueof has joined #openstack-swift02:17
*** henriqueof has quit IRC02:21
*** zaitcev_ has quit IRC02:24
*** zaitcev_ has joined #openstack-swift02:36
*** ChanServ sets mode: +v zaitcev_02:36
*** early has quit IRC02:37
*** early has joined #openstack-swift02:39
*** early has quit IRC02:56
*** early has joined #openstack-swift02:57
*** tdasilva has quit IRC03:07
*** tdasilva has joined #openstack-swift03:07
*** ChanServ sets mode: +v tdasilva03:07
*** tdasilva has quit IRC03:15
*** tdasilva has joined #openstack-swift03:16
*** ChanServ sets mode: +v tdasilva03:16
*** BjoernT has joined #openstack-swift03:17
*** tdasilva has quit IRC03:19
*** tdasilva has joined #openstack-swift03:20
*** ChanServ sets mode: +v tdasilva03:20
*** BjoernT has quit IRC03:21
*** BjoernT has joined #openstack-swift03:21
*** zaitcev_ is now known as zaitcev03:22
zaitcevYou guys won't believe the patch I'm preparing03:22
*** gkadam has joined #openstack-swift03:31
*** zaitcev has quit IRC03:31
*** psachin has joined #openstack-swift03:32
*** gkadam has quit IRC03:40
*** zaitcev has joined #openstack-swift03:44
*** ChanServ sets mode: +v zaitcev03:44
*** zaitcev has quit IRC03:52
openstackgerritPete Zaitcev proposed openstack/swift master: py3: Fix testUtf8Container (test.functional.tests.TestContainer)  https://review.opendev.org/67524404:03
*** zaitcev has joined #openstack-swift04:06
*** ChanServ sets mode: +v zaitcev04:06
*** tdasilva has quit IRC04:08
*** tdasilva has joined #openstack-swift04:09
*** ChanServ sets mode: +v tdasilva04:09
*** zaitcev has quit IRC04:11
*** BjoernT has quit IRC04:17
*** BjoernT has joined #openstack-swift04:26
*** henriqueof has joined #openstack-swift04:56
*** spsurya has joined #openstack-swift05:02
*** BjoernT has quit IRC05:15
*** ccamacho has quit IRC06:03
*** tdasilva has quit IRC06:11
*** rcernin has quit IRC06:18
*** tdasilva has joined #openstack-swift06:31
*** ChanServ sets mode: +v tdasilva06:31
*** ianychoi has quit IRC07:01
*** ianychoi has joined #openstack-swift07:01
*** ccamacho has joined #openstack-swift07:03
*** baojg has joined #openstack-swift07:08
*** ianychoi has quit IRC07:24
*** ianychoi has joined #openstack-swift07:25
*** tesseract has joined #openstack-swift07:31
*** tdasilva has quit IRC07:32
*** tdasilva has joined #openstack-swift07:32
*** ChanServ sets mode: +v tdasilva07:32
*** tdasilva has quit IRC08:00
*** tdasilva has joined #openstack-swift08:01
*** ChanServ sets mode: +v tdasilva08:01
*** takamatsu has quit IRC08:06
*** threestrands has quit IRC08:19
*** mikecmpbll has joined #openstack-swift08:25
*** tdasilva has quit IRC08:33
*** tdasilva has joined #openstack-swift08:33
*** ChanServ sets mode: +v tdasilva08:33
*** e0ne has joined #openstack-swift09:04
*** henriqueof has quit IRC09:28
*** tdasilva has quit IRC09:29
*** tdasilva has joined #openstack-swift10:52
*** ChanServ sets mode: +v tdasilva10:52
*** baojg has quit IRC11:39
*** baojg has joined #openstack-swift11:41
*** baojg has quit IRC11:41
evrardjphey, sorry to bother you with paperwork... Can I use part of the "What is Swift" text here: https://wiki.openstack.org/wiki/Swift to describe Swift mission in governance?11:48
evrardjpanyone seeing something against that?11:48
tdasilvarledisez: kota_: curious why we have `grpcio>=1.10.0` if we know it will fail starting with 1.3.5? https://github.com/openstack/swift/blob/feature/losf/requirements.txt#L17,L2011:51
*** ccamacho has quit IRC12:24
*** ccamacho has joined #openstack-swift13:23
*** BjoernT has joined #openstack-swift13:28
timburkeevrardjp, that sounds great. sorry i hadn't gotten around to submitting that (or even acking that i'd seen the request)14:46
*** diablo_rojo has joined #openstack-swift14:58
evrardjptimburke: I have proposed something15:02
evrardjphttps://review.opendev.org/#/c/675307/15:03
patchbotpatch 675307 - governance - Add Swift mission into governance - 1 patch set15:03
evrardjpsorry I didn't re-ping here, I thought I did.15:03
*** psachin has quit IRC15:15
*** cschwede has joined #openstack-swift15:15
*** ChanServ sets mode: +v cschwede15:15
*** spsurya has quit IRC15:34
*** gyee has joined #openstack-swift15:43
openstackgerritTim Burke proposed openstack/swift master: Bring our IPv6 setup more in line with devstack's  https://review.opendev.org/67539516:02
*** zaitcev has joined #openstack-swift16:16
*** ChanServ sets mode: +v zaitcev16:16
*** e0ne has quit IRC16:16
*** zaitcev_ has joined #openstack-swift16:17
*** ChanServ sets mode: +v zaitcev_16:17
*** zaitcev has quit IRC16:20
*** mikecmpbll has quit IRC16:34
*** diablo_rojo has quit IRC16:37
timburkehuh. this stable check's bizarro -- how'd we get a retry? https://logs.opendev.org/periodic-stable/opendev.org/openstack/swift/stable/ocata/openstack-tox-py27/b7f4cf7/job-output.txt.gz#_2019-08-08_06_44_31_12804116:46
timburkelike, a 503 would make some sense -- two timeouts popped -- but how'd we get to a 202??16:48
*** zaitcev_ has quit IRC16:57
*** tdasilva has quit IRC16:59
*** BjoernT has quit IRC17:02
*** BjoernT has joined #openstack-swift17:03
*** BjoernT has quit IRC17:03
*** e0ne has joined #openstack-swift17:05
*** BjoernT has joined #openstack-swift17:06
openstackgerritTim Burke proposed openstack/swift master: py3: Finish porting s3 func tests  https://review.opendev.org/67522717:06
*** BjoernT has quit IRC17:06
*** BjoernT has joined #openstack-swift17:07
*** BjoernT has quit IRC17:07
timburkeclayg, idk -- "hardlink" is a decent shorthand for the semantic we're going for, but i don't want to conflate the feature *too* much with the filesystem notion of a hardlink -- they're so different!17:08
*** BjoernT has joined #openstack-swift17:08
timburkenaming things is hard17:08
*** BjoernT has quit IRC17:09
claygtimburke: 100% agree, but also I think it's really wordy/confusing to keep calling them "Symlinks created with a X-Symlink-Target-Etag header"17:09
*** BjoernT_ has joined #openstack-swift17:10
*** zaitcev_ has joined #openstack-swift17:10
*** ChanServ sets mode: +v zaitcev_17:10
*** BjoernT_ has quit IRC17:10
claygIs the suggestion then that we *should* avoid the name hardlink and need to find another noun?  do you have a suggestion?  Or were you just describing/highlighting the reasons we've avoided it in the past and might want to continue to do so?17:11
*** BjoernT has joined #openstack-swift17:11
*** BjoernT has quit IRC17:11
*** BjoernT_ has joined #openstack-swift17:12
*** BjoernT_ has quit IRC17:12
DHEvalidating symlink?17:13
*** BjoernT has joined #openstack-swift17:13
*** BjoernT has quit IRC17:14
timburkeclayg, mostly highlighting why we might want to avoid it. i don't actually have a better suggestion, though17:14
timburke"symlinks with integrity" -- unlike those other cads :P17:14
*** BjoernT_ has joined #openstack-swift17:14
*** BjoernT_ has quit IRC17:14
*** BjoernT has joined #openstack-swift17:15
*** BjoernT has quit IRC17:16
*** BjoernT has joined #openstack-swift17:16
*** BjoernT has quit IRC17:17
*** BjoernT_ has joined #openstack-swift17:17
*** BjoernT_ has quit IRC17:17
*** BjoernT has joined #openstack-swift17:18
*** BjoernT has quit IRC17:18
*** BjoernT has joined #openstack-swift17:19
*** BjoernT has quit IRC17:20
*** BjoernT_ has joined #openstack-swift17:20
*** BjoernT_ has quit IRC17:20
*** BjoernT has joined #openstack-swift17:21
*** BjoernT has quit IRC17:21
*** BjoernT has joined #openstack-swift17:22
*** BjoernT has quit IRC17:22
*** BjoernT has joined #openstack-swift17:22
*** BjoernT has quit IRC17:23
*** BjoernT_ has joined #openstack-swift17:23
*** tesseract has quit IRC17:27
*** BjoernT_ has quit IRC17:28
*** e0ne has quit IRC17:35
openstackgerritTim Burke proposed openstack/python-swiftclient stable/rocky: Fix SLO re-upload  https://review.opendev.org/67332217:37
*** ianychoi has quit IRC17:43
*** ianychoi has joined #openstack-swift17:47
claygliteral link?17:49
timburkestatic link? akin to static large object?17:50
timburke(there's the overloading with compiler-speak, but at least it's not a filesystem thing)17:50
*** ianychoi_ has joined #openstack-swift17:54
*** ianychoi_ has quit IRC17:55
*** ianychoi_ has joined #openstack-swift17:55
*** BjoernT has joined #openstack-swift17:57
*** ianychoi has quit IRC17:58
*** psachin has joined #openstack-swift18:08
*** psachin has quit IRC18:11
openstackgerritTim Burke proposed openstack/swift stable/stein: container-replicator: Add a timeout for get_shard_ranges  https://review.opendev.org/67543218:21
openstackgerritTim Burke proposed openstack/swift stable/stein: reconciler: Enqueue right work for shard containers  https://review.opendev.org/67543318:22
openstackgerritTim Burke proposed openstack/swift stable/stein: Remove invalid dict entries from hashes.pkl  https://review.opendev.org/67543418:22
openstackgerritTim Burke proposed openstack/swift stable/stein: Fix quarantine when object path is not a directory  https://review.opendev.org/67543518:23
openstackgerritTim Burke proposed openstack/swift stable/stein: Make GreenAsyncPile not hang  https://review.opendev.org/67543618:23
openstackgerritTim Burke proposed openstack/swift stable/stein: s3api: Allow clients to upload with UNSIGNED-PAYLOAD  https://review.opendev.org/67543718:26
openstackgerritTim Burke proposed openstack/swift stable/stein: s3api: Block ETag header from CompleteMultipartUpload requests  https://review.opendev.org/67543818:27
*** zaitcev_ has quit IRC18:28
claygi like how static link sounds - i'll probably take to talking about symlinks and statlinks18:36
*** zaitcev_ has joined #openstack-swift18:42
*** ChanServ sets mode: +v zaitcev_18:42
*** BjoernT has quit IRC19:15
*** e0ne has joined #openstack-swift19:16
*** BjoernT has joined #openstack-swift19:16
openstackgerritClay Gerrard proposed openstack/swift master: Consolidate Container-Update-Override headers  https://review.opendev.org/67545119:16
claygmattoliverau: p 675451 - i'd appreciate your thoughts!19:17
patchbothttps://review.opendev.org/#/c/675451/ - swift - Consolidate Container-Update-Override headers - 1 patch set19:17
*** BjoernT has quit IRC19:20
claygThe obj server code said the sysmeta takes precedence, but if it’s not tested who knows. 🤷‍♂️20:11
timburkeseems about right. here's the case i worry about, though: https://github.com/openstack/swift/blob/2.22.0/test/unit/obj/test_server.py#L1791-L179520:21
*** diablo_rojo has joined #openstack-swift20:24
timburketesting locally, my fears seem unfounded, though? at any rate, i see the symlink func tests passing on a saio with a defaul EC policy...20:27
*** diablo_rojo__ has joined #openstack-swift20:30
*** diablo_rojo has quit IRC20:33
*** e0ne has quit IRC20:37
*** zaitcev_ has quit IRC20:41
*** diablo_rojo__ is now known as diablo_rojo20:52
*** zaitcev_ has joined #openstack-swift20:54
*** ChanServ sets mode: +v zaitcev_20:54
*** diablo_rojo has quit IRC21:07
*** diablo_rojo has joined #openstack-swift21:08
*** diablo_rojo__ has joined #openstack-swift21:16
*** diablo_rojo has quit IRC21:19
openstackgerritTim Burke proposed openstack/swift master: slo: Add X-Manifest-Etag to responses  https://review.opendev.org/67546921:20
*** diablo_rojo__ is now known as diablo_rojo21:25
*** kei-ichi has quit IRC22:03
*** rcernin has joined #openstack-swift22:53
*** zaitcev_ is now known as zaitcev23:05
zaitcev    AssertionError: Errors in response headers:23:05
zaitcev      Expected keep-alive but got Keep-Alive for connection23:05
zaitcevin test.functional.tests.TestFileUTF8.testGetResponseHeaders23:05
zaitcevrings a bell for anyone?23:05
*** diablo_rojo has quit IRC23:05
*** diablo_rojo has joined #openstack-swift23:06
openstackgerritMerged openstack/swift master: Bring our IPv6 setup more in line with devstack's  https://review.opendev.org/67539523:10
timburkezaitcev, py3 actually remembers what the casing was on the wire, where py2 had a .lower() buried in the parsing somewhere (httplib, probably?)23:18
zaitcevtimburke: this came up with my functests running under py2, so I was just wondering if you already had a solution for it, like with encoding_dance23:19
timburkethere's a bunch of places in the func tests where i do something like `headers = {h.lower(): v for h, v in resp.getheaders()}` where the code used to just wrap it in a dict(...)23:19
timburkei know i did a lot of that in https://github.com/openstack/swift/commit/506279235d12f1590bd928de51d991b7fb425e6e23:20
zaitcevOkay. Maybe just one or two places slipped through... I have Apache as the front proxy and it may be camel-casing.23:21
timburkeoh! it's in the value!23:23
zaitcevFinding that encoding_dance was difficult and I spent a few days on it, thanks to my sheer incompetence. First, I looked for errors=, but encoding_dance used positional arguments, so I missed it.23:24
zaitcevIn the value? Is it one of the headers that list other headers, like CORS23:24
timburketake a look at the test: https://github.com/openstack/swift/blob/2.22.0/test/functional/tests.py#L131723:25
zaitcevOh. "connection" is actually name, so "Connection: Keep-Alive"23:26
timburkeyup23:26
timburkethe curious thing is that we definitely have some case-sensitive checks; look at x-object-meta-fruit a few lines up23:27
zaitcevBut Banana is cased23:28
zaitcevSomething is tinkering with the _values_, I can't believe it.23:29
timburkeright?? weird...23:29
zaitcevI'll look into this later.23:29
zaitcevDon't worry about this, please. I just asked, in case.23:29
timburkeassuming that eventlet really is the one sending it (and my understanding is that it *should* be), it seems to be lower: https://github.com/eventlet/eventlet/blob/master/eventlet/wsgi.py#L52823:30
zaitcevgo go gadget tcpdump23:35
zaitcevOkay, so this is Apache's fault. It's Connection: keep-alive on the loopback but Connection: Keep-Alive on the outside.23:49
zaitcevI don't know if this is valid and something we need to work around in functests.23:50
zaitcevgtg23:50
timburkehttps://github.com/apache/httpd/blob/a709b2df87f3feaf631569c51d927901f8d2f32e/modules/http/http_protocol.c#L295 i guess?23:50
timburkehttps://tools.ietf.org/html/rfc7230#section-6.1 says "Connection options are case-insensitive." so it sure sounds (to me) like the func tests ought to tolerate either23:53
zaitcevI see.23:54
*** diablo_rojo has quit IRC23:57
« Wednesday, 2019-08-07 Index Friday, 2019-08-09 »

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!