Wednesday, 2017-12-13

*** edausq has joined #openstack-swift00:01
*** joeljwright has quit IRC00:02
*** tovin07_ has joined #openstack-swift00:42
*** two_tired has joined #openstack-swift01:09
m_kazuhirogood morning01:13
claygtimburke: I was wondering why it was the way it was... strange thing was it seemed to fix the behavior... maybe my test is bogus...01:19
claygm_kazuhiro: you’re doing great!  Can you double check timburke and I on that symlink copy issue I referenced in my review and squash in whatever needs fixin01:20
timburkedunno... maybe there's some *other* place where we happen to grab and set req.params again? so modifying the cached version ultimately modifies the real query string, too?01:20
timburkei feel like i'd really prefer the req.params thing to be a context manager or something... but it'd still probably suck01:23
*** vint_bra has joined #openstack-swift01:35
*** cshastri has joined #openstack-swift01:36
*** gyee has quit IRC01:36
m_kazuhiroclayg: ok. I will check it.01:40
kota_good morning01:42
mattoliveraukota_, m_kazuhiro: morning02:01
kota_mattoliverau: o/02:04
* kota_ is looking the irc logs...02:04
*** JimCheung has quit IRC02:16
*** JimCheung has joined #openstack-swift02:24
m_kazuhiromattoliverau: morning02:26
*** JimCheung has quit IRC02:41
kota_timburke, clayg: Did https://gist.github.com/clayg/a883d21723bda1a947109be2d33e2be2 fail w/o the fix on your environment? I tried to run it but those 2 tests didn't fail in my local.02:47
claygHrmmm02:49
claygThat’s good though!  :P02:49
*** m_kazuhiro_ has joined #openstack-swift02:50
kota_let me take time more to consider what's problem there. Exactly, I found we are lack of the unit tests of copy + symink assertions.02:52
kota_clayg: ^^02:52
*** m_kazuhiro has quit IRC02:54
*** JimCheung has joined #openstack-swift03:06
*** JimCheung has quit IRC03:11
*** dancn has joined #openstack-swift03:24
*** kei_yama has quit IRC03:44
*** bkopilov has quit IRC04:00
*** kei_yama has joined #openstack-swift04:01
*** threestrands has joined #openstack-swift04:14
*** threestrands has quit IRC04:14
*** threestrands has joined #openstack-swift04:14
*** jappleii__ has quit IRC04:15
*** threestrands has quit IRC04:15
*** threestrands has joined #openstack-swift04:16
*** threestrands has quit IRC04:16
*** threestrands has joined #openstack-swift04:16
*** links has joined #openstack-swift04:30
*** openstackgerrit has joined #openstack-swift04:31
openstackgerritMerged openstack/swift master: Refactor proxy-server conf loading to a helper function  https://review.openstack.org/52572804:31
*** links has quit IRC04:32
*** threestrands has quit IRC04:36
*** threestrands has joined #openstack-swift04:43
*** threestrands has quit IRC04:43
*** threestrands has joined #openstack-swift04:43
*** threestrands has quit IRC04:44
*** threestrands has joined #openstack-swift04:44
*** threestrands has quit IRC04:44
*** threestrands has joined #openstack-swift04:44
*** threestrands has quit IRC04:45
*** threestrands has joined #openstack-swift04:46
*** threestrands has quit IRC04:46
*** threestrands has joined #openstack-swift04:46
*** threestrands has quit IRC04:47
*** threestrands has joined #openstack-swift04:47
*** threestrands has quit IRC04:47
*** threestrands has joined #openstack-swift04:47
openstackgerritMerged openstack/swift master: Ringbuilder: Forbid writing empty rings  https://review.openstack.org/52519205:00
*** two_tired has quit IRC05:15
*** frank_young has joined #openstack-swift05:22
*** frank_young has quit IRC05:23
claygok... i think i'm crazy :'(05:31
claygi think i had a bug in test but I also thought I had successfully confirmed the issue with curl 😫05:32
claygso if put w/ x-copy-from *works* with symlink=get - does anyone know *why* ???  how does the qs makes it's way onto the new GET request?05:32
claygok, the source_req = req.copy_get() is just copying the headers - i obviously did not verify with curl (or rather tried to and hand a typo and made a mistake)05:36
claygI'll purpose the versioned writes tests as a follow up!05:36
kota_clayg: the fix you pushed looks like possibly a bug exists in the query_args is `symlink=get&multipart-manifest=get`to me.05:38
kota_I'm just starting the assertion in my local via adding unit test.05:39
claygyeah that "fix" isn't useful - timburke said there was some *other* problems too :'(05:39
kota_yup05:39
claygthe fuctests are good tho - they prove symlink + versioned writes are awesome!05:39
claygi'm fixing my review now05:39
openstackgerritClay Gerrard proposed openstack/swift master: functest for symlink + versioned writes  https://review.openstack.org/52758305:41
claygeverything says it's in merge conflict tho - not sure what landed - but I think after a rebase patch 232162 is going to be ready to go!  sorry it took me a good 3 days!?05:46
patchbothttps://review.openstack.org/#/c/232162/ - swift - Symlink implementation.05:46
*** m_kazuhiro has joined #openstack-swift05:53
*** bkopilov has joined #openstack-swift05:54
*** m_kazuhiro_ has quit IRC05:54
*** threestrands has quit IRC06:03
*** d0ugal has quit IRC06:04
*** d0ugal has joined #openstack-swift06:08
kota_clayg, m_kazuhiro: can i work on the rebase?06:10
*** armaan_ has joined #openstack-swift06:11
*** armaan has quit IRC06:11
*** tovin07_ has quit IRC06:12
*** armaan_ has quit IRC06:16
kota_oh i hit the ring part swapping test failure that probably mattoliverau is working on the fix for the flakey test failure.06:17
kota_i confirmed m_kazuhiro locally i can push the rebased version.06:21
openstackgerritKota Tsuyuzaki proposed openstack/swift master: Symlink implementation.  https://review.openstack.org/23216206:22
kota_clayg: it's just a rebased version of your +2 to resolve the merge conflict.06:22
openstackgerritKota Tsuyuzaki proposed openstack/swift master: Add more assertions for Symlink + Copy unit tests  https://review.openstack.org/52759506:26
openstackgerritKota Tsuyuzaki proposed openstack/swift master: Add more assertions for Symlink + Copy unit tests  https://review.openstack.org/52759506:27
*** tovin07_ has joined #openstack-swift06:33
*** cshastri has quit IRC06:57
*** bkopilov has quit IRC07:11
*** rcernin has quit IRC07:12
*** armaan has joined #openstack-swift07:12
*** cshastri has joined #openstack-swift07:15
*** bkopilov has joined #openstack-swift07:29
*** hseipp has joined #openstack-swift07:41
*** armaan has quit IRC07:53
*** armaan has joined #openstack-swift07:53
openstackgerritKota Tsuyuzaki proposed openstack/swift master: Symlink implementation.  https://review.openstack.org/23216207:53
kota_sorry, I failed to rebase, i resolved pep8 issues there.07:54
*** rcernin has joined #openstack-swift08:13
*** cshastri has quit IRC08:21
*** oshritf has joined #openstack-swift08:24
openstackgerritKota Tsuyuzaki proposed openstack/swift master: Add more assertions for Symlink + Copy unit tests  https://review.openstack.org/52759508:25
*** tesseract has joined #openstack-swift08:25
*** cshastri has joined #openstack-swift08:34
*** kei_yama has quit IRC08:46
*** tovin07_ has quit IRC08:46
*** tovin07_ has joined #openstack-swift08:46
hseipptdasilva: Great to see that you were aware of that IoT session, did see it as well but was actually running a "Mini Maker Fair" at our site yesterday presenting the following cool IoT app: https://github.com/IBM/air-traffic-control?cm_sp=IBMCode-_-cloud-based-air-traffic-control-application-_-Get-the-Code, so was not able not join ;-)08:55
tdasilvahseipp: that sounds really cool, going to have to check it out08:56
*** m_kazuhiro has quit IRC08:58
hseipptdasilva: If you do so, feel free to PM me in case you run into issues, I've got some patches to the Swift code (Apple Swift != OpenStack Swift) :-)08:58
tdasilvahseipp: cool, will do. Yeah I did notice the swift words on the readme but noticed it referred to apple's swift :)09:00
*** cbartz has joined #openstack-swift09:00
*** links has joined #openstack-swift09:00
*** oshritf has quit IRC09:07
*** oshritf has joined #openstack-swift09:08
*** oshritf has quit IRC09:13
*** cebruns_ has quit IRC09:20
mattoliverauhseipp: everytime you mention swift in the readme can you say something like (language not the object storage system) so we can slowly fight back ;)09:21
tdasilvaheh09:21
*** cebruns has joined #openstack-swift09:22
hseippmattoliverau: :-) I give up here, as I can't address the Chinese version of the readme09:23
acolesgood morning09:42
tdasilvaacoles: o/09:46
acolestdasilva: up early?09:46
tdasilvaacoles: i'm in israel this week09:47
acolestdasilva: oh. up late?09:47
acoles;)09:47
tdasilvaheh09:47
tdasilvaacoles, rledisez, cschwede I just mentioned to the fosdem sds chairs that we are willing to be flexible with presentation times09:48
*** tovin07_ has quit IRC10:07
openstackgerritAlistair Coles proposed openstack/swift master: Use symlink in container-sync internal client pipeline  https://review.openstack.org/52712610:13
mattoliverauLol10:20
*** aluria has quit IRC10:24
*** dosaboy has quit IRC10:24
*** aluria has joined #openstack-swift10:25
*** dosaboy has joined #openstack-swift10:25
*** dosaboy has quit IRC10:31
*** cshastri has quit IRC10:35
*** dosaboy has joined #openstack-swift10:37
kota_hello acoles and tdasilva o/10:40
acoleskota_: o/10:41
kota_interesting fosdem10:42
kota_acoles: thanks for rebasing that patch10:43
* kota_ means patch 52712610:44
patchbothttps://review.openstack.org/#/c/527126/ - swift - Use symlink in container-sync internal client pipe...10:44
acoleskota_: NP. I think it worth considering including it in the parent, I left comment on patch 232162. It needs other opinions for a decision.10:46
patchbothttps://review.openstack.org/#/c/232162/ - swift - Symlink implementation.10:46
kota_acoles: i see thanks. that's the one I minded to the last before I added my +2.10:46
* kota_ is looking acoles comment10:47
kota_hmm... it looks fair point.10:49
kota_however, I'm still not sure the configuration should be in the pipeline for now. I think fortunately, we will have swift-meeting tomorrow morning so we can have the discussion over there.10:50
kota_intuitively, i don't like to have special config for container-sync internal client as possible.10:50
kota_from my look, we have a couple of different solution so far10:51
kota_a) encryption10:51
kota_that should be in internal client pipeline. that is why acoles proposed the patch.10:52
kota_b) slo10:52
*** SkyRocknRoll has joined #openstack-swift10:52
kota_however, slo resolves the transfer problem w/o middleware config pipeline10:52
kota_i know that is able because slo doesn't use sysmeta from histrical reason so that remote cluster can resolve the problem w/o local internal client pipeline change.10:53
kota_tbh, i'd like to avoid operator configuration as possible though. not sure which is suitable for us, right now.10:55
kota_anyway, thx acoles for bringing up this10:57
*** mvk has quit IRC10:58
acoleskota_: slo is also weird because we need to send the remote cluster the internal format and bypass slo on the remote cluster, to avoid the remote trying to check the manifest. so it is fortunate that x-static-large-object *can* be passed to the remote cluster, whereas sysmeta cannot.11:02
acoleskota_: let's discuss at the meeting11:05
kota_acoles: yes, it's true.11:05
kota_acoles: :)11:05
acolesI'll add to agenda11:05
kota_thanks!11:06
* kota_ has to re-check my alarm to ring my morning11:10
*** armaan has quit IRC11:11
*** armaan has joined #openstack-swift11:12
*** SkyRocknRoll has quit IRC11:15
*** geaaru has joined #openstack-swift11:16
acoleskota_: good night11:19
kota_g'night11:19
*** mvk has joined #openstack-swift11:23
*** SkyRocknRoll has joined #openstack-swift11:27
*** silor has joined #openstack-swift11:28
*** armaan has quit IRC11:41
*** armaan has joined #openstack-swift11:41
*** armaan has quit IRC11:41
*** armaan has joined #openstack-swift11:42
*** bkopilov has quit IRC12:07
*** armaan has quit IRC12:31
*** armaan has joined #openstack-swift12:32
*** armaan has quit IRC12:37
*** armaan has joined #openstack-swift12:38
*** vint_bra has quit IRC12:39
*** armaan has quit IRC12:56
*** armaan has joined #openstack-swift12:57
*** rcernin has quit IRC13:04
acolesmattoliverau: timburke: I'd like to get https://review.openstack.org/#/c/510978 merged - I'm sure it is not perfect but I'm finding I am writing conflicts as I try to move forwards on other fronts (e.g. container delete) - any objections to merge-and-iterate?13:28
patchbotpatch 510978 - swift (feature/deep) - Root driven shrinking13:28
*** tesseract has quit IRC13:51
*** aluria has quit IRC13:55
*** aluria has joined #openstack-swift13:55
*** links has quit IRC13:58
*** links has joined #openstack-swift14:00
*** JimCheung has joined #openstack-swift14:00
*** shaner has quit IRC14:02
*** JimCheung has quit IRC14:04
*** links has quit IRC14:05
*** openstackstatus has quit IRC14:36
*** openstack has quit IRC14:39
*** openstack has joined #openstack-swift14:41
*** ChanServ sets mode: +o openstack14:41
*** openstack has quit IRC14:43
*** openstack has joined #openstack-swift14:46
*** ChanServ sets mode: +o openstack14:46
*** armaan has quit IRC15:01
*** armaan has joined #openstack-swift15:02
*** ianychoi has quit IRC15:04
*** SkyRocknRoll has quit IRC15:05
*** bkopilov has joined #openstack-swift15:12
*** links has joined #openstack-swift15:19
*** SkyRocknRoll has joined #openstack-swift15:22
*** geaaru has quit IRC15:28
*** dosaboy has quit IRC15:28
*** dosaboy has joined #openstack-swift15:28
*** Sukhdev has joined #openstack-swift16:00
*** ianychoi has joined #openstack-swift16:06
*** sc has quit IRC16:08
*** silor has quit IRC16:27
*** Sukhdev has quit IRC16:40
notmynamegood morning16:41
notmynameplease pay attention to this ML thread: http://lists.openstack.org/pipermail/openstack-dev/2017-December/125473.html16:44
*** armaan has quit IRC16:57
*** armaan has joined #openstack-swift16:58
openstackgerritChristopher Bartz proposed openstack/swift master: DLO Delete of whole object  https://review.openstack.org/52774417:01
*** geaaru has joined #openstack-swift17:06
*** Sukhdev has joined #openstack-swift17:06
*** Sukhdev has quit IRC17:08
*** cbartz has quit IRC17:17
*** hseipp has quit IRC17:21
openstackgerritClay Gerrard proposed openstack/swift master: functest for symlink + versioned writes  https://review.openstack.org/52758317:23
openstackgerritClay Gerrard proposed openstack/swift master: Assert X-Newest and X-Backend headers are propagated to symlink target  https://review.openstack.org/52747717:27
openstackgerritClay Gerrard proposed openstack/swift master: Symlink doc clean up  https://review.openstack.org/52747517:28
*** JimCheung has joined #openstack-swift17:31
*** mvk has quit IRC17:34
*** links has quit IRC17:40
*** shaner has joined #openstack-swift17:45
timburkelooking at http://lists.openstack.org/pipermail/openstack-dev/2017-December/125495.html -- i'm not sure what we have could be called a "cycle"17:48
*** chsc has joined #openstack-swift17:49
*** chsc has joined #openstack-swift17:49
timburkehttps://github.com/openstack/swift/commits/master/CHANGELOG shows between one and 5 (!) months between releases since 2.3.0; that seems better described as an "arrhythmia" :P17:50
*** SkyRocknRoll has quit IRC17:57
notmynametimburke: you can see the exact dates on https://wiki.openstack.org/wiki/Swift/version_map and visually as the blue vertical lines on http://d.not.mn/total_contribs.png18:11
*** mvk has joined #openstack-swift18:18
*** catintheroof has joined #openstack-swift18:44
*** openstackstatus has quit IRC19:27
*** Sukhdev has joined #openstack-swift19:28
*** oshritf has joined #openstack-swift19:45
*** oshritf has quit IRC19:46
timburkekota_: (when you wake up) what do you think about landing https://review.openstack.org/#/c/511964/ ? i know we want to freeze swift3 in favor of the s3api work, but (1) it only really touches infrastructure that i'm pretty sure we don't need on the s3api branch and (2) it obviates the need for https://review.openstack.org/#/c/522251/ (which seems to be causing -infra trouble?)19:47
patchbotpatch 511964 - swift3 - Use Pike Keystone19:47
patchbotpatch 522251 - swift3 - Update tox_install.sh to align for sphinx jobs19:47
timburkeswift3 is in such a terrible split-brain wrt to testing... we've got pike swift, newton constraints, and *mitaka* keystone :-(19:49
*** oshritf has joined #openstack-swift19:50
*** oshritf has quit IRC19:52
*** shaner has quit IRC19:58
*** shaner has joined #openstack-swift20:07
claygtorlololo20:08
claygit's just like a real world deployment!20:08
mattoliveraumorning20:24
mattoliverauacoles: re root driven shrinking, I say merge it. We have the old version in got history if we need to refer to it before merge :) and it is a great idea!20:25
mattoliverau*git history20:25
*** openstackstatus has joined #openstack-swift20:28
*** ChanServ sets mode: +v openstackstatus20:28
*** torgomatic has quit IRC20:43
*** charz has quit IRC20:43
*** saltsa has quit IRC20:43
*** saltsa has joined #openstack-swift20:43
*** torgomatic has joined #openstack-swift20:43
*** ChanServ sets mode: +v torgomatic20:43
*** charz has joined #openstack-swift20:43
*** Anticime1 is now known as Anticimex20:48
kota_good morning20:58
timburkegood morning kota_ (and mattoliverau! i've not been paying enough attention)20:58
kota_timburke: o/20:59
kota_timburke: thx for head up me on the swift3 change.20:59
notmynameMEETING TIME21:00
timburkesure -- just a thing to think over. seems like it might help out -infra without causing any pain for us21:00
kota_timburke: the latter one was in my radar and i think we can get it merged even we're in soft freeze because the tox install won't be affected to the s3api migration because it will be removed in swift upstream repo isn't it?21:01
acolesmattoliverau: ok, thanks - timburke has left a comment so I'll check that tomorrow and merge if appropriate21:01
kota_oh, it looks like we need previous one to pass the gate though. I'll check it in this week. Probably at Fri in my time.21:03
timburkeand if we have the first one, the second one won't be needed at all21:06
*** m_kazuhiro has joined #openstack-swift21:11
*** catintheroof has quit IRC21:15
*** catintheroof has joined #openstack-swift21:15
*** threestrands has joined #openstack-swift21:18
*** threestrands has quit IRC21:18
*** threestrands has joined #openstack-swift21:18
*** catintheroof has quit IRC21:19
acoleskota_: m_kazuhiro just running functional tests on my saio then I will push rebased patch21:35
*** saint_ has joined #openstack-swift21:35
timburkeoh man, we should go look at p 435929 and p 503262 too!21:36
patchbothttps://review.openstack.org/#/c/435929/ - swift - Functionnal tests for domain_remap middleware21:36
patchbothttps://review.openstack.org/#/c/503262/ - swift - Functional tests for staticweb middleware21:36
kota_acoles: great21:36
m_kazuhiroacoles: OK. Thank you!21:36
acolestimburke: oh yeah, domain-remap func tests, I should take a look21:39
openstackgerritAlistair Coles proposed openstack/swift master: Symlink implementation.  https://review.openstack.org/23216221:41
acolesm_kazuhiro: kota_ clayg ^^ someone should run probe tests?21:41
claygalways a good idea21:41
mattoliverautimburke: tempurl patch looks interesting, nice one21:42
acolesI kicked them off on my saio but not sure how long they take these days, getting late here21:42
kota_nice21:42
notmynameacoles: about 1200 seconds for me yesterday21:43
m_kazuhiroacoles: ok21:44
acolesnotmyname: now you have me feeling competitive :)21:45
* acoles urges on his testrunner21:45
kota_unit/func all passed in my local w/ the newest symlink21:57
acoleslol test.probe.test_container_sync failed - guess why?21:58
*** rcernin has joined #openstack-swift22:00
kota_sounds curious22:07
kota_ah... let me leave for now to prepare breakfast to my fam...22:07
acolesno symlink in container internal client conf! :O22:07
kota_acoles: oh, that path include new TestContainerSyncAndSymlink. I just thought another one failed.22:08
kota_make sense.22:08
acolesyep, also I had encryption enabled so all the container sync tests failed. I was using wrong saio vm.22:09
acolesits ok now .. test_sync_cross_acc_symlink (test.probe.test_container_sync.TestContainerSyncAndSymlink) ... ok22:10
kota_great22:10
*** m_kazuhiro has quit IRC22:13
*** saint_ has quit IRC22:22
*** chsc has quit IRC22:57
*** vint_bra has joined #openstack-swift23:06
*** kei_yama has joined #openstack-swift23:34
*** itlinux has joined #openstack-swift23:35

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