Monday, 2016-11-21

*** vinsh has quit IRC01:00
*** ChubYann has quit IRC01:07
*** ChubYann has joined #openstack-swift01:08
*** vinsh has joined #openstack-swift01:09
*** MooingLemur has joined #openstack-swift01:10
*** vinsh has quit IRC01:13
kota_good morning02:09
mattoliveraukota_ morning02:16
kota_mattoliverau: hello02:24
*** winggundamth has joined #openstack-swift02:40
*** Jeffrey4l has quit IRC02:51
*** sudorandom has quit IRC02:56
*** ntata has quit IRC02:56
*** sudorandom has joined #openstack-swift02:56
*** vinsh has joined #openstack-swift03:09
*** hugokuo has quit IRC03:12
*** vinsh has quit IRC03:14
*** hugokuo has joined #openstack-swift03:17
*** sanchitmalhotra has joined #openstack-swift03:23
*** sanchitmalhotra has quit IRC03:28
*** tqtran has joined #openstack-swift03:35
*** tqtran has quit IRC03:39
*** links has joined #openstack-swift03:46
charz_mattoliverau: kota_ morning03:56
mattoliveraucharz_: morning03:56
*** links has quit IRC04:17
*** Jeffrey4l has joined #openstack-swift04:42
*** Jeffrey4l has quit IRC04:43
*** Jeffrey4l has joined #openstack-swift04:44
*** MooingLemur has quit IRC05:06
*** MooingLemur has joined #openstack-swift05:12
*** MooingLemur has quit IRC05:14
*** MooingLemur has joined #openstack-swift05:16
*** vinsh has joined #openstack-swift05:16
*** ppai has joined #openstack-swift05:35
*** MooingLemur has quit IRC05:39
*** MooingLemur has joined #openstack-swift05:40
*** vinsh has quit IRC05:53
*** takashi_ has joined #openstack-swift06:03
lifelessnotmyname: to get -x behaviour you need the backend runner to support that today; we could in principle use that to cancel backends once we see a failure - file a wishlist ticket06:07
lifelessnotmyname: SkipTest is totally backend specific - the subunit protocol defines skips semantically, so as long as the backend emits valid subunit there you're golden06:07
*** ChubYann has quit IRC06:57
*** sams-gleb has joined #openstack-swift07:11
*** pcaruana has joined #openstack-swift07:17
*** tqtran has joined #openstack-swift07:35
*** tqtran has quit IRC07:35
*** silor has joined #openstack-swift08:02
*** oshritf has joined #openstack-swift08:06
*** hseipp has joined #openstack-swift08:06
*** oshritf has quit IRC08:06
*** silor1 has joined #openstack-swift08:07
*** silor has quit IRC08:08
*** silor1 is now known as silor08:08
*** sams-gleb has quit IRC08:09
*** sams-gleb has joined #openstack-swift08:09
*** oshritf has joined #openstack-swift08:10
*** sams-gle_ has joined #openstack-swift08:12
*** sams-gleb has quit IRC08:12
*** oshritf has quit IRC08:14
*** rledisez has joined #openstack-swift08:24
*** amoralej|off is now known as amoralej08:27
*** sams-gle_ has quit IRC08:31
*** sams-gleb has joined #openstack-swift08:32
*** sams-gleb has quit IRC08:36
*** ahale has quit IRC08:44
*** ppai_ has joined #openstack-swift08:45
*** ppai has quit IRC08:45
*** cbartz has joined #openstack-swift08:46
*** geaaru has joined #openstack-swift08:47
*** oshritf has joined #openstack-swift08:47
*** joeljwright has joined #openstack-swift08:48
*** ChanServ sets mode: +v joeljwright08:48
*** mahatic_ has joined #openstack-swift08:50
*** mahatic has quit IRC08:51
*** oshritf has quit IRC08:51
openstackgerritTuan Luong-Anh proposed openstack/swift-bench: Replace assertEquals with assertEqual  https://review.openstack.org/40012608:59
*** sanchitmalhotra has joined #openstack-swift09:02
*** sanchitmalhotra has quit IRC09:03
*** sanchitmalhotra has joined #openstack-swift09:03
*** ahale has joined #openstack-swift09:08
*** ahale has quit IRC09:13
onovycschwede: hi, are you there?09:18
*** asettle has joined #openstack-swift09:35
*** abalfour has quit IRC09:47
*** abalfour has joined #openstack-swift09:48
cschwedeonovy: Hi! Yes, I’m here for a bit10:00
*** ahale has joined #openstack-swift10:02
*** sams-gleb has joined #openstack-swift10:02
openstackgerritMathias Bjoerkqvist proposed openstack/swift: WIP: Storing encryption root secret in Barbican  https://review.openstack.org/36487810:15
openstackgerritMathias Bjoerkqvist proposed openstack/swift: WIP: Storing encryption root secret in Barbican  https://review.openstack.org/36487810:21
*** gabor_antal has quit IRC10:25
openstackgerritKota Tsuyuzaki proposed openstack/swift: Fix stats calculation in object-reconstructor  https://review.openstack.org/28394610:54
*** geaaru has quit IRC10:56
kota_acoles: thanks for doing great work at patch 39725911:01
patchbothttps://review.openstack.org/#/c/397259/ - swift - Use per policy diskfile manager in object replicator11:01
kota_acoles: i confirmed it works well so that I squash the patch into patch 28394611:02
patchbothttps://review.openstack.org/#/c/283946/ - swift - Fix stats calculation in object-reconstructor11:02
*** jordanP has joined #openstack-swift11:03
*** tesseract has joined #openstack-swift11:10
*** tesseract is now known as Guest9374011:11
*** geaaru has joined #openstack-swift11:12
onovycschwede: ad your review. what about add "connection timeout" for replicator daemon to config?11:13
onovyi think it's much better to have solo connection and read timeout11:14
onovybecause if host is down, connection timeout mark them as bad really quickly11:14
onovyhttps://review.openstack.org/#/c/391617/11:14
patchbotpatch 391617 - swift - object-replicator cleanup11:14
cschwedeonovy: if you do that you still need a big warning note for deployers that the default behavior changes11:16
cschwedeonovy: because the behavior is different, no matter what the timeout is11:16
cschwedeonovy: even if both timeouts are by default 60 seconds, which is probably not wanted as well11:16
cschwedeonovy: it’s a bit tricky imo. either we keep the current behavior, or we change it but require config updates from every deployer11:18
*** geaaru has quit IRC11:19
*** mvk has quit IRC11:29
*** acoles_ is now known as acoles11:31
*** geaaru has joined #openstack-swift11:39
*** Jeffrey4l has quit IRC11:45
*** takashi_ has quit IRC12:11
*** vint_bra has joined #openstack-swift12:12
acolesgate failures? if your patch is failing gate func tests then you most likely will need to rebase since this landed https://review.openstack.org/39988812:13
patchbotpatch 399888 - swift - Removed "in-process-" from func env tox name (MERGED)12:13
-openstackstatus- NOTICE: We are currently having capacity issues with our ubuntu-xenial nodes. We have addressed the issue but will be another few hours before new images have been uploaded to all cloud providers.12:20
*** kei_yama has quit IRC12:21
openstackgerritAlistair Coles proposed openstack/swift: breaking down tests.py into smaller pieces  https://review.openstack.org/39927512:24
acolestdasilva: ^^ rebase only12:24
*** catintheroof has joined #openstack-swift12:31
*** mvk has joined #openstack-swift12:39
*** Jeffrey4l has joined #openstack-swift12:41
*** Jeffrey4l has quit IRC12:42
*** Jeffrey4l has joined #openstack-swift12:45
*** Jeffrey4l has quit IRC12:47
*** Jeffrey4l has joined #openstack-swift12:49
*** sanchitmalhotra has quit IRC12:50
*** Jeffrey4l has quit IRC12:52
*** Jeffrey4l has joined #openstack-swift12:52
openstackgerritMahati Chamarthy proposed openstack/swift: Move documented reclaim_age option to correct location  https://review.openstack.org/37441912:58
openstackgerritMahati Chamarthy proposed openstack/swift: Reduce backend requests for SLO If-Match / HEAD requests  https://review.openstack.org/34753812:59
openstackgerritMahati Chamarthy proposed openstack/swift: Confirm receipt of SLO PUT with etag  https://review.openstack.org/39090113:05
mahatic_sorry for the noise, rebasing13:08
mahatic_timburke: did one of yours - 347538 since 390901 is dependent on it13:10
*** SkyRocknRoll has joined #openstack-swift13:12
*** amoralej is now known as amoralej|lunch13:24
*** Jeffrey4l has quit IRC13:26
*** mvk has quit IRC13:29
*** ppai_ has quit IRC13:29
*** klamath has joined #openstack-swift14:07
onovycschwede: but if it is good change, why don't do it?14:09
onovyyep, it's not backward compatible, but i should get better timeout when node is done14:10
onovy*it14:10
*** lxsli_away is now known as lxsli14:20
*** _JZ_ has joined #openstack-swift14:27
tdasilvaacoles: thanks!14:31
acolestdasilva: np, good morning14:32
tdasilvaacoles: good afternoon to you sir :)14:33
tdasilvaacoles: just noticed you and jrichli +A patch 399275, thanks! I saw your comments there only after I had sent patchset 2, so I didn't ignore your comments14:34
patchbothttps://review.openstack.org/#/c/399275/ - swift - breaking down tests.py into smaller pieces14:34
acolestdasilva: I think my comments reduced to "IDK, anything <5k lines works" ;)14:34
tdasilvaacoles: agree with keeping TestFile, TestContainers and TestAccounts in tests.py (at least for now) and also agree with not doing any refactoring right now, but i think it would be best to break up the features (middleware) in different files14:35
tdasilvajust cause we keep adding new tests (e.g., symlinks will add about 1K lines of tests) so keeping all feature tests in one file seems like we would just end up in the same place we are today14:36
*** amoralej|lunch is now known as amoralej14:37
acolestdasilva: yes, agree. I'm sure there will be times when its hard to know which feature a test should be added under (e.g symlinks+versions), but maybe we just go with arbitrary precedence - e.g. symlinks+versions tests (if there are any) would go in test_versioned_writes, other symlink tests in test_symlinks ??14:39
acolestdasilva: jrichli made a point about duplicated test filenames, might be worth considering14:40
tdasilvaacoles: yeah, honestly I can't come up with a great solution for the multiple feature test, so I agree it will end up being pretty arbitrary where those tests go. Hopefully there won't be too many of those that create a real problem. but I think what you suggested is better than having a test_symlink_versions.py file14:43
*** mvk has joined #openstack-swift14:45
tdasilvaacoles, jrichli: in terms of that duplicated names I wasn't too concerned tbh. We will now have test/unit/common/middleware/test_versioned_writes.py and test/functional/test_versioned_writes.py, but IMO that's ok, i think it's less confusing than naming it something else, plust we already have obj/server.py and container/server.py and so on14:45
openstackgerritMathias Bjoerkqvist proposed openstack/swift: WIP: Storing encryption root secret in Barbican  https://review.openstack.org/36487814:46
*** ntata has joined #openstack-swift14:50
*** ntata has quit IRC14:53
openstackgerritAlistair Coles proposed openstack/swift: Update README.md to clarify policies allowed  https://review.openstack.org/40028614:57
openstackgerritAlistair Coles proposed openstack/swift: go: Update README.md to clarify policies allowed  https://review.openstack.org/40028614:59
*** gabor_antal has joined #openstack-swift15:12
*** oshritf has joined #openstack-swift15:12
*** sams-gle_ has joined #openstack-swift15:12
*** sams-gleb has quit IRC15:15
*** sams-gle_ has quit IRC15:16
*** oshritf has quit IRC15:23
*** oshritf has joined #openstack-swift15:24
*** sams-gleb has joined #openstack-swift15:27
cschwedeonovy: i’m not against it per se, but in the case with two configurable timeouts the patch will need some notes as well, and we need to think about some sane defaults. a 5 second default timeout with the patch applied is to short imo, when it was 60 seconds before15:29
onovycschwede: we have 500ms timeout from proxy->store15:30
onovyis 10x really too short?15:30
cschwedeonovy: if you have global clusters then it is likely that the proxy never talks to a remote storage server directly, and is using a slow wan link - 5 second might be enough, but it was 60 seconds before and we should take that into account15:31
onovyah, right. global clusters15:31
*** vinsh has joined #openstack-swift15:33
cschwedeonovy: the challenge here is that the replication itself might have finished successfully, but then we can’t finalize it because this small requests hits a timeout - and boom you need to start over later again15:34
onovyyep, that's right. my motivation is to timeout as fast as possible if node is not available15:34
onovyso maybe we should have bigger timeout AFTER rsync, and smaller before15:35
cschwedeonovy: sounds like a good idea!15:35
onovyi think i will split it into two patches15:36
onovyfirst: refactor it for direct_client15:36
onovysecond: change it to two timeouts15:36
*** siva_krish has joined #openstack-swift15:39
*** siva_krish has left #openstack-swift15:41
*** oshritf has quit IRC15:43
*** rledisez has quit IRC15:45
*** oshritf has joined #openstack-swift15:46
*** oshritf has quit IRC15:53
*** pcaruana has quit IRC16:06
*** Guest93740 has quit IRC16:11
openstackgerritOndřej Nový proposed openstack/swift: object-replicator cleanup  https://review.openstack.org/39161716:13
onovycschwede: ^ thanks16:16
*** oshritf has joined #openstack-swift16:16
*** douglascorrea has joined #openstack-swift16:24
*** tuan_luong has joined #openstack-swift16:28
*** oshritf has quit IRC16:29
*** oshritf has joined #openstack-swift16:33
notmynamegood morning16:34
*** Tuan_ has joined #openstack-swift16:36
*** tuan_luong has quit IRC16:39
*** Tuan_ has quit IRC16:47
notmynameFYI https://review.openstack.org/#/c/394261/ is a doc spec that seems to imply the swift team is signing up for writing an upgrade guide in the ocata cycle16:47
patchbotpatch 394261 - docs-specs - [ops-guide] Add project upgrade notes16:47
*** asettle has quit IRC17:00
*** mmotiani1 has joined #openstack-swift17:14
*** hseipp has quit IRC17:28
*** nikivi has joined #openstack-swift17:39
*** nikivi has quit IRC17:52
*** jordanP has quit IRC18:01
*** dmorita has joined #openstack-swift18:01
*** dmorita has quit IRC18:03
*** dmorita_ has joined #openstack-swift18:03
*** oshritf has quit IRC18:03
*** vinsh has quit IRC18:13
onovynotmyname: hi18:15
onovynotmyname: sry for this: https://review.openstack.org/#/c/399892/ if you don't want it18:15
patchbotpatch 399892 - openstack-infra/project-config - Swift: Make Python 3.5 check voting (MERGED)18:15
onovybut i wrote: please wait for John's comment, before +A this :) thanks18:16
onovythey merged :]18:16
patchbotError: Spurious "]".  You may want to quote your arguments with double quotes in order to prevent extra brackets from being evaluated as nested commands.18:16
notmynameonovy: meh. I do appreciate your comment on it :-)18:16
*** joeljwright has quit IRC18:16
onovyso you are fine with py35 gating (voting)?18:16
notmynameprobably18:16
notmynameI mean, I am now! ;-)18:17
onovycool, thanks :)18:17
*** ChubYann has joined #openstack-swift18:20
notmynameacoles: tdasilva: mattoliverau: kota_: jrichli: cschwede: I want to make sure you saw the change that onovy made this weekend (that I pushed through) to the tox.ini. had to fix the gate, but it might affect your local test environment https://review.openstack.org/#/c/399888/18:21
patchbotpatch 399888 - swift - Removed "in-process-" from func env tox name (MERGED)18:21
notmynameonovy: thanks again for being on top of that18:21
*** torgomatic has joined #openstack-swift18:22
*** ChanServ sets mode: +v torgomatic18:22
*** tqtran has joined #openstack-swift18:23
acolesnotmyname: saw it, gate was failing patches that weren't rebased on it18:24
tdasilvanotmyname: ack18:25
timburkegood morning18:25
*** oshritf has joined #openstack-swift18:27
openstackgerritAlistair Coles proposed openstack/swift: Fix stats calculation in object-reconstructor  https://review.openstack.org/28394618:33
*** oshritf has quit IRC18:33
openstackgerritMerged openstack/swift-bench: Replace assertEquals with assertEqual  https://review.openstack.org/40012618:33
acoleskota_: ^^ improved the unit test that got impacted by replicator change18:33
onovynotmyname: np!18:40
acolesonovy: notmyname out of curiosity,what was the problem that required the tox env name change?18:40
notmynameacoles: linux has a 128 char limit on the shebang line18:41
notmynameand infra builds the tox env to include the job name and the env name18:41
notmynameso with a long env name (which also is generally similar to the job name), you go over rthe limit18:41
notmynamefunc-in-process-encryption, IIRC, was the longest18:42
notmynamealso because I added -xfs-tmpdir to all of the job names. that was the change since last week18:42
acolesnotmyname: OIC ^^ so it was your fault ;)18:43
notmynameyep. all my fault18:43
acoleshummingbird will need a merge from master to get the new tox envs18:43
*** amoralej is now known as amoralej|off18:44
*** vinsh has joined #openstack-swift18:48
notmynameah, good point18:50
*** acoles is now known as acoles_18:50
*** jamielennox is now known as jamielennox|away18:52
torgomaticWeinberg's Second Law: If builders built buildings the way programmers wrote programs, then the first woodpecker that came along would destroy civilization.18:53
*** jamielennox|away is now known as jamielennox19:00
*** cbartz has quit IRC19:02
*** chlong has joined #openstack-swift19:03
onovya! john's fail, /me fixed! syc!19:06
notmyname:-)19:08
*** geaaru has quit IRC19:08
notmynamehmm... https://review.openstack.org/400378 didn't get announced19:10
patchbotpatch 400378 - swift (stable/newton) - 2.10.1 authors/changelog updates19:10
notmynameah, probably hasn't been configured for that branch19:10
notmynameI'll hold off on the mitaka one until the 2 in queue get landed19:11
*** stream10 has joined #openstack-swift19:14
*** diogogmt has joined #openstack-swift19:26
openstackgerritMathias Bjoerkqvist proposed openstack/swift: WIP: Storing encryption root secret in Barbican  https://review.openstack.org/36487819:35
*** silor has quit IRC19:41
*** vinsh has quit IRC19:45
*** vinsh has joined #openstack-swift19:47
*** douglascorrea has quit IRC20:07
*** douglascorrea has joined #openstack-swift20:08
*** vinsh has quit IRC20:10
*** douglascorrea has quit IRC20:12
*** asettle has joined #openstack-swift20:26
*** zaitcev has joined #openstack-swift20:38
*** ChanServ sets mode: +v zaitcev20:38
*** ndk_ has quit IRC20:49
*** ndk_ has joined #openstack-swift20:51
mattoliverauMorning20:55
*** asettle has quit IRC21:08
*** asettle has joined #openstack-swift21:08
*** vinsh has joined #openstack-swift21:12
*** asettle has quit IRC21:13
openstackgerritOndřej Nový proposed openstack/swift: Added comment for "user" option in drive-audit config  https://review.openstack.org/40040621:14
onovytimburke: :)21:15
timburkeonovy: i didn't mean you had to do it *right now* :P21:15
onovy:)21:15
*** portante has quit IRC21:19
*** portante has joined #openstack-swift21:20
*** stream10 has quit IRC21:36
*** sams-gleb has quit IRC21:52
*** sams-gleb has joined #openstack-swift21:52
*** sams-gleb has quit IRC21:57
*** mmotiani1 has quit IRC21:59
*** vint_bra has quit IRC22:14
openstackgerritTim Burke proposed openstack/swift: Reduce backend requests for SLO If-Match / HEAD requests  https://review.openstack.org/34753822:41
*** chlong has quit IRC22:52
notmynamehttps://review.openstack.org/400435 for 2.7.122:53
patchbotpatch 400435 - swift (stable/mitaka) - 2.7.1 changelog updates22:53
notmynamekota_: hisashi: hope everyone is ok22:57
kota_notmyname: thx, I'm safe and around Tokyo it seems not so impacted23:15
*** SkyRocknRoll has quit IRC23:21
*** kei_yama has joined #openstack-swift23:23
*** _JZ_ has quit IRC23:25
*** douglascorrea has joined #openstack-swift23:28
*** tqtran has quit IRC23:42
*** tqtran has joined #openstack-swift23:42
*** openstack has joined #openstack-swift23:45
*** douglascorrea has quit IRC23:48
*** douglascorrea has joined #openstack-swift23:49
*** douglascorrea has quit IRC23:53

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!