Wednesday, 2019-10-02

*** gyee has quit IRC00:08
*** NM has joined #openstack-swift01:02
kota_hello  world01:17
*** NM has quit IRC01:29
*** NM has joined #openstack-swift01:32
*** NM has quit IRC01:41
*** spsurya has joined #openstack-swift01:47
*** irclogbot_1 has quit IRC02:09
*** irclogbot_2 has joined #openstack-swift02:13
*** BjoernT has quit IRC03:16
*** Jeffrey4l has quit IRC03:38
*** Jeffrey4l has joined #openstack-swift03:40
*** henriqueof has quit IRC04:33
*** henriqueof1 has joined #openstack-swift04:33
*** dasp has quit IRC04:40
*** pcaruana has joined #openstack-swift04:52
*** rcernin_ has joined #openstack-swift05:59
*** rcernin has quit IRC06:01
*** dasp has joined #openstack-swift06:40
*** rcernin_ has quit IRC06:45
*** rcernin has joined #openstack-swift06:45
*** tesseract has joined #openstack-swift06:57
*** rcernin has quit IRC07:04
*** diablo_rojo has joined #openstack-swift07:18
*** binwiederhier has quit IRC07:27
*** e0ne has joined #openstack-swift07:57
*** tkajinam has quit IRC08:04
*** mikecmpbll has joined #openstack-swift08:06
*** e0ne has quit IRC08:20
*** mkrai_ has joined #openstack-swift09:12
mkrai_Hi, I am facing issue running devstack due to some error in swift setup.cfg file on stein release09:13
*** diablo_rojo has quit IRC09:13
*** diablo_rojo has joined #openstack-swift09:14
mkrai_This is my conf file
mkrai_and the error is here
mkrai_Can someone please help?09:15
alecuyermkrai_:  This may be caused by an old version of setuptools, would check this : , see if it helps ?09:18
mkrai_alecuyer, Thanks, it worked :)09:24
*** e0ne has joined #openstack-swift09:24
alecuyermkrai_: you're welcome!09:26
*** diablo_rojo has quit IRC10:08
*** mkrai_ has quit IRC10:20
*** rcernin has joined #openstack-swift10:25
*** e0ne has quit IRC10:51
*** diablo_rojo has joined #openstack-swift10:54
*** csmart has quit IRC11:07
*** openstackstatus has quit IRC11:09
*** csmart has joined #openstack-swift11:21
*** rdejoux has joined #openstack-swift11:51
*** ianychoi has quit IRC12:21
*** ianychoi has joined #openstack-swift12:25
*** diablo_rojo has quit IRC12:37
*** diablo_rojo has joined #openstack-swift12:49
*** new_student1411 has joined #openstack-swift12:57
new_student1411Hi, after I disable any account/project in keystone, I am still able to access the containers/objects using previously generated ec2 credentials. Any reason, why is it happening?12:58
*** NM has joined #openstack-swift13:04
*** e0ne has joined #openstack-swift13:13
*** henriqueof1 has quit IRC13:51
*** mkrai_ has joined #openstack-swift13:52
*** diablo_rojo has quit IRC13:52
*** mkrai_ has quit IRC14:11
*** gkadam has joined #openstack-swift14:12
*** gkadam has quit IRC14:12
*** baffle has joined #openstack-swift14:14
*** mkrai_ has joined #openstack-swift14:17
*** BjoernT has joined #openstack-swift14:26
*** openstackstatus has joined #openstack-swift14:28
*** ChanServ sets mode: +v openstackstatus14:28
*** rcernin has quit IRC14:30
*** NM has quit IRC14:34
*** mkrai_ has quit IRC14:47
*** mkrai__ has joined #openstack-swift14:47
*** mkrai__ has quit IRC14:51
*** henriqueof has joined #openstack-swift14:55
timburkenew_student1411, there are two potential causes i could see: first, if you configured secret_cache_duration for s3token, any already-cached credentials will continue to be valid until they fall out of the cache15:06
timburkesecond, the project/user disablement in keystone may not be reflected the credentials api -- i'm not sure; haven't looked closely at that code15:08
timburkehow are you disabling things? i could try to repro...15:11
openstackgerritTim Burke proposed openstack/swift master: Fix some request-smuggling vectors on py3
*** gyee has joined #openstack-swift15:30
*** ianychoi has quit IRC15:51
*** BjoernT has quit IRC15:53
*** ianychoi has joined #openstack-swift15:54
*** BjoernT has joined #openstack-swift15:56
*** mikecmpbll has quit IRC16:01
*** mikecmpbll has joined #openstack-swift16:04
claygtimburke: so I think we can get the prefix behavior we want by throwing a "and '\x00' not in name" on the end of query... I'll have to double check that doesn't blow the index16:08
*** pcaruana has quit IRC16:10
*** pcaruana has joined #openstack-swift16:11
rledisezhi, we are enabling s3api and we are having an issue related to the Authorization header. some users were used to store this header in objects metadata and now the requests are rejected because s3token raise the the header is invalid. did you ever hit that case?16:20
*** rdejoux has quit IRC16:23
timburke"some users were used to store this header in objects metadata" -- wait, what?? what would that even *mean*?16:24
rlediseztimburke:  I got wrong informations, sorry, they do a get with that "invalid" header. they do not store it :D16:25
rledisezso, it's totally on them if they have issues16:25
timburkewhew! had me worried there ;-)16:25
timburkefwiw, i added a reject_non_aws_authorization config option as a drive-by in -- i should probably refresh that16:26
patchbotpatch 613452 - swift - s3api: Move authenticator logic to separate module - 4 patch sets16:26
*** tesseract has quit IRC16:29
*** NM has joined #openstack-swift16:37
timburkei might look into breaking that out as a separate patch... then you've got more options, anyway16:40
timburkei'll be really interested in seeing what parts of the API your users use most, and which features they really want :-)16:41
*** mikecmpbll has quit IRC16:52
*** e0ne has quit IRC16:55
*** zaitcev_ is now known as zaitcev17:00
timburkerelease team's getting antsy about creating a stable/train branch -- anybody got spare review bandwidth for ? then i'll turn the crank on getting 2.23.0 out17:08
patchbotpatch 684041 - swift - Fix some request-smuggling vectors on py3 - 3 patch sets17:08
claygtimburke: fwiw thinking about 'nulls in object paths is scary' - I'm starting to think it might be *good* for us17:19
claygwhen we're writing object names out to console like devices (logging, stdout, etc) we *should* be worried about control characters that users could *already* have been sneaking into object names17:19
timburkecertainly true17:20
claygbetter for us to find a "bug" with how we're writing out names on an object that we created than something trying to do something sneaky17:20
claygbut if we have these bugs I think we already have them - null is *a* tricky byte for sure - but I'm unconvinced it's any worse than... backspace?17:21
timburkedid you notice that you can colorize your text listings by dropping some ANSI escape codes into your names? 🤣17:22
claygthat might be debateable... but I for one am glad you're finding these issues and hope we can address a good deal of them by the time we're ready to merge in the feature17:22
*** BjoernT has quit IRC17:22
timburkeah, good ol'
patchbotpatch 300577 - swift - Colorize Account/Container listings (ABANDONED) - 1 patch set17:24
claygyou're the best timburke17:24
*** pcaruana has quit IRC17:45
*** dcourtoi has joined #openstack-swift17:51
openstackgerritMerged openstack/swift stable/pike: Fix stable gate
openstackgerritClay Gerrard proposed openstack/swift master: Allow internal clients to use null namespace
*** rdejoux has joined #openstack-swift18:17
*** e0ne has joined #openstack-swift18:34
*** NM has quit IRC18:37
*** e0ne has quit IRC18:40
*** BjoernT has joined #openstack-swift18:45
*** NM has joined #openstack-swift18:47
*** BjoernT has quit IRC18:49
*** BjoernT has joined #openstack-swift18:52
openstackgerritClay Gerrard proposed openstack/swift master: DNM: Frustrated rant about OOO under a shifting py2/3 base
*** spsurya has quit IRC19:09
*** ccamacho has joined #openstack-swift19:24
*** ccamacho has quit IRC19:28
timburkezaitcev, mattoliverau, clayg: thanks for all the reviews on -- sorry that software is so terrible19:31
patchbotpatch 684041 - swift - Fix some request-smuggling vectors on py3 - 3 patch sets19:31
claygtimburke: yeah I guess you can pipeline a two requests that don't have a body w/o having to send content-length or transfer-encoding... I'd forgotten we don't like connection: close bodies and opt to 41119:33
claygso Q: "why not explicitly set connection: close" seems to A: "because that's unneccessariy and incorrect" 🤣19:34
claygI can see the justification for blocking out whole methods with not PY2 - at first - but also recongize there's probably some limit...19:37
clayglooks like we're mostly done now - so that's good!19:37
claygall we can do is keep trying to make it better - thanks for fixing that nasty bug19:37
*** BjoernT has quit IRC19:37
*** BjoernT has joined #openstack-swift19:47
*** rdejoux has quit IRC19:49
*** NM has quit IRC19:51
*** DHE has quit IRC20:05
*** remi_ness has joined #openstack-swift20:14
*** new_student1411 has quit IRC20:30
*** remi_ness has quit IRC20:54
timburkemeeting time!21:00
timburkeclayg, tdasilva, kota_ ^^^21:01
*** BjoernT has quit IRC21:30
*** BjoernT has joined #openstack-swift21:35
*** e0ne has joined #openstack-swift21:49
*** BjoernT has quit IRC21:50
*** BjoernT has joined #openstack-swift21:54
claygPretty much that time most days. Same time MTuF would be slightly better. If the meeting was more like now I’d have a much better time making it consistently.21:54
claygBut like you said after “fall back” maybe it’s moot.21:54
timburkemakes sense21:55
timburkestill... that's another 4 meetings... just as you're trying to sell us all on versioning...21:58
*** BjoernT has quit IRC22:24
*** e0ne has quit IRC22:35
*** rcernin has joined #openstack-swift22:48
*** tkajinam has joined #openstack-swift22:54
openstackgerritMerged openstack/swift master: Fix some request-smuggling vectors on py3
*** NM has joined #openstack-swift23:11
openstackgerritTim Burke proposed openstack/swift master: Authors/changelog for 2.23.0
*** NM has quit IRC23:15
kota_sorry, I was absent from the meeting23:56
kota_will check the log23:56

Generated by 2.15.3 by Marius Gedminas - find it at!