Monday, 2018-03-12

*** yashmurty has joined #openstack-swift00:00
*** Guest27 has joined #openstack-swift00:05
*** Guest27 has quit IRC00:05
*** kallenp has joined #openstack-swift00:17
DHEso if I'm understanding this right, swift does support ipv6 but only in a pure ipv6 environment... no dual stack00:33
*** itlinux has quit IRC00:39
*** tovin07_ has joined #openstack-swift00:44
*** zhongjun has joined #openstack-swift01:17
*** cshastri has joined #openstack-swift01:26
*** gerardgine has joined #openstack-swift02:04
*** openstackgerrit has joined #openstack-swift02:20
openstackgerritwangqi proposed openstack/swift master: Follow the new PTI for document build  https://review.openstack.org/55178602:20
kota_good morning02:29
*** psachin has joined #openstack-swift02:44
*** gerardgine has quit IRC02:50
*** tovin07_ has quit IRC03:03
*** tovin07_ has joined #openstack-swift03:17
*** kallenp has quit IRC03:17
*** armaan has quit IRC03:18
*** armaan has joined #openstack-swift03:18
*** psachin has quit IRC03:18
*** psachin has joined #openstack-swift03:33
mattoliveraukota_: morning03:33
openstackgerritMatthew Oliver proposed openstack/swift master: Experimental swift-ring-composer CLI to build composite rings  https://review.openstack.org/45150704:00
mattoliverau^ just a rebase04:00
*** kei-ichi has quit IRC04:26
openstackgerritwangqi proposed openstack/swift master: Follow the new PTI for document build  https://review.openstack.org/55178604:26
*** kallenp has joined #openstack-swift04:29
*** cshastri has quit IRC05:02
*** germs has joined #openstack-swift05:06
*** germs has quit IRC05:06
*** germs has joined #openstack-swift05:06
*** germs has quit IRC05:11
*** kei-ichi has joined #openstack-swift05:13
*** cshastri has joined #openstack-swift05:16
*** cshastri_ has joined #openstack-swift05:21
*** cshastri has quit IRC05:21
*** geaaru_ has joined #openstack-swift05:43
*** geaaru has quit IRC05:44
*** kallenp has joined #openstack-swift06:02
*** kallenp has joined #openstack-swift06:02
openstackgerritMatthew Oliver proposed openstack/swift master: py3: port cli recon tests  https://review.openstack.org/55104006:03
mattoliverau^ timburke that's a fixed version after running a py3 recon cli.06:03
mattoliverau^ there was fun with urllib2 objects getheader method that doesn't exist. So that will come up again. Worse, in Py3 the urllib request object has a .get_header method, but the urllib HTTPmessage object only has a .get for getting headers.. fun times :P06:06
openstackgerritKota Tsuyuzaki proposed openstack/swift feature/s3api: Avoid global CONF instance  https://review.openstack.org/55127306:15
kota_sounds fantastic journey to py3, mattoliverau06:16
*** yashmurty has quit IRC06:17
*** threestrands_ has joined #openstack-swift06:26
*** yashmurty has joined #openstack-swift06:27
*** andymccr_ has joined #openstack-swift06:33
*** threestrands has quit IRC06:34
*** zaitcev has quit IRC06:34
*** andymccr has quit IRC06:34
*** kallenp has joined #openstack-swift06:35
*** kallenp has quit IRC06:35
*** yashmurty has left #openstack-swift06:38
*** zaitcev has joined #openstack-swift06:41
*** ChanServ sets mode: +v zaitcev06:41
*** tovin07_ has quit IRC06:54
*** tovin07_ has joined #openstack-swift06:55
*** geaaru_ has quit IRC07:02
*** germs has joined #openstack-swift07:07
*** germs has quit IRC07:07
*** germs has joined #openstack-swift07:07
*** kallenp has joined #openstack-swift07:07
*** kallenp has quit IRC07:08
*** rcernin has quit IRC07:09
*** germs has quit IRC07:11
*** threestrands_ has quit IRC07:13
*** kallenp has joined #openstack-swift07:29
*** kallenp has joined #openstack-swift07:33
*** armaan has quit IRC07:43
*** armaan has joined #openstack-swift07:43
*** kallenp has joined #openstack-swift07:44
*** hseipp has joined #openstack-swift07:50
*** pcaruana has joined #openstack-swift07:56
*** tesseract has joined #openstack-swift08:10
*** kallenp has left #openstack-swift08:31
*** hoonetorg has quit IRC08:36
*** geaaru has joined #openstack-swift08:41
*** hoonetorg has joined #openstack-swift08:50
acolesgood morning08:51
*** germs has joined #openstack-swift09:08
*** germs has quit IRC09:08
*** germs has joined #openstack-swift09:08
*** germs has quit IRC09:12
*** mahatic has quit IRC09:16
*** d0ugal_ has quit IRC09:19
*** d0ugal has joined #openstack-swift09:19
*** root_743 has joined #openstack-swift09:20
*** EmilienM has left #openstack-swift09:23
kota_good morning acoles09:44
* kota_ is now at home because worker's union force to get me out of my office09:45
acoleskota_: lol - why?09:59
kota_acoles: annual spring laborer offensive10:00
kota_I'm not sure if the culture exists only in Japan or not.10:01
acolesoh, so not only you?10:01
kota_just googling the translation to English10:01
kota_acoles: yes. I'm not only guy that out of my office. all of coworkers should be out to show their opinions to our company10:02
kota_i don't know it's really effective (or not)10:03
*** tovin07_ has quit IRC10:03
acolesit happens less in uk than in the past, since the laws controlling workers unions were changed in the 1980s10:04
kota_interesting10:04
openstackgerritAlistair Coles proposed openstack/swift feature/deep: Make ShardRange.update_state idempotent  https://review.openstack.org/55133210:06
openstackgerritAlistair Coles proposed openstack/swift feature/deep: Add exclude_states arg to ContainerBroker.get_shard_ranges  https://review.openstack.org/55135510:06
openstackgerritAlistair Coles proposed openstack/swift feature/deep: Add EXPANDING state to ShardRange  https://review.openstack.org/55133110:06
openstackgerritAlistair Coles proposed openstack/swift feature/deep: Include other shard ranges when updating root  https://review.openstack.org/55133310:06
openstackgerritAlistair Coles proposed openstack/swift feature/deep: Use explicit state rather than subtlety to control when we update root  https://review.openstack.org/55135610:06
*** geaaru has quit IRC10:15
*** cshastri_ has quit IRC10:20
*** geaaru has joined #openstack-swift10:28
openstackgerritKota Tsuyuzaki proposed openstack/python-swiftclient master: Add force auth retry mode in swiftclient  https://review.openstack.org/55195610:52
*** germs has joined #openstack-swift11:08
*** germs has quit IRC11:08
*** germs has joined #openstack-swift11:08
*** germs has quit IRC11:13
*** Supun has joined #openstack-swift11:33
zigoHi guys, I'm having weird unit tests failures in Debian Sid with swift: http://paste.openstack.org/show/698716/12:20
zigoCan someone have a look to this cut/paste?12:20
*** mvk has quit IRC12:21
*** Supun has quit IRC12:25
*** Supun has joined #openstack-swift12:26
*** Supun has quit IRC12:37
*** mvk has joined #openstack-swift12:37
*** Supun has joined #openstack-swift12:38
*** Supun has quit IRC12:55
*** Supun has joined #openstack-swift12:55
openstackgerritMerged openstack/swift feature/deep: Make ShardRange.update_state idempotent  https://review.openstack.org/55133213:00
openstackgerritMerged openstack/swift feature/deep: Add exclude_states arg to ContainerBroker.get_shard_ranges  https://review.openstack.org/55135513:08
*** Supun has quit IRC13:45
*** arunman has joined #openstack-swift13:54
*** Supun has joined #openstack-swift13:57
*** ianychoi has joined #openstack-swift13:59
*** pcaruana has quit IRC14:01
*** zhongjun has quit IRC14:06
*** pcaruana has joined #openstack-swift14:15
*** arunman has quit IRC14:30
*** pcaruana has quit IRC14:30
*** Supun has quit IRC14:33
*** hseipp has quit IRC14:33
*** kallenp has joined #openstack-swift14:33
*** arunman has joined #openstack-swift14:41
*** pcaruana has joined #openstack-swift14:43
*** cshastri_ has joined #openstack-swift14:48
*** psachin has quit IRC15:16
*** cshastri_ has quit IRC15:28
*** gyee has joined #openstack-swift15:41
openstackgerritAlistair Coles proposed openstack/swift feature/deep: Use state rather than subtlety to make shards of shards authoritative  https://review.openstack.org/55135615:43
openstackgerritAlistair Coles proposed openstack/swift feature/deep: Use state rather than subtlety to make shards of shards authoritative  https://review.openstack.org/55135615:53
*** hseipp has joined #openstack-swift15:59
*** kallenp has left #openstack-swift16:00
timburkezigo: what fs is /tmp using? i wonder if it doesn't support O_TMPFILE...16:00
notmynamegood morning16:09
*** pcaruana has quit IRC16:12
notmynamezigo: interesting. looks maybe related to https://github.com/openstack/swift/commit/773edb4a5dd22b22f5fcab57820bcbdd4176dfcd ? does debian do something with XFS and O_TMPFILE support?16:15
notmynamezigo: oh! maybe your /tmp (or TMPFILE) doesn't do the right thing? tmpfs isn't good. let me find that patch...16:17
notmynamezigo: https://github.com/openstack/swift/commit/728b4ba14008f59ffe64fd28bc061c12a0c6446716:18
notmynamespecifically, from that commit message:16:18
notmynameWith this patch, every test environment will require TMPDIR to be16:18
notmynameusing a filesystem that supports at least 4k of extended attributes.16:18
notmynameNeither ext4 nor tempfs support this. XFS is recommended.16:18
zigonotmyname: Yeah, using tmpfs for /tmp is quite common.16:19
notmynamezigo: if that's the case, then you've found an area that should be skipped but was missed in that patch16:19
*** armaan_ has joined #openstack-swift16:19
zigonotmyname: :)16:19
zigoonovy: See above ? ^16:19
notmynameyeah, unfortunately our choice was to require something different for TMPDIR or mock out an entire filesystem16:19
notmynameso we chose to require tests to have TMPDIR to support at least 4k of xattrs. practically, this means "XFS"16:20
zigoI first thought about removing my ramdisk for my build env, but forgot about /tmp...16:20
zigoSilly me.16:20
zigonotmyname: What if I do export TMPDIR=`pwd`/debian/tmpdir when running the tests?16:21
zigoWill that fix things?16:21
notmynameone way to find out :-)16:21
zigo(this will still fail under *some* build env. though...)16:21
zigonotmyname: Yeah, /me tries ...16:21
notmynameI'd expect that to work, if the xattr thing is the issue16:22
zigonotmyname: it can only be a temp fix though...16:22
notmynamehowever, the pastebin is talking about linkat... so, maybe?16:22
*** armaan__ has joined #openstack-swift16:22
zigonotmyname: Where I build, everything is in tmpfs, as I run sbuild in AUFS over tmpfs ...16:22
zigos/AUFS/overlayfs/16:23
*** kallenp has joined #openstack-swift16:23
*** armaan has quit IRC16:23
notmynamezigo: swift requires an XFS TMPDIR to runa nd pass all tests. without that, we intend to pass test runs bus skip a vast number of them16:23
zigonotmyname: Got ya.16:23
kallenpWhere to get / find / download script swift-init reffered in step 6 on official swift documentation ? …Debian 9, link to doc https://docs.openstack.org/swift/queens/install/finalize-installation-ubuntu-debian.html16:24
notmynamekallenp: that comes with swift16:24
kallenpwhere ? It's not present in my Debian 916:25
notmynamehow did you install it?16:25
kallenpfollowed this guide: https://docs.openstack.org/swift/queens/install/finalize-installation-ubuntu-debian.html16:26
zigonotmyname: If I understand you well, we'd just need to add the @requires_o_tmpfile_support to fix things, right?16:26
zigo(I mean, in the failing tests...)16:26
*** armaan_ has quit IRC16:26
notmynamezigo: oh, is that a decorator we have? :-)16:26
notmynameI haven't looked at that part of the code in a long time16:27
zigonotmyname: within test/unit/common/test_linkat.py, line 40 of that patch you linked to.16:27
timburkethat doesn't quite look specific enough... and at least one of the failing tests should already have the decorator16:28
* zigo checks16:28
*** Supun has joined #openstack-swift16:29
zigotimburke: Indeed, test_linkat_success has the decorator...16:29
kallenpnotmyname: any idea where to get this script ?16:29
zigoThey all do ...16:30
kallenpnotmyname: not present in Debian 9. I try to install apt-file && apt-file update && apt-file search swift-init - not exist in any Debian 9 * package16:30
zigo:/16:30
* zigo tries the export TMPDIR hack.16:30
*** gerardgine has joined #openstack-swift16:31
timburkezigo: the decorator isn't enough for us, because o_tmpfile_supported doesn't actually look at any fs: https://github.com/openstack/swift/blob/2.17.0/swift/common/utils.py#L4326-L433916:31
kallenpwhat ? I doen understand16:31
zigoOh !16:31
zigotimburke: I know how to fix that then ! :)16:31
timburkei expect that exporting TMPDIR to point at an xfs mount should work fine. that's what i typically do, just because i haven't gotten around to rebuilding my vsaio in a while :P16:32
notmynamekallenp: it's included with swift. whatever packages you installed will have them. I'd assume they would be installed into `/bin`. check your PATH variable16:33
zigotimburke: I'm not sure what o_tmpfile_supported returns though. Is it simply True / False?16:35
timburkeyeah, True/False16:36
kallenpnotmyname: Really not exist in newly installed environment. Try to locat *swift* across the system and NOT EXIST.16:37
kallenpapt list --installed | grep swift16:38
kallenppython-swift/stable,now 2.10.2-1~deb9u1 all [installed]16:38
kallenppython-swiftclient/stable,now 1:3.1.0-3 all [installed]16:38
kallenppython3-swiftclient/stable,now 1:3.1.0-3 all [installed]16:38
kallenpswift/stable,now 2.10.2-1~deb9u1 all [installed]16:38
kallenpswift-account/stable,now 2.10.2-1~deb9u1 all [installed]16:38
kallenpswift-container/stable,now 2.10.2-1~deb9u1 all [installed]16:38
kallenpswift-drive-audit/stable,now 2.10.2-1~deb9u1 all [installed,automatic]16:38
kallenpswift-object/stable,now 2.10.2-1~deb9u1 all [installed]16:38
kallenpswift-proxy/stable,now 2.10.2-1~deb9u1 all [installed]16:38
zigotimburke: Would something like this be ok as a better replacement? https://github.com/openstack/nova/blob/master/nova/utils.py#L133816:38
zigoI was annoyed by it in Nova, because /var/lib/nova/instance was using tmpfs for me when running tests, and then failing, so I know this function a little bit ! :)16:39
zigoThat's for testing O_DIRECT, though it'd be the same idea.16:40
zigoJust try to open a file, and see if that works...16:40
zigoIf you're ok with it, then I can write the patch.16:42
timburkeprobably? i'm generally a fan of trying and handling the failure...16:42
zigo:)16:42
* zigo starts writing the patch then.16:42
*** Supun has quit IRC16:43
timburkethanks16:43
notmynamekallenp: look for other binaries like `swift-proxy-server`. `swift-init` should be in the same place16:53
kallenpreally not16:55
*** arunman has quit IRC17:03
openstackgerritThomas Goirand proposed openstack/swift master: Fix tests using O_TMPFILE  https://review.openstack.org/55209317:10
zigoThere you go !17:11
zigoLet me test the patch in my local build env now ...17:11
zigoIs it only me, or is gerrit super slow? :(17:11
*** pcaruana has joined #openstack-swift17:18
*** hseipp has quit IRC17:24
*** pcaruana has quit IRC17:27
notmynamezigo: that LOG global is going to cause problems17:36
*** SkyRocknRoll has joined #openstack-swift17:43
*** itlinux has joined #openstack-swift17:49
*** pcaruana has joined #openstack-swift17:53
*** armaan__ has quit IRC18:00
*** armaan has joined #openstack-swift18:01
*** d0ugal has quit IRC18:05
*** ggine has joined #openstack-swift18:06
*** tesseract has quit IRC18:06
*** gerardgine has quit IRC18:07
*** kallenp has left #openstack-swift18:11
-openstackstatus- NOTICE: Most jobs in zuul are currently failing due to a recent change to zuul; we are evaluating the issue and will follow up with a recommendation shortly. For the moment, please do not recheck.18:19
*** ChanServ changes topic to "Most jobs in zuul are currently failing due to a recent change to zuul; we are evaluating the issue and will follow up with a recommendation shortly. For the moment, please do not recheck."18:19
*** d0ugal has joined #openstack-swift18:20
*** SkyRocknRoll has quit IRC18:29
*** ChanServ changes topic to "https://bugs.not.mn/project/Swift | Logs: http://eavesdrop.openstack.org/irclogs/%23openstack-swift/ | Meetings: https://wiki.openstack.org/wiki/Meetings/Swift | Review Dashboard: http://not.mn/reviews.html | https://etherpad.openstack.org/p/Dublin_PTG_Swift"18:42
-openstackstatus- NOTICE: Zuul has been restarted without the breaking change; please recheck any changes which failed tests with the error "Accessing files from outside the working dir ... is prohibited."18:42
*** geaaru has quit IRC18:46
openstackgerritAlistair Coles proposed openstack/swift feature/deep: Use state rather than subtlety to make shards of shards authoritative  https://review.openstack.org/55135618:54
openstackgerritAlistair Coles proposed openstack/swift feature/deep: Root leader decides when shard containers shard  https://review.openstack.org/55212018:54
openstackgerritTim Burke proposed openstack/swift feature/deep: Fix up probetests check on feature/deep  https://review.openstack.org/55029119:03
*** openstackgerrit has quit IRC19:04
*** armaan has quit IRC19:16
zigonotmyname: What do you suggest? That I just remove it?19:22
acolestimburke: nice idea on p 550291, I spent a while looking fruitlessly at that today19:36
patchbothttps://review.openstack.org/#/c/550291/ - swift (feature/deep) - Fix up probetests check on feature/deep19:36
timburkewe'll see where it goes19:37
*** openstackgerrit has joined #openstack-swift19:59
openstackgerritThomas Goirand proposed openstack/swift master: Fix tests using O_TMPFILE  https://review.openstack.org/55209319:59
*** mvk has quit IRC20:01
*** Sukhdev has joined #openstack-swift20:04
*** armaan has joined #openstack-swift20:13
openstackgerritThomas Goirand proposed openstack/swift master: Fix tests using O_TMPFILE  https://review.openstack.org/55209320:20
openstackgerritAlistair Coles proposed openstack/swift feature/deep: Remove unnecessary timestamp shenanigans when creating shard container  https://review.openstack.org/55213320:26
*** Sukhdev has quit IRC20:29
openstackgerritThomas Goirand proposed openstack/swift master: Fix tests using O_TMPFILE  https://review.openstack.org/55209320:31
*** armaan has quit IRC20:35
*** armaan has joined #openstack-swift20:35
openstackgerritThomas Goirand proposed openstack/swift master: Fix tests using O_TMPFILE  https://review.openstack.org/55209320:36
openstackgerritTim Burke proposed openstack/swift feature/deep: Fix up probetests check on feature/deep  https://review.openstack.org/55029120:39
openstackgerritThomas Goirand proposed openstack/swift master: Fix tests using O_TMPFILE  https://review.openstack.org/55209320:41
tdasilvatimburke: just curious, why do you need: delegate_to: localhost20:44
timburkeno idea. cribbing from a similar job in zuul-jobs20:48
tdasilvakk20:50
openstackgerritThomas Goirand proposed openstack/swift master: Fix tests using O_TMPFILE  https://review.openstack.org/55209320:53
zigoEventually I'll stop flooding the channel with wrong patches ... :)20:54
*** itlinux has quit IRC20:57
*** ggine has quit IRC20:58
openstackgerritThomas Goirand proposed openstack/swift master: Fix tests using O_TMPFILE  https://review.openstack.org/55209320:59
*** mvk has joined #openstack-swift21:03
zigonotmyname: timburke: This last patch looks good to me, no unit test error anymore in my Debian package ! :)21:05
*** gerardgine has joined #openstack-swift21:18
*** Sukhdev has joined #openstack-swift21:20
*** pcaruana has quit IRC21:36
*** armaan has quit IRC21:38
*** armaan has joined #openstack-swift21:39
*** Sukhdev has quit IRC21:47
*** d0ugal has quit IRC21:48
*** kei-ichi has quit IRC21:48
openstackgerritTim Burke proposed openstack/swift feature/deep: Fix up probetests check on feature/deep  https://review.openstack.org/55029121:49
*** kei-ichi has joined #openstack-swift21:53
*** stewie925 has quit IRC21:53
openstackgerritTim Burke proposed openstack/swift feature/deep: Fix up probetests check on feature/deep  https://review.openstack.org/55029122:20
*** rcernin has joined #openstack-swift22:24
*** Sukhdev has joined #openstack-swift22:33
mattoliveraumorning22:43
*** geaaru has joined #openstack-swift22:47
openstackgerritTim Burke proposed openstack/swift master: Capture probe test logs  https://review.openstack.org/55217423:12
*** germs has joined #openstack-swift23:14
*** germs has quit IRC23:14
*** germs has joined #openstack-swift23:14
*** germs has quit IRC23:18
*** Sukhdev has quit IRC23:20
notmynametimburke: that seams like a reasonable idea23:22
timburkenotmyname: may as well get the goodness on master first. i still gotta run down what's going on with https://review.openstack.org/#/c/550291/ though23:23
patchbotpatch 550291 - swift (feature/deep) - Fix up probetests check on feature/deep23:23
*** kei_yama has joined #openstack-swift23:24
timburkeit seems like some container updates aren't landing until *after* we run the sharder? i guess? it's weird to me that we get "0 scanned" in the stats...23:25
timburkeand we never get the "Starting processing ..." message :-/23:26
*** geaaru has quit IRC23:34
timburkeyay! http://logs.openstack.org/74/552174/1/check/swift-probetests-centos-7/e0d097d/swift/23:35
timburkestill a bit annoying trying to cross-correlate everything, but it's a start. i should maybe also figure out how to gzip these things so they aren't quite so big?23:36
openstackgerritMerged openstack/swift master: Apply remote metadata in _handle_sync_response  https://review.openstack.org/46408423:42
*** itlinux has joined #openstack-swift23:51
kei-ichiI sent review request to both openstack-infra/project-config and openstack/governance just now.23:55
kei-ichiEven this repository's main review is infra-team, I added notmyname mattoliverau kota_ as reviewer of this just in case.23:55
kei-ichiIf you find some mistake or my misunderstanding, please let me know o/23:55
kei-ichiAnyway, good morning !23:55
mattoliveraukei-ichi: great, nice work. I'll take a look :)23:56
mattoliveraukei-ichi: and good morning :)23:56
kei-ichiThanks and good morning  !o!23:56
notmynamekei-ichi: I think you may want to add it as a separate project apart from swift, similar to how pyeclib and liberasurecode and storlets work.23:58

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