Monday, 2019-10-14

timburkei'm guessing soemthing like http://paste.openstack.org/show/783506/ -- apparently we need to open in binary mode if we want relative seeks...00:27
timburkemahatic_, still around? long time no see!00:34
*** BjoernT has joined #openstack-swift02:11
*** psachin has joined #openstack-swift03:35
mahatic_timburke, hi! congrats on PTL job! how's it going?03:48
timburkemahatic_, i find it similar to when i became core -- things are different, but in many ways much the same ;-)04:05
timburkehow are you doing?04:06
mahatic_timburke, :) I'm sure you're doing a great job04:09
mahatic_I'm alright. For now, moved on from working on object storage to block storage04:10
timburkecool -- how's that going? working on cinder, or something else?04:10
mahatic_rados block storage04:10
mahatic_it's been good. c++ multi-threading & memory management is making my head spin ;)04:12
timburkeyeah, it usually takes me a bit to reorient myself when i come back to looking at liberasurecode or pyeclib :-)04:14
timburkey'know, i've still got one of the ceph papers sitting on my desk... was kind of hoping to find some insights applicable to swift's ring, but i've got this growing suspicion that such lessons are hard to learn except through experience and patches04:15
mahatic_+1 but reading doesn't hurt. it might help you get to one of those `aha` idea moments04:17
mahatic_I was just reading the ring related doc for fun last month, and I felt that `partitions` are equivalent to perhaps what ceph calls `placement groups`04:18
timburkeyeah, it's been a bit since i tried to get ceph loaded in my head, but that sounds about right04:21
*** BjoernT has quit IRC04:24
mahatic_timburke, I got a meeting in a bit, ttyl. It was nice catching up. good night!04:27
timburkemahatic_, yeah, good to catch up a bit! hope you drop in again some time :-)04:27
mahatic_sure :)04:28
*** viks___ has joined #openstack-swift04:38
*** idlemind has quit IRC04:48
openstackgerritTim Burke proposed openstack/swift master: py3: Fix swift-drive-audit  https://review.opendev.org/68832604:57
timburkebaffle, ^^^04:57
openstackgerritTim Burke proposed openstack/swift master: Fix kms_keymaster under Python 3  https://review.opendev.org/68811304:59
timburkealso, zigo, if you get a chance to test out my changes to ^^^, i'd appreciate it05:00
*** ormandj has quit IRC05:42
mattoliverauOh I forgot to mention, the family and I are taking a vacation this week. Till next week then :)06:00
*** ccamacho has joined #openstack-swift06:10
*** ormandj has joined #openstack-swift06:24
*** pcaruana has joined #openstack-swift06:43
*** aluria has joined #openstack-swift06:57
*** tesseract has joined #openstack-swift07:04
*** rcernin has quit IRC07:11
*** rdejoux has joined #openstack-swift07:13
baffletimburke: Seems to work now, I had to set python3.6 as interepter at the top, but that is distribution specific. Ref. swift-drive-audit.07:19
*** rpittau|afk is now known as rpittau07:40
*** mikecmpbll has joined #openstack-swift07:55
*** pcaruana has quit IRC11:34
*** pcaruana has joined #openstack-swift11:41
*** FlorianFa has joined #openstack-swift12:05
*** Guest31368 has joined #openstack-swift12:58
*** Guest31368 is now known as redrobot_13:00
*** redrobot_ is now known as redrobot13:01
*** Guest69423 has quit IRC13:06
*** openstackstatus has quit IRC13:15
*** BjoernT has joined #openstack-swift13:26
*** baojg has quit IRC13:46
*** baojg has joined #openstack-swift13:47
*** BjoernT_ has joined #openstack-swift13:47
*** BjoernT has quit IRC13:49
*** baojg has quit IRC14:07
*** baojg has joined #openstack-swift14:08
*** e0ne has joined #openstack-swift14:16
claygnova is dropping py2!!! :jealous:14:23
*** e0ne has quit IRC14:29
*** e0ne has joined #openstack-swift15:28
*** rdejoux has quit IRC15:28
*** e0ne has quit IRC15:34
*** ccamacho has quit IRC15:47
*** e0ne has joined #openstack-swift16:03
*** openstackstatus has joined #openstack-swift16:04
*** ChanServ sets mode: +v openstackstatus16:04
*** rpittau is now known as rpittau|afk16:06
timburkegood morning16:31
*** BjoernT_ is now known as BjoernT16:34
*** mikecmpbll has quit IRC16:36
timburkeclayg, if we'd prioritized it sooner...16:44
claygvictim blaming16:50
*** e0ne has quit IRC17:00
*** NM has joined #openstack-swift17:04
*** e0ne has joined #openstack-swift17:16
*** BjoernT has quit IRC17:20
*** e0ne has quit IRC17:22
*** gmann is now known as gmann_afk17:25
*** tesseract has quit IRC17:25
timburkeanybody have opinions on repo layout for swiftclient? we haven't actually been running swiftclient's func tests with tempauth because of a difference between test/sample.conf and tests/sample.conf: https://review.opendev.org/#/c/687773/17:27
patchbotpatch 687773 - python-swiftclient - Actually run tempauth tests in swiftclient dsvm jobs - 2 patch sets17:27
*** e0ne has joined #openstack-swift17:54
*** e0ne has quit IRC17:59
*** psachin has quit IRC18:05
tdasilvatimburke: how bad would it be to rename one of the two?18:07
*** NM has quit IRC18:07
timburkei don't have a strong opinion either way -- but i do often kick myself a bit when working on swiftclient and trying to do tab completion...18:12
timburkeso... sure? i'd propose moving swiftclient to test: less disruptive (since we get more patches against swift than swiftclient), better completion (since test-requirements.txt already makes it stop at "test"), ...18:13
tdasilvayeah, i actually never realized the difference until now18:14
tdasilvatimburke: on a somewhat related note...do you remember the history of why python-swiftclient is running tests with stestr and swift with ostestr?18:15
timburkenot really -- i think it comes down to us slowly moving toward the "OpenStack Way" over so long a time period as to have the target change on us :-/18:17
openstackgerritTim Burke proposed openstack/swift master: swift-account-audit: work with encryption  https://review.opendev.org/68852718:20
openstackgerritTim Burke proposed openstack/swift master: py3: fix swift-account-audit  https://review.opendev.org/68852818:21
openstackgerritTim Burke proposed openstack/swift master: swift-account-audit: clean up some error formatting  https://review.opendev.org/68852918:21
tdasilvatimburke: got it, noticed that earlier today, i miss not having coverage for in-process func tests18:21
tdasilvagoing to try to spend some time moving to stestr18:22
timburke"Support for Python 2.7 will be dropped in stestr release 3.0.0 which is being planned for early 2020." -- I wonder if we should put in version pins now, before it becomes a problem...18:24
*** NM has joined #openstack-swift18:26
tdasilvawhat's the benefit of stestr over nosetests?18:27
*** e0ne has joined #openstack-swift18:28
* timburke shrugs18:28
timburkeman, we have so many little extra utils that don't really get tested... swift-drive-audit came up over the weekend, this morning i noticed swift-account-audit and swift-dispersion-populate, not so long ago i had to fix up swift-orphans...18:28
timburkei think stable/train is going to see a decent bit of activity...18:29
tdasilvatimburke: yeah, i think our cli tools don't get as much test coverage18:30
tdasilva:(18:30
*** e0ne has quit IRC18:30
openstackgerritTim Burke proposed openstack/swift master: py3: fix swift-dispersion-populate  https://review.opendev.org/68853118:34
openstackgerritTim Burke proposed openstack/swift master: py3: Fix swift-recon  https://review.opendev.org/68853318:40
*** zaitcev has joined #openstack-swift18:44
*** ChanServ sets mode: +v zaitcev18:44
zaitcevhttps://twitter.com/bloodeagle40234/status/118321236424760115218:44
timburkewhew -- i was thinking about kota_ this weekend for sure -- glad to hear he's doing ok18:45
timburkeadded a bunch of patches to the top of https://wiki.openstack.org/wiki/Swift/PriorityReviews -- i feel like we should consider a fast-follow 2.23.1 for stable/train...18:52
*** NM has quit IRC19:13
*** NM has joined #openstack-swift19:14
*** e0ne has joined #openstack-swift19:16
*** Jeffrey4l has quit IRC19:23
*** Jeffrey4l has joined #openstack-swift19:24
*** gmann_afk is now known as gmann20:01
zaitcevtest/functional/swift_test_client.py has both io.BytesIO and six.BytesIO. Not like it changes anything, but could we just do one or the other?20:10
openstackgerritThiago da Silva proposed openstack/swift master: WIP: New Object Versioning mode  https://review.opendev.org/68238220:11
*** BjoernT has joined #openstack-swift20:24
*** pcaruana has quit IRC20:26
*** NM has quit IRC20:27
*** benj has joined #openstack-swift20:28
*** benj is now known as Guest3991020:28
*** BjoernT_ has joined #openstack-swift20:32
*** BjoernT has quit IRC20:32
timburkezaitcev, yeah, sounds like a good plan20:35
zigotimburke: I'll try it tonight.20:39
*** NM has joined #openstack-swift20:51
*** diablo_rojo has joined #openstack-swift20:56
openstackgerritMerged openstack/swift master: py3: Fix swift-drive-audit  https://review.opendev.org/68832620:57
*** rcernin has joined #openstack-swift20:57
openstackgerritTim Burke proposed openstack/swift master: Consistently use io.BytesIO  https://review.opendev.org/68857121:03
timburkezaitcev, ^^^ -- though the changes in swob and wsgi make me a little nervous; iirc there's some performance benefit to py2's StringIO -- so i could be talked out of those ones21:05
timburkestill, if we *really* wanted performance, we'd be going for cStringIO...21:06
*** NM has quit IRC21:27
*** NM has joined #openstack-swift21:38
*** jistr has quit IRC21:52
*** rcernin has quit IRC21:59
*** NM has quit IRC22:00
*** jistr has joined #openstack-swift22:01
*** jistr has quit IRC22:02
*** jistr has joined #openstack-swift22:03
*** diablo_rojo has quit IRC22:19
*** NM has joined #openstack-swift22:23
*** NM has quit IRC22:23
*** e0ne has quit IRC22:29
*** e0ne has joined #openstack-swift22:31
*** e0ne has quit IRC22:38
*** diablo_rojo has joined #openstack-swift22:46
*** e0ne has joined #openstack-swift22:57
*** e0ne has quit IRC23:02
*** e0ne has joined #openstack-swift23:05
*** rcernin has joined #openstack-swift23:07
*** e0ne has quit IRC23:14
*** tkajinam has joined #openstack-swift23:17
*** tkajinam_ has joined #openstack-swift23:17
*** e0ne has joined #openstack-swift23:19
*** tkajinam has quit IRC23:21
*** e0ne has quit IRC23:24
*** e0ne has joined #openstack-swift23:25
*** e0ne has quit IRC23:31
*** e0ne has joined #openstack-swift23:35
*** e0ne has quit IRC23:39
*** e0ne has joined #openstack-swift23:46
*** e0ne has quit IRC23:53
*** e0ne has joined #openstack-swift23:54
*** e0ne has quit IRC23:59

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