Tuesday, 2016-01-19

*** ho has joined #openstack-swift00:01
*** ho is now known as Guest9151900:02
*** kei_yama has joined #openstack-swift00:04
*** nadeem has quit IRC00:06
*** Jeffrey4l has joined #openstack-swift00:07
*** Guest91519 is now known as ho_00:10
*** diazjf has quit IRC00:10
ho_good morning00:12
*** dmorita has quit IRC00:13
mattoliverauho_: morning00:15
*** diogogmt has quit IRC00:15
ho_mattoliverau: morning!00:16
*** dslevin_ has joined #openstack-swift00:17
*** dslevin_ has quit IRC00:17
*** nadeem has joined #openstack-swift00:18
*** dslevin has quit IRC00:18
*** dslevin has joined #openstack-swift00:18
*** asettle has quit IRC00:19
*** asettle has joined #openstack-swift00:19
*** dmorita has joined #openstack-swift00:21
*** haomaiwang has joined #openstack-swift00:22
*** dmorita has quit IRC00:22
*** asettle has quit IRC00:24
*** haomaiwang has quit IRC00:27
*** dmorita has joined #openstack-swift00:30
ho_hugokuo: morning!00:31
*** dmorita has quit IRC00:31
*** dmorita has joined #openstack-swift00:32
*** dslevin has quit IRC00:35
*** dmorita has quit IRC00:35
*** dmorita has joined #openstack-swift00:35
*** dslevin has joined #openstack-swift00:38
mattoliverauhugokuo: morning00:39
kota_ho_, mattolivearu, hugokuo, and all: good morning00:39
ho_kota_: morning. how about snowing tokyo today?00:40
kota_ho_: it's fine today, snowing just yesterday early morning00:42
kota_ho_: however snow is still remaining everywhere00:42
kota_ho_: most traffic (basically car, bus and so on) still affected a bit.00:43
ho_kota_: i see. please be careful there are many troubles in a few day after big snow :-)00:45
kota_ho_: thanks!00:45
*** dslevin has quit IRC00:48
*** lcurtis_ has joined #openstack-swift00:48
openstackgerritTakashi Kajinami proposed openstack/swift: Update parameters about authtoken middleware in proxy-server.conf.sample  https://review.openstack.org/26139500:49
*** dslevin has joined #openstack-swift00:51
*** m_kazuhiro has joined #openstack-swift00:54
*** diazjf has joined #openstack-swift00:54
*** lcurtis_ has quit IRC00:55
*** lcurtis_ has joined #openstack-swift00:58
*** diogogmt has joined #openstack-swift00:58
*** asettle has joined #openstack-swift01:00
*** Jeffrey4l has quit IRC01:01
ho_got many merge conflicts01:01
notmynameho_: yeah. I'm not sure why. I saw the same thing this morning01:02
*** diazjf has quit IRC01:03
ho_notmyname: is there any big patch landed?01:03
notmynameno. that's the weird part01:03
ho_notmyname: ok, i will rebase them.01:04
mattoliveraunotmyname: how's your LCA talk coming? not long now01:05
notmynameheh. don't remind me!01:05
notmynameI've got notes. and a rough outline. lots to fill in, though ;-)01:05
openstackgerritHisashi Osanai proposed openstack/swift: Improve Keystone v3 token support  https://review.openstack.org/20146101:05
mattoliveraunotmyname: at least I'm not the only one in the same boat :)01:06
notmynameclayg: I'm now trying to come up with a changelog entry for patch 241571. "rings do more of what you really thought they did" doesn't quite have the right sound to it. any help is appreciated01:06
patchbotnotmyname: https://review.openstack.org/#/c/241571/ - Put part-replicas where they go (MERGED)01:06
mattoliverautrying to force myself to do a little bit each day01:06
notmynamemattoliverau: that's a good plan01:06
notmynamemattoliverau: oh, will we be able to pool back to melb on saturday? or should I take the lca bus to the airport?01:07
mattoliveraunotmyname: depends on when your flight is I guess.01:07
openstackgerritTim Burke proposed openstack/python-swiftclient: Include response headers in ClientExceptions  https://review.openstack.org/26938201:07
notmynamemattoliverau: oh my. I just looked01:08
mattoliverauThe bus will take you right to the airport, where as the train will take you to a station which you then need to bus/or taxi to the airport01:08
openstackgerritHisashi Osanai proposed openstack/swift: Add functional test for access control (RBAC) with Keystone  https://review.openstack.org/20241101:08
mattoliverauif its not till later that night, then you can come back to Melb and I could drive you to the airport.01:08
openstackgerritHisashi Osanai proposed openstack/swift: Add functional test for access control (container ACL) with Keystone  https://review.openstack.org/21360801:08
notmynamemattoliverau: nah. it's super early in the morning01:09
openstackgerritHisashi Osanai proposed openstack/swift: Add functional test for access control (RBAC INFO) with Keystone  https://review.openstack.org/21281001:09
openstackgerritHisashi Osanai proposed openstack/swift: Add functional test for access control (COPY) with Keystone  https://review.openstack.org/21282501:09
openstackgerritHisashi Osanai proposed openstack/swift: Performance improvement for RBAC related functional tests  https://review.openstack.org/25337101:09
* notmyname waits for ho_ to finish rebasing ;-)01:09
openstackgerritHisashi Osanai proposed openstack/swift: Add container quota functional test  https://review.openstack.org/25515201:09
openstackgerritHisashi Osanai proposed openstack/swift: Add X-Delete-After and X-Delete-At functional test  https://review.openstack.org/25562601:09
mattoliveraunotmyname: if I was driving to LCA then I'd just drive you to the airport, but not going to leave Bec without the car in summer and that pregnant.01:10
notmynameah, right01:10
ho_notmyname: mattoliverau: sorry for waiting my rebasing. i just done01:11
notmynamemy flight home is at 7am. so I'll be leaving geelong in the middle of the night01:11
notmynameho_: no worries :-)01:11
mattoliveraunotmyname: bugger. hmm, we should look into finding out how to get you to the airport. Worst case Taxi.01:12
notmynamehow's transport in general between MEL and geelong? are we talking well-established transit infrastructure? or dirt road in the desert?01:13
mattoliveraunotmyname: lol, yeah, no dirt road, there are train links and a highway :)01:16
notmynamepeterlisak: I'm going through all the stuff that's landed since the last release. you've got a *lot* of great commits into swift. thanks!01:20
notmynameoh, for the other-side-of-the-world shift that's coming online now, likely to be a swift release in 2 days (my wednesday)01:21
*** haomaiwa_ has joined #openstack-swift01:23
*** nadeem has quit IRC01:27
*** haomaiwa_ has quit IRC01:28
*** nadeem has joined #openstack-swift01:30
*** dslevin has quit IRC01:31
*** dslevin has joined #openstack-swift01:36
kota_notmyname: do we have remaining patches (likely blockers) for the next swift release?01:37
notmynamekota_: blockers? no. stuff that would be nice? probably01:37
*** nadeem has quit IRC01:37
kota_ok, if you think of need some reviewers, please ping me :)01:37
notmynamekota_: the RBAC patch from ho_!01:38
kota_so huge (not open the review page yet :P)01:38
ho_notmyname: kota_: yeah! that absolutely my blocker! lol01:38
notmynamekota_: yeah, but it's mostly repeated stuff that shouldn't be as hard to review as the length of the patch would imply01:39
kota_notmyname: all right, will try during this week.01:40
notmynamethanks :-)01:40
ho_kota_: notmyname: thanks!01:40
notmynameI've been putting together all the responses to my questions I sent out01:40
notmynametons of great information and feeback01:41
*** dmorita has quit IRC01:41
openstackgerritKota Tsuyuzaki proposed openstack/swift: WIP: Fix EC ring validation at ring reload  https://review.openstack.org/26899301:43
*** dmorita has joined #openstack-swift01:45
*** sc68cal has quit IRC01:49
*** sc68cal has joined #openstack-swift01:54
*** Jeffrey4l has joined #openstack-swift01:56
*** dslevin has quit IRC01:56
*** diogogmt has quit IRC02:00
*** dmorita has quit IRC02:03
*** diogogmt has joined #openstack-swift02:06
*** dslevin has joined #openstack-swift02:07
*** asettle has quit IRC02:08
*** sc68cal has quit IRC02:09
*** sc68cal has joined #openstack-swift02:12
*** klrmn has quit IRC02:14
*** david-lyle has joined #openstack-swift02:22
*** mingdang1 has joined #openstack-swift02:24
*** haomaiwa_ has joined #openstack-swift02:26
*** dmorita has joined #openstack-swift02:27
openstackgerritTim Burke proposed openstack/python-swiftclient: Get rid of getexpect cruft  https://review.openstack.org/26939902:36
*** dmorita has quit IRC02:37
*** dmorita has joined #openstack-swift02:39
*** dmorita has quit IRC02:57
*** dmorita has joined #openstack-swift03:00
openstackgerritTim Burke proposed openstack/python-swiftclient: Include response headers in ClientExceptions  https://review.openstack.org/26938203:01
*** dmorita has quit IRC03:02
*** km_ has quit IRC03:08
*** klrmn has joined #openstack-swift03:08
*** km_ has joined #openstack-swift03:08
*** forcebanana has quit IRC03:12
*** sanchitmalhotra has joined #openstack-swift03:13
*** natarej has joined #openstack-swift03:15
*** dmorita has joined #openstack-swift03:16
*** yatin has joined #openstack-swift03:19
*** links has joined #openstack-swift03:27
openstackgerritJanie Richling proposed openstack/swift: Support for http footers - Replication and EC  https://review.openstack.org/23739303:27
*** venkat has joined #openstack-swift03:28
jrichlijlhinson ^^ ready for rebasing if-match03:28
*** mragupat has joined #openstack-swift03:37
*** asettle has joined #openstack-swift03:49
*** dmorita has quit IRC03:59
*** dmorita has joined #openstack-swift04:26
*** silor has joined #openstack-swift04:26
*** proteusguy has quit IRC04:40
*** lcurtis_ has quit IRC04:42
*** lcurtis_ has joined #openstack-swift04:44
*** dmorita has quit IRC04:45
*** dmorita has joined #openstack-swift04:49
*** proteusguy has joined #openstack-swift04:53
*** dmorita has quit IRC04:59
*** haomaiwa_ has quit IRC04:59
*** dmorita has joined #openstack-swift05:02
*** haomaiwang has joined #openstack-swift05:05
*** mragupat has quit IRC05:07
*** asettle has quit IRC05:11
*** ppai has joined #openstack-swift05:12
*** dmorita has quit IRC05:14
*** dmorita has joined #openstack-swift05:15
openstackgerritTim Burke proposed openstack/python-swiftclient: Get rid of FakeConn cruft  https://review.openstack.org/26939905:16
*** lcurtis_ has quit IRC05:20
*** Jeffrey4l has quit IRC05:31
*** SkyRocknRoll has joined #openstack-swift05:33
*** klrmn1 has joined #openstack-swift05:41
*** klrmn has quit IRC05:42
*** peluse has quit IRC05:46
*** peluse has joined #openstack-swift05:47
*** dmorita has quit IRC05:48
*** Jeffrey4l has joined #openstack-swift05:52
*** klrmn1 has quit IRC05:55
*** jamielennox is now known as jamielennox|away05:59
*** dimasot has joined #openstack-swift06:06
*** haomaiwang has quit IRC06:21
*** 6A4ABP35J has joined #openstack-swift06:24
openstackgerritBing Hu proposed openstack/python-swiftclient: Support uploading to an object in swift from stdin update code and add unittest use mkstemp instead of NamedTemporaryFile, upadte unittest Closes-bug: #1521342  https://review.openstack.org/25941006:40
openstackbug 1521342 in python-swiftclient "support uploading from a stdin stream" [Wishlist,In progress] https://launchpad.net/bugs/1521342 - Assigned to Bing Hu (hubingsh)06:40
*** eranrom has joined #openstack-swift06:41
*** ChubYann has quit IRC06:41
eranrommattoliverau: Hey matt, are you here?06:48
*** dmorita has joined #openstack-swift06:49
eranrommattoliverau: I basically wanted to nag you (politely) regarding the container sync patch (https://review.openstack.org/#/c/205803/)06:50
eranrommattoliverau: many thanks in advance06:50
*** Jeffrey4l has quit IRC06:52
*** dmorita has quit IRC06:53
*** zaitcev has quit IRC06:55
*** asettle has joined #openstack-swift06:56
*** 6A4ABP35J has quit IRC07:09
*** haomaiwang has joined #openstack-swift07:09
openstackgerritMustafa ÇELİK proposed openstack/swift: X-Timestamp missing from response headers  https://review.openstack.org/26816307:14
*** eranrom has quit IRC07:26
openstackgerritKota Tsuyuzaki proposed openstack/swift: WIP: Fix EC ring validation at ring reload  https://review.openstack.org/26899307:30
dimasotgood morning07:31
*** haomaiwang has quit IRC07:34
*** haomaiwang has joined #openstack-swift07:36
mattoliveraudimasot: morning07:39
*** Jeffrey4l has joined #openstack-swift07:42
*** jmccarthy has quit IRC07:53
*** jmccarthy has joined #openstack-swift07:53
*** Jeffrey4l has quit IRC07:56
*** arnox has joined #openstack-swift07:59
*** rcernin has joined #openstack-swift08:01
*** asettle has quit IRC08:14
*** Jeffrey4l has joined #openstack-swift08:14
*** rledisez has joined #openstack-swift08:18
*** wanghua has joined #openstack-swift08:19
*** eranrom has joined #openstack-swift08:19
*** chlong has quit IRC08:27
*** geaaru has joined #openstack-swift08:31
*** asettle has joined #openstack-swift08:32
*** dimasot has quit IRC08:38
*** daemontool has joined #openstack-swift08:41
*** km_ has quit IRC08:42
*** jordanP has joined #openstack-swift08:48
*** dmorita has joined #openstack-swift08:51
*** dmorita has quit IRC08:57
*** acoles_ is now known as acoles08:58
acolesgood morning09:02
openstackgerritMustafa ÇELİK proposed openstack/swift: X-Timestamp missing from response headers  https://review.openstack.org/26816309:02
*** daemontool has quit IRC09:02
mahaticgood morning acoles09:03
openstackgerritChristopher Bartz proposed openstack/swift: Use the correct split_path in handle_request  https://review.openstack.org/26000109:03
*** daemontool has joined #openstack-swift09:03
*** mrmoje has joined #openstack-swift09:03
acolesmahatic: so do we have consensus on removing all the content-type code? looks like jrichli is happy with that.09:05
*** jordanP has quit IRC09:08
*** [1]eranrom has joined #openstack-swift09:12
*** eranrom has quit IRC09:15
*** [1]eranrom is now known as eranrom09:15
*** mrmoje has quit IRC09:17
*** mrmoje has joined #openstack-swift09:19
*** asettle has quit IRC09:20
*** jistr has joined #openstack-swift09:20
*** nakagawamsa has joined #openstack-swift09:27
*** nakagawamsa has quit IRC09:27
*** trifon has joined #openstack-swift09:35
*** aix has joined #openstack-swift09:39
*** daemontool has quit IRC09:39
*** daemontool has joined #openstack-swift09:39
*** haomaiwang has quit IRC09:40
mahaticacoles: yup :)09:42
*** haomaiwang has joined #openstack-swift09:43
*** aix_ has joined #openstack-swift09:44
*** hseipp has joined #openstack-swift09:44
acolesmahatic: great09:45
*** aix_ has quit IRC09:46
*** number80 has quit IRC09:46
*** mrmoje has quit IRC09:49
*** mrmoje has joined #openstack-swift09:51
*** geaaru has quit IRC09:54
*** bartolo86 has joined #openstack-swift09:57
*** cebruns has quit IRC09:58
*** Jeffrey4l has quit IRC10:00
*** cebruns has joined #openstack-swift10:01
*** geaaru has joined #openstack-swift10:06
*** daemontool has quit IRC10:08
*** daemontool has joined #openstack-swift10:08
*** dimasot has joined #openstack-swift10:10
*** daemontool has quit IRC10:11
*** daemontool has joined #openstack-swift10:11
*** jordanP has joined #openstack-swift10:12
*** eranrom has quit IRC10:14
*** daemontool has quit IRC10:17
*** daemontool has joined #openstack-swift10:17
*** Jeffrey4l has joined #openstack-swift10:18
*** daemontool has quit IRC10:19
*** daemontool has joined #openstack-swift10:19
*** hyakuhei has joined #openstack-swift10:20
*** dimasot has quit IRC10:21
*** silor has quit IRC10:23
*** aix has quit IRC10:23
*** aix has joined #openstack-swift10:26
*** dimasot has joined #openstack-swift10:30
*** Jeffrey4l has quit IRC10:31
*** number80 has joined #openstack-swift10:35
*** cebruns has quit IRC10:37
*** daemontool has quit IRC10:38
openstackgerritKota Tsuyuzaki proposed openstack/swift: Fix EC ring validation at ring reload  https://review.openstack.org/26899310:38
*** cebruns has joined #openstack-swift10:39
*** daemontool has joined #openstack-swift10:42
*** mingdang1 has quit IRC10:46
*** aix has quit IRC10:46
*** mrmoje has quit IRC10:49
*** dmorita has joined #openstack-swift10:54
*** aix has joined #openstack-swift10:56
*** dmorita has quit IRC10:59
openstackgerritMahati Chamarthy proposed openstack/swift: Decrypt multipart mime doc  https://review.openstack.org/24837711:08
openstackgerritMahati Chamarthy proposed openstack/swift: Remove content-type encryption/decryption  https://review.openstack.org/26129211:08
*** ho_ has quit IRC11:09
nttHi, how can I obtain the size of a pseudo folder without listing files each time?11:16
*** eranrom has joined #openstack-swift11:20
mahaticacoles: jrichli ^ there we go11:27
*** eranrom has quit IRC11:30
acolesmahatic: thanks11:31
acolesnotmyname: added you as reviewer on patch 261395 to seek your opinion on maintaining options vs linking to docs11:31
patchbotacoles: https://review.openstack.org/#/c/261395/ - Update parameters about authtoken middleware in pr...11:31
acolesnotmyname: by the time you read this I am sure I will feel less agitated about it :)11:32
*** sanchitmalhotra has quit IRC11:33
*** kei_yama has quit IRC11:35
*** silor has joined #openstack-swift11:41
*** silor1 has joined #openstack-swift11:44
*** jmccarthy1 has joined #openstack-swift11:46
*** jmccarthy has quit IRC11:46
*** SkyRocknRoll has quit IRC11:47
*** silor has quit IRC11:48
*** silor1 is now known as silor11:48
acolesmahatic: you still around?11:50
mahaticacoles: yeah11:50
acolesmahatic: about the two func tests you mention as failing - i see them fail on v1 of the patch and on feature/crypto11:51
acolesmahatic: did you think they were new errors?11:51
acoleswith version 6?11:51
mahaticacoles: no, I noticed them with v1 itself ( i mentioned it in the comment - that it is on patchset 1 itself)11:52
mahaticacoles: I couldn't pin point what was causing them though11:52
acolesmahatic: OIC I misunderstood your comment11:52
acolesmahatic: but I see them on feature/crypto too so not due to this patch? do you see different on feature/crypto?11:53
mahaticacoles: oops, I didn't check on feature/crypto though11:53
mahaticacoles: thanks, let me do that. Maybe they are intermittent failures(?)11:54
*** yatin has quit IRC11:55
*** proteusguy has quit IRC11:56
*** m_kazuhiro has quit IRC12:00
mahaticacoles: I don't see them on feature/crypto12:01
acolesmahatic: ooohhh12:01
mahaticacoles: ERROR: test_slo_get_ranged_manifest_repeated_segment ->this one12:02
*** bartolo86 has left #openstack-swift12:02
mahaticacoles: there is another one ERROR: test_slo_get_ranged_submanifest -> this one I do see12:02
*** SkyRocknRoll has joined #openstack-swift12:03
acolesmahatic: is you feature/crypto branch up to date?12:03
acoleslast commit is 688aa7d12:03
*** daemontool has quit IRC12:05
mahaticacoles: well that's not the last commit :/12:05
*** daemontool has joined #openstack-swift12:06
*** aix has quit IRC12:06
mahaticacoles: If it helps, I didn't see these on the parent commit as well i.e "Modify unit tests to include real crypto"12:06
*** dimasot has quit IRC12:07
mahaticbut that commit itself wasn't upto date with feature/crypto I think12:08
acolesmahatic: what do you see returned from 'git log --pretty=oneline -n 1'12:08
mahaticacoles: give me a min please12:09
mahaticacoles: a6a8bb - add round trip encrypter12:13
acolesmahatic: ok your branch is a little behind12:14
acolesso on feature/crypto branch you need to 'git pull --ff-only origin'12:14
acolesmahatic: although, its not obvious why that would make difference to the test error12:15
mahaticacoles: done12:15
mahaticacoles: yeah12:15
acolesbut the nothing is obvious round here :)12:15
acolesthen* !12:15
mahaticacoles: I will run them again12:15
acolesdon't forget to restart services12:16
*** mrmoje has joined #openstack-swift12:17
*** aix has joined #openstack-swift12:18
mahaticacoles: okay, I see them on feature/crpyto as well. Failures are consistent. Happy time? ;)12:24
acolesmahatic: yay! consistency is good12:24
acoles(consistent success is better)12:25
acolesmahatic: approved. thanks.12:28
mahaticacoles: cool, thanks as well12:29
*** proteusguy has joined #openstack-swift12:33
openstackgerritAlistair Coles proposed openstack/swift: EC: Avoid conflicts when ssync'ing fragments  https://review.openstack.org/18140712:35
*** eranrom has joined #openstack-swift12:36
openstackgerritAlistair Coles proposed openstack/swift: Don't ssync data when only a durable is missing  https://review.openstack.org/23268412:41
openstackgerritAlistair Coles proposed openstack/swift: Stop staticweb revealing container existence to unauth'd requests  https://review.openstack.org/24886712:42
openstackgerritAlistair Coles proposed openstack/swift: Enable object server to return non-durable data  https://review.openstack.org/21527612:43
*** cbartz has joined #openstack-swift12:46
*** links has quit IRC12:46
*** haomaiwang has quit IRC12:51
*** dmorita has joined #openstack-swift12:56
*** dmorita has quit IRC13:00
*** daemontool_ has joined #openstack-swift13:10
*** daemontool has quit IRC13:11
*** alejandrito has joined #openstack-swift13:14
openstackgerritAlistair Coles proposed openstack/swift: Make functional tests tolerate pre-existing versioned container  https://review.openstack.org/16338513:15
*** janonymous has quit IRC13:15
*** cbartz has left #openstack-swift13:16
*** cbartz has joined #openstack-swift13:17
*** ekarlso has quit IRC13:18
*** ekarlso has joined #openstack-swift13:18
acolesclayg: stole your idea ^^ hope that's ok!13:18
*** SkyRocknRoll has quit IRC13:21
openstackgerritAlistair Coles proposed openstack/swift: Make pipeline_property less brittle  https://review.openstack.org/10492613:21
* acoles is done with rebase13:24
*** shakamunyi has joined #openstack-swift13:24
*** Kanshi has joined #openstack-swift13:28
KanshiHello everybody, I'm looking for someone who uses Swift with a Nginx reverse proxy in front13:29
*** mingdang1 has joined #openstack-swift13:31
*** janonymous has joined #openstack-swift13:46
*** daemontool__ has joined #openstack-swift13:49
*** daemontool_ has quit IRC13:50
*** haomaiwang has joined #openstack-swift13:51
*** daemontool__ is now known as daemontool13:54
*** Kanshi has quit IRC14:00
*** daemontool has quit IRC14:06
*** daemontool has joined #openstack-swift14:07
*** venkat has quit IRC14:12
*** petertr7_away is now known as petertr714:19
*** diogogmt has quit IRC14:28
*** ppai has quit IRC14:29
*** daemontool_ has joined #openstack-swift14:30
*** daemontool has quit IRC14:31
*** daemontool_ has quit IRC14:34
*** daemontool_ has joined #openstack-swift14:34
*** daemontool_ has quit IRC14:37
*** daemontool_ has joined #openstack-swift14:37
*** mingdang1 has quit IRC14:51
*** diogogmt has joined #openstack-swift14:54
*** dmorita has joined #openstack-swift14:58
*** dmorita has quit IRC15:02
*** mragupat has joined #openstack-swift15:04
*** blmartin has joined #openstack-swift15:06
*** diogogmt has quit IRC15:12
*** daemontool_ has quit IRC15:12
*** dslevin has quit IRC15:18
*** daemontool_ has joined #openstack-swift15:24
*** bsdkurt has quit IRC15:31
*** cbartz has left #openstack-swift15:37
*** bsdkurt has joined #openstack-swift15:39
openstackgerritGábor Antal proposed openstack/swift: Added missing super() calls in obj/ classes  https://review.openstack.org/26971915:49
*** hyakuhei has quit IRC15:56
*** mrmoje has quit IRC15:57
*** dmorita has joined #openstack-swift15:59
*** diazjf has joined #openstack-swift16:01
*** diogogmt has joined #openstack-swift16:04
*** dmorita has quit IRC16:04
*** hyakuhei has joined #openstack-swift16:05
*** aix has quit IRC16:05
*** trifon has quit IRC16:19
*** jordanP has quit IRC16:22
*** petertr7 is now known as petertr7_away16:23
*** sanchitmalhotra has joined #openstack-swift16:26
*** Jeffrey4l has joined #openstack-swift16:27
*** Jeffrey4l has quit IRC16:28
*** Jeffrey4l has joined #openstack-swift16:29
*** klamath has joined #openstack-swift16:35
klamathhowdy, getting this error when trying to reduce weight on some devices to zero: Balance of 999.99 indicates you should push this       ring, wait at least 1 hours, and rebalance/repush. after one hour i try another rebalance and get : Cowardly refusing to save rebalance as it did not change at least 1%.16:35
*** alejandrito has quit IRC16:41
*** eranrom has quit IRC16:45
*** jmccarthy1 has quit IRC16:45
*** nexusz99 has joined #openstack-swift16:46
*** nexusz99 has quit IRC16:46
*** nadeem has joined #openstack-swift16:49
*** gyee has joined #openstack-swift16:53
*** rcernin has quit IRC16:53
*** david-lyle has quit IRC16:58
*** dslevin has joined #openstack-swift16:59
*** dmorita has joined #openstack-swift17:03
*** diazjf has quit IRC17:03
*** david-lyle has joined #openstack-swift17:06
*** acoles is now known as acoles_17:07
*** rcernin has joined #openstack-swift17:08
openstackgerritBéla Vancsics proposed openstack/swift: Reduce code duplication  https://review.openstack.org/25426617:14
*** diazjf has joined #openstack-swift17:16
*** dmorita has quit IRC17:19
notmynamegood morning, everyone17:21
*** dmorita has joined #openstack-swift17:21
*** Jeffrey4l has quit IRC17:22
*** diogogmt has quit IRC17:25
*** dmorita has quit IRC17:25
*** dmorita has joined #openstack-swift17:25
notmynameif you missed the email, next 2 openstack releases are "newton" and "ocata"17:26
openstackgerritBéla Vancsics proposed openstack/swift: Reduce code duplication  https://review.openstack.org/25426617:27
notmynamedoes anyone have python 2.7.3?17:28
*** petertr7_away is now known as petertr717:30
*** arnox has quit IRC17:35
pdardeaunotmyname: great, now i've got "it's the big fig newton!" running in endless loop in my head :D17:36
*** diogogmt has joined #openstack-swift17:36
notmynamepdardeau: I think you might have a good tshirt design idea there17:36
*** dslevin has quit IRC17:37
*** ChubYann has joined #openstack-swift17:45
*** jistr has quit IRC17:48
*** rledisez has quit IRC18:00
*** geaaru has quit IRC18:02
*** diogogmt has quit IRC18:04
*** diogogmt has joined #openstack-swift18:05
notmynameFYI I've got a patch I'm working on that fixes a release-blocking bug (introduced with the IPv6 memcache patch)18:10
*** klrmn has joined #openstack-swift18:13
*** hseipp has quit IRC18:13
*** dimasot has joined #openstack-swift18:19
notmynamehere's the summary: we need to bump the eventlet version18:25
*** lcurtis has joined #openstack-swift18:26
notmynamebasically, the min version in our requirements doesn't support IPv6. current swift requirements is >=0.16.1. current global requirements is >=0.17.418:26
notmyname0.17.4 is also the current release you get with `pip install -U eventlet`18:27
notmynamerunning tests..will submit a patch shortly18:29
notmyname@acoles_ @dfg @redbo_ @kota_ @tdasilva @cschwede ^18:30
notmynameeg https://gist.github.com/notmyname/23c5f6a656ea0f2a5a93 and https://gist.github.com/notmyname/8a4d548389ca7f15a3ab18:33
*** zhiyan has quit IRC18:33
*** siva_krishnan has quit IRC18:34
*** aerwin3 has quit IRC18:34
*** aerwin3 has joined #openstack-swift18:36
*** zhiyan has joined #openstack-swift18:36
diazjfnotmyname, jrichli, whenever you get a chance please review https://review.openstack.org/#/c/241068/18:38
diazjf^ completed Castellan auth support spec18:39
diazjfnotmyname, jrichli, I included examples of what can be added to the swift keymaster to be able to access different authentication types18:42
diazjfit will support password, token, and certificate cerdentials18:43
*** dmorita has quit IRC18:46
*** dmorita has joined #openstack-swift18:46
*** hyakuhei has quit IRC18:48
*** hyakuhei has joined #openstack-swift18:49
*** hyakuhei has quit IRC18:49
openstackgerritJason Gu proposed openstack/swift: Functional testing for Swift API response headers  https://review.openstack.org/26981018:49
openstackgerritJohn Dickinson proposed openstack/swift: Bump eventlet min version to 0.17.4  https://review.openstack.org/26981118:50
notmynametorgomatic clayg acoles_ dfg redbo_ kota_ tdasilva cschwede ^18:51
*** eranrom has joined #openstack-swift18:54
jrichlidiazjf: thanks, i will take a look18:55
*** petertr7 is now known as petertr7_away18:56
diazjfjrichli, awesome let me know what you think :)18:57
notmynamediazjf: note that swift doesn't currently use oslo.config19:00
diazjfnotmyname, it's just for backward compatibility with other projects19:01
diazjfnotmyname, swift in this case will use the castellan credential objects which are generated from reading the swift conf19:01
*** petertr7_away is now known as petertr719:08
notmynamemattoliverau: we need to get reverse listings on http://developer.openstack.org/api-ref-objectstorage-v1.html19:09
*** dmorita has quit IRC19:19
*** sgundur has joined #openstack-swift19:22
*** dmorita has joined #openstack-swift19:27
*** dmorita has quit IRC19:27
*** dmorita has joined #openstack-swift19:27
*** lpabon has joined #openstack-swift19:42
*** lpabon_ has joined #openstack-swift19:43
*** lpabon_ has quit IRC19:46
*** lpabon has quit IRC19:46
*** lpabon has joined #openstack-swift19:46
*** lpabon has quit IRC19:47
*** lpabon has joined #openstack-swift19:47
*** lpabon_ has joined #openstack-swift19:47
*** gyee has quit IRC19:49
*** eranrom has quit IRC19:50
*** lpabon_ has quit IRC19:51
*** petertr7 is now known as petertr7_away19:52
*** lpabon has quit IRC19:53
*** lpabon has joined #openstack-swift19:53
openstackgerritSivasathurappan Radhakrishnan proposed openstack/swift: Added unit test cases for common/bufferedhttp.py  https://review.openstack.org/24467819:54
*** petertr7_away is now known as petertr719:56
*** PsionTheory has joined #openstack-swift19:56
openstackgerritSivasathurappan Radhakrishnan proposed openstack/swift: Deleted comment about part power in FakeRing  https://review.openstack.org/25363120:01
ZyricGood Morning20:02
*** diazjf1 has joined #openstack-swift20:03
*** diazjf has quit IRC20:05
*** remix_tj has joined #openstack-swift20:13
*** eranrom has joined #openstack-swift20:13
*** [1]dimasot has joined #openstack-swift20:22
*** mrmoje has joined #openstack-swift20:24
*** dimasot has quit IRC20:25
*** [1]dimasot is now known as dimasot20:25
blmartinSo while writing some probe tests for container sharding, I found that swiftclient does not support arbitrary query strings and there are only a few supported keywords.20:30
blmartinWould there be any desire for arbitrary query strings or is there a reason for the restriction?20:30
*** diogogmt has quit IRC20:36
blmartinAh I need to amend my statement20:37
blmartinThere is not an arbitrary querystring for container operations20:38
notmynameblmartin: are you wanting to make a request that doesn't have first-class support in swiftclient? or are you doing new stuff?20:39
blmartinnew stuff20:39
blmartinit is for container sharding so I could see this not being normal/desired20:39
notmynameseems very similar to supporting arbitrary headers passed in. not somethat that should be normal. but pretty reasonable to ask for20:40
*** diogogmt has joined #openstack-swift20:40
timburkeworth noting: from blmartin's issue, it follows that there's no way to do reverse listings currently in swiftclient20:40
notmynameso in answer to your question, no there's no reason other than nobody has typed that in20:41
blmartinnotmyname: thanks, I'll see if I cannot find a similar issue in the blueprints. If I nothing turns up I can add one20:44
*** dimasot has quit IRC20:48
jrichlihttps://wiki.openstack.org/wiki/ObjectEncryption is very out of date.  I am not sure of the practice for wikis.  Do we mark this as deprecated or something?20:52
notmynamejrichli: there's no official practice. actually I'm not sure the wikis are technically authoritative (just informative). if it's wrong, delete it. or update. or add a reference to something that's right20:54
jrichlinotmyname: ok, thx20:55
*** daemontool_ has quit IRC20:59
*** silor has quit IRC21:07
*** dslevin has joined #openstack-swift21:09
*** eranrom has quit IRC21:14
*** gyee has joined #openstack-swift21:17
*** siva_krishnan has joined #openstack-swift21:18
*** siva_krishnan has joined #openstack-swift21:20
*** siva_krishnan has quit IRC21:20
diazjf1jrichli, so +1 on the Castellan spec?21:23
*** siva_krishnan has joined #openstack-swift21:24
*** mingdang1 has joined #openstack-swift21:29
jrichlidiazjf1: see last comment on review21:30
onovyclayg: hi, are you here?21:37
*** dslevin has quit IRC21:38
claygonovy: bangin'21:40
mattoliveraunotmyname: rgr I'll get on that (reverse listings in api doc)21:41
onovyclayg: https://docs.google.com/spreadsheets/d/1dzcMutNwRYcDiSMNOhKLM1tk8QX3uqFg9BJszJ6plr4/edit#gid=0 // tests results for ionice21:41
blmartinmattoliverau: I'm trying to get these basic sharding tests out the door, I should be able to play with the range code again tomorrow21:43
claygonovy: Can you describe the difference between scearnio B & C21:45
mattoliveraublmartin: awesome thanks man, I'm trying to get the sharders auditing code working, manly looking for overlaps in the root containers shard table. and then I may need to tweek misplaced objects a little more. But getting closer.. I think21:45
onovyclayg: ionice priority set21:45
onovyidle vs. default21:45
onovyon object-auditor21:45
blmartinmattoliverau: yeah last I looked at it there were some ranges missing in the table, a strange case with two ranges covering None on the right side21:46
blmartinmattoliverau: all the objects were in the sharded containers though21:47
onovyclayg: peter tested performance when auditor is not running (A). then after auditor start without any change (B)21:47
onovysry, C21:47
claygonovy: ok!  that's great!21:47
mattoliveraublmartin: great, so movement happened but the table syncing isn't quite right yet. that's a start ;)21:48
onovythen changed priority to idle of object-auditor : B21:48
claygwith deadline does the ionice priority controls not really make any difference?21:48
onovywhen you are using deadline sched in kernel, priorities are not working21:48
onovywe are comparing CFQ with deadline sched here21:48
claygonovy: understood21:48
claygonovy: sure that makes sense21:48
notmynamemattoliverau: or let asettle know ;-)21:49
onovythere are ~ 30 % perferomance boost for ionice idle21:49
claygonovy: so you can definately observe an impact even from a mile away - that's great - incredibly validating21:49
onovyand of course: slower auditor :)21:49
onovyBUT. only for GETs21:49
onovyfor which i don't have explaining (now)21:49
mattoliveraunotmyname: lol, hmm... she is on the other side of the country ;)21:50
onovyit's nice that you have performance almost same as when "no auditor running"21:50
notmynamemattoliverau: :-)21:51
claygonovy: please associate that spreadsheet with the change21:51
notmynamemattoliverau: no particular rush, just something I noticed when putting together changelog updates21:51
onovybtw: if you want reproduce it, there are few caveats:21:52
onovy1. don't use loopback device (no IO priority)21:52
onovy2. use CFQ (obvious)21:52
onovy3. "unshape" object-auditor to "max"21:52
onovy4. kernel buffers/cache are hell for testing :)21:53
onovy /loopback/loop/ (mount -o loop)21:53
onovyclayg: ok21:53
notmynameonovy: thanks for putting that together21:54
onovypeterlisak: ^^ this thanks is for you :)21:54
claygnotmyname: +1 thank you peterlisak !21:54
onovypeter worked on it hard on friday and monday21:54
onovyi'm just "selling" it :)21:55
onovythat first caveats point (loop) was first fail21:55
onovysaio, loop, hey - it doesn't do anything! :)21:55
*** fthiagogv has quit IRC21:56
notmynameonovy: so if I'm reading that right, it means that the auditors were running and not impacting client requests when using ionice (scenario B). whereas without ionice, they caused a hit to client request performance22:01
*** dmorita has quit IRC22:01
*** dmorita has joined #openstack-swift22:03
*** dmorita has quit IRC22:03
*** dmorita has joined #openstack-swift22:03
*** alejandrito has joined #openstack-swift22:05
claygnotmyname: i mean FWIW the auditors are cranked way up in that test - but that's kinda nice maybe?  You can run your auditors at full bore all the time (thats a thing you might want, right?) and not worry about client facing impact as much.  It's different from the ratelimit approach we've taken against backend-load in the past... but it's not very fine grained - the pattern may not apply nearly as cleanly to other daemons -22:08
notmynameright. I was just trying to confirm I was reading it correctly22:10
claygnotmyname: yeah totally!  booooo auditor IO22:10
notmynameand for the test presented, it looks pretty interesting: like you said, crank up the auditors and not worry as much about client impact. that's great!22:10
torgomaticif that's the only actual benefit to come out of ionice settings, that's still worth it IMO22:16
notmynameI agree22:17
*** finisherr has joined #openstack-swift22:17
finisherrWhen you upload an object with a colon character in the name, does it get character encoded for you as a %3a?22:18
*** klamath has quit IRC22:18
*** mingdang1 has quit IRC22:19
openstackgerritJanie Richling proposed openstack/swift: Support for http footers - Replication and EC  https://review.openstack.org/23739322:20
*** rcernin has quit IRC22:21
finisherrI've noticed that some of my proxy nodes do this22:21
finisherrswift-proxy nodes to be specific22:22
notmynamefinisherr: I saw you asking about that last night in #openstack. looked like you were getting some help, and I was busy with family stuff, so i didn't jump in22:22
notmynamefinisherr: however, basic initial tests and I couldn't duplicate it22:22
finisherrOh yeah, still beating my head against this wall22:22
notmynamefinisherr: it seems like the request is getting double-quoted at some point22:23
*** mrmoje has quit IRC22:23
notmynamedo you have anything between you and the proxy server? ideally, you'd duplicate this with curl as the client and talking directly to the swift proxy. no LB, no apache/nginx, etc22:23
finisherrso, the resulted name of the object ends up with a %3A22:23
finisherrand the PUT calls get logged in proxy.conf as %25253A22:24
*** mrmoje has joined #openstack-swift22:24
finisherrI redirected FIFO sockets running remote ssh calls to two nodes, one with the behavior and another without the behavior. These ssh calls were redirected to vimdiff in readonly mode22:25
finisherrI basicallyed compared every single debian package22:25
finisherrand python packge22:25
finisherrI did noticed that the version of requests/urllib3 was different22:25
*** diazjf1 has quit IRC22:25
openstackgerritSamuel Merritt proposed openstack/swift: Let developers/operators add watchers to object audit  https://review.openstack.org/21282422:26
finisherrbut when i downgraded the version to the one that was working on the other node without the escaping (and after a swift proxy reload), it didn't matter22:26
finisherrno dice22:27
*** blmartin has left #openstack-swift22:27
notmynamefinisherr: are you able to duplicate this with curl?22:28
*** asettle has joined #openstack-swift22:28
finisherrI can duplicate it every single time22:28
openstackgerritTim Burke proposed openstack/python-swiftclient: Prevent test runs from cluttering current directory  https://review.openstack.org/26989122:28
finisherrif I hit one proxy node with the upload, it works without the escaping22:28
notmynamewhat client are you using?22:28
finisherrif i hit the other proxy node, it escapes it22:28
finisherrboto client22:28
finisherrbut the same exact call each time22:28
notmynameso you're also using the swift3 middleware?22:28
finisherri believe so yes22:29
notmynameand are you running the swift proxy directly? or running it behind apache, nginx, or something else?22:29
finisherrone of our software engineers hooked me up with a script to to validate the issue22:29
finisherror repro the issue22:29
finisherrno other proxy22:29
finisherrhitting the swift proxies directly22:29
notmynameany load balancer?22:29
notmynamethere's 2 things you can do to eliminate some variable and hopefully isolate the problem22:30
finisherrwhat's that?22:30
notmynamefirst, don't use boto. use curl. do handcrafted requests. that get's tricky when you're doing the s3-style auth, but that leads to my second suggestion22:31
finisherri was gonna md5sum every single python package22:31
finisherreven though I've got them identical, i'd like to know if the code is in fact the same22:31
notmynamesecond, duplicate it by using the swift API instead of the S3 API. that will eliminate (or identify) swift3 as the culperate22:31
finisherrbut how would that change things if the actual call is the same?22:31
finisherrfrom the same workstation22:32
timburkefinisherr: it wouldn't surprise me if this was a latent swift3 issue. i know i've seen problems with it not finding copy sources as a result of url-encoding being applied one-too-many times; see patch 24601822:32
patchbottimburke: https://review.openstack.org/#/c/246018/ - Accept URL-encoded x-amz-copy-source headers (MERGED)22:32
notmynamebecause you're eliminating code paths22:32
*** petertr7 is now known as petertr7_away22:32
finisherrI haven't looked at the script he gave me yet22:32
finisherrmaybe I should22:32
finisherrhe meaning a software engineer I work with22:32
notmynamefinisherr: here's what I see on swift: when sending %3A in an object name, that's translate to a ":". when sending a :, I get back :. ie swift is unquoting the path for stuff that is valid (url safe)22:34
notmynamelikewise, when I send %25253A, I'll get %252A stored (one unquote)22:35
*** jamielennox|away is now known as jamielennox22:35
finisherrit appears that some of our nodes is translating the OTHER way22:35
finisherrwhen we put with a : we get the %3A22:36
notmynamemaybe. or maybe the client (ie boto)22:36
finisherrI wonder why only when interacting with certain nodes22:36
finisherrswift proxy instances to be specific22:36
*** vinsh has joined #openstack-swift22:36
finisherrif it were the same result every time I'd be more inclined to thing it were the client22:38
finisherrbut who knows22:38
notmynamefinisherr: https://gist.github.com/notmyname/0a4d6be9a608642c959f22:39
notmynamefinisherr: yeah, I agree. but I like to eliminate those variables22:39
finisherrFor sure22:39
notmynamefinisherr: specifically note the last example in that gist. where I use %79, that gets translated to "y" because hex(ord(y)) = 0x79 -> %7922:40
openstackgerritSamuel Merritt proposed openstack/swift: Monkeypatch everything the same  https://review.openstack.org/24537522:42
notmynamefinisherr: so to check if the issue is with swift3 or not, you should use the swift API (not s3). ideally via curl. if the problem goes away, then you can look in swift3, and that has a *lot* smaller surface area for bugs that swift does22:43
finisherrYeah, I'll try it out with curl22:43
notmynameand like timburke said, similar issues have been closed in swift3 before22:43
finisherrThanks for you output22:44
notmynamefinisherr: also, specifically checking that the swift3 middleware is the same on both servers is an extra check you can do22:44
finisherrI did that much22:44
finisherrsame version22:44
finisherrbut need to md5sum to be sure22:44
notmyname`python -c "import swift3; print swift3.__file__" | xargs md5sum`22:46
notmynamewell, I guess that's actually just the __init__.pyc file, and you want everything in the containing directory22:46
*** alejandrito has quit IRC22:51
finisherrSorry, i'm kind of in a meeting but I'm checking in22:52
finisherrthank you for your help22:52
finisherrtoday is just being a huge meeting bomb22:52
finisherrI'll have to get any real work done during my 'night shift' : )22:52
notmynamehurricanerix: around?22:59
*** dmorita has quit IRC23:03
*** km__ has joined #openstack-swift23:05
*** km__ is now known as Guest4080923:06
*** mragupat has quit IRC23:08
*** dmorita has joined #openstack-swift23:09
openstackgerritJohn Dickinson proposed openstack/swift: authors and changelog updates for 2.6.0  https://review.openstack.org/26991123:12
notmynamefirst draft of changelog for 2.6.0 release ^23:12
notmynameclayg: I'm still working on some way to summarize your changes to the ring for the changelog ^23:14
notmyname"Now does more of what you expect"23:14
torgomatic"if it was doing an annoying thing before, it probably stopped"23:16
claygtorgomatic: +223:16
claygtorgomatic: if your ring hurts take 2 of the new rebalance and call me in the morning23:17
torgomaticclayg: exactly23:17
openstackgerritSamuel Merritt proposed openstack/swift: Port FileLikeIter to Python 3  https://review.openstack.org/23701923:23
*** kei_yama has joined #openstack-swift23:25
*** dmorita has quit IRC23:32
openstackgerritSamuel Merritt proposed openstack/swift: Fix a comment's indentation  https://review.openstack.org/26992723:33
*** dmorita has joined #openstack-swift23:34
*** dmorita has quit IRC23:35
*** lcurtis has quit IRC23:35
*** dmorita has joined #openstack-swift23:36
*** finisherr has left #openstack-swift23:36
*** bill_az has joined #openstack-swift23:38
notmynameclayg: "The ring rebalancing algorithm was updated to better handle edge cases and to give better (more balanced) rings in the general case."23:39
notmynameanything to warn about for upgrades?23:39
notmynameI think that sentence is correct, but I also think it's almost simplifies it too much. ok, the ring's better, so why is this in the changelog?23:40
torgomatic"ring builder: set switch to 'more magic'"23:40
openstackgerritTim Burke proposed openstack/swift: Stop logging "client disconnected" for 304 / 412 responses  https://review.openstack.org/24439423:41
*** dmorita has quit IRC23:41
*** dmorita has joined #openstack-swift23:42
notmynamethanks timburke. I had already found one or two of those :-)23:42
notmynameoh, and as for the accents on onovy and peterlisak...I don't know. normally I try to coerce things to ascii, but that's not always done either.23:43
notmynameI do know I can type one of them and I have to copy/paste the other23:43
notmynameclayg: torgomatic: "The ring rebalancing algorithm was updated to better handle edge cases and to give better (more balanced) rings in the general case. New rings will have better initial placement, capacity adjustments will move less data for beter balance, and existing rings that were imbalanced should start to become better balanced as it goes through rebalance cycles."23:45
claygnotmyname: well that makes it sound like it's just better all around23:47
notmynameclayg: yeah. you should definitely upgrade!23:50
notmynameclayg: do you think that's a reasonable changelog entry, or should it be reworked?23:51
pdardeaunotmyname: optimized balance and placement?23:54
onovynotmyname: ionice: yes23:54
onovyaccents: you can write this? šíáýčříáěščýřáíěýčšř?23:54
onovyyes i know. i have funny keyboard23:54
notmynamewell, not without some weird learning of multiple keypresses per letter23:55
onovyi think c'n'p is best way23:55
onovyand you are not writting my/peter name too often, right? :)23:55
notmynameonovy: so to the specific question, I ascii-fied your name and peterlisak's name in the AUTHORS file. I can add back the accents, or I can keep it ascii. your call23:55
onovyi'm fine with ascii23:56
onovy(and i think peter too)23:56
onovybut will it work with your scripts for authors file?23:56
notmynamepretty much every time I update the authors file, I think of this http://www.kalzumeus.com/2010/06/17/falsehoods-programmers-believe-about-names/23:57
onovyok. anything else? it's 1am23:58
notmynameonovy: go to bed!23:58
onovygn :)23:58

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