Wednesday, 2015-08-05

*** haomaiwang has quit IRC00:00
*** haomaiwang has joined #openstack-swift00:00
*** esker has joined #openstack-swift00:02
*** kota_ has joined #openstack-swift00:08
*** ChanServ sets mode: +v kota_00:08
*** zul has joined #openstack-swift00:09
kota_good morning00:09
*** esker has quit IRC00:20
*** esker has joined #openstack-swift00:21
*** haomaiwang has quit IRC00:35
*** delattec has quit IRC00:42
*** cdelatte has quit IRC00:43
*** delattec has joined #openstack-swift00:43
*** nadeem has joined #openstack-swift00:43
*** changbl has quit IRC00:59
*** bill_az has quit IRC01:00
tdasilvakota_: hi!01:16
*** david-lyle has joined #openstack-swift01:19
kota_tdasilva: hi!01:23
tdasilvakota_: thanks for your help with patch 13434701:24
patchbottdasilva: https://review.openstack.org/#/c/134347/01:24
kota_tdasilva: thanks to work a lot for versions_writes and am a bit worried I was overkilling for that.01:25
tdasilvakota_: no no, it's all good...you have been helping find real issues01:25
tdasilvakota_: maintaining backwards compatibility is a real pain01:26
kota_tdasilva: thanks and I think it's ok that we pends something like refactoring the code to follow up patch01:27
kota_and now01:27
kota_I'm looking at patch 134347 addressed the issues you fixed between patch set 41 and 42.01:28
patchbotkota_: https://review.openstack.org/#/c/134347/01:28
*** changbl has joined #openstack-swift01:28
kota_in my quick view, it seems ok but now testing in my env.01:28
tdasilvakota_: sounds good...hopefully we are closer to the end :-)01:29
kota_tdasilva: I think so, too. There are awesome # of patch sets01:30
tdasilvakota_: awesome or awful ? :P01:32
kota_tdasilva: might be awful :/01:32
kota_tdasilva: I don't want to try such a number, I'd like to do handclap for your effort.01:34
tdasilvakota_: it was a series of very small increments and since it's been going for a while, there are a lot of patchsets that are just rebases, so the number is probably a little inflated01:38
kota_tdasilva: I see.01:39
*** nadeem has quit IRC01:40
openstackgerritKota Tsuyuzaki proposed openstack/swift: Refactor versioned_writes middleware  https://review.openstack.org/20931501:57
*** blmartin has joined #openstack-swift01:59
*** zul has quit IRC02:09
kota_tdasilva: still here?02:16
mattoliverauI think he's gone to bed :)02:16
kota_mattoliverau: I think so.02:17
kota_mattoliverau: time zone difference sometimes blocks or conversation.02:17
kota_mattoliverau: can i talk about patch 13434702:18
patchbotkota_: https://review.openstack.org/#/c/134347/02:18
kota_?02:18
kota_mattoliverau: that's versiones_writes middleware stuff.02:18
mattoliveraukota_: sure you can.. I should take a look at the latest revision :)02:19
kota_thanks but I found I might miss something before talking with you.02:19
kota_sorry02:19
kota_with the latest one, I found my curl reports there is X-Versions-Location in the resposne header twice.02:20
kota_but now, I cannot reproduce the situation...02:20
kota_I thought that is because using pure env/start_response in the container_request but...not? I'm not sure.02:21
kota_so I have to dig more before asking you sorry :/02:22
*** annegentle has quit IRC02:22
mattoliverauLol, no probs02:22
*** zaitcev has joined #openstack-swift02:23
*** ChanServ sets mode: +v zaitcev02:23
*** annegentle has joined #openstack-swift02:23
openstackgerritBen Martin proposed openstack/swift: +Document method to avoid rsync filling root drive  https://review.openstack.org/19920302:31
*** gyee has quit IRC02:35
openstackgerritRui Zang proposed openstack/python-swiftclient: Replace assertEquals() with assertEqual() assertEquals() is deprecated in Python 3.x Using it logs a DeprecationWarning in Python 3.x. The same goes for assertNotEquals()  https://review.openstack.org/20931902:36
kota_I got it, I am ready to ask mattliverau!02:41
mattoliverauLol, fire away02:42
kota_mattoliverau: thanks02:42
kota_mattoliveau: so considerting for using versioned_middleware.02:43
kota_at first we have already a config "allow_versions" in container config.02:43
kota_assuming operator set allow_versioning to True already.02:43
kota_and an user set X-Version-Location something like... 'target' container02:44
kota_like PUT /a/c -H 'X-Versions-Location: target'02:44
kota_and then02:45
kota_operator install and setup new versions_write middelware with setting "allow_versions_writes" = True.02:45
kota_"allow_versions_writes" is a config in the middleware to turn on/off the new middleware.02:46
kota_After that, the user who already set versions-location to "target" container change the location to "target2"02:47
kota_like as PUT /a/c -H "X-Versions-Location: target2"02:47
kota_I want to talk about this situation.02:47
kota_In the case, HEAD /a/c shows duplicated X-Versions-Location headers02:48
kota_one is target, the other is target202:48
kota_I already know the reason and the case is not recommended in the middleware documentation.02:49
kota_but I'm not sure I may put good score I leaves it as it is.02:50
kota_and that is absolutely fixed in the follow up patch I wrote up at patch 20931502:51
patchbotkota_: https://review.openstack.org/#/c/209315/02:51
mattoliverauYeah I see the problem02:51
charzpeluse: here02:53
kota_mattoliverau: thanks, ok we can reproduce that.02:54
*** annegentle has quit IRC02:54
kota_mattoliverau: and raise it again for tdasilva at gerrit.02:55
kota_IMO, the duplicated headers might cause another problem around security software.02:56
mattoliverauSeeing as you already have a follow up patch to fix it written and submitted to gerrit, maybe mention the bug in thiagos change, that this patch fixes it. Then give it a good score (but don't +A) (not that you can cause 0 +2s ATM) so thiago can decide if he wants to merge your patch into his.. But personally I think its fine as a follow up, so long as the follow up is merged before the next release02:56
kota_I have an experience that security software raised a bunch of warning when there is duplicated header key.02:57
*** blmartin has quit IRC02:57
mattoliverauTrue, its a good find02:58
kota_mattoliverau: nice comments!02:58
kota_mattoliverau: ok, I'm going to add +2 with my suggestion after taking my lunch :P02:59
kota_mattoliverau: thanks a lot!03:00
mattoliverauSounds great and I won't +A until someone else takes a look ;)03:00
mattoliverauAnytime kota_, we timezone mates need to stick together ;)03:00
zaitcevThere's still a lot of overlap.03:02
mattoliverauIn timezones?03:04
zaitcevYes.03:04
mattoliverauTrue and most the Americans don't seem to sleep ;)03:05
openstackgerritDaisuke Morita proposed openstack/swift: WIP: Changing Policies  https://review.openstack.org/20932903:15
*** dmorita_ has quit IRC03:16
*** sanchitmalhotra has joined #openstack-swift03:20
*** annegentle has joined #openstack-swift03:29
*** jrichli has joined #openstack-swift03:32
*** SkyRocknRoll has joined #openstack-swift03:35
pelusehey can anyone out there explain the semantics behind HTTP_CONFLICT in the case of PUT?  I'm missing something really obvious I think, what is the scenario we are catching here?03:37
peluse(wrt timestamps)03:37
*** annegentle has quit IRC03:38
*** sanchitmalhotra1 has joined #openstack-swift03:46
*** sanchitmalhotra has quit IRC03:48
*** hrou has quit IRC03:50
openstackgerritKota Tsuyuzaki proposed openstack/swift: Refactor versioned_writes middleware  https://review.openstack.org/20931504:01
*** zaitcev has quit IRC04:05
*** kota_ has quit IRC04:06
*** mahatic has joined #openstack-swift04:09
*** janonymous has quit IRC04:32
*** fifieldt has joined #openstack-swift04:32
*** SkyRocknRoll has quit IRC04:35
*** ppai has joined #openstack-swift04:35
*** ppai has quit IRC04:41
*** bill_az has joined #openstack-swift04:45
*** remix_tj has quit IRC04:50
*** jrichli has quit IRC04:58
*** pcaruana has quit IRC04:59
*** haomaiwang has joined #openstack-swift05:08
hopeluse: i guess time difference b/w proxy-serves. one request via proxy-server A and next request via proxy-server to account/container/object-servers. if timestamp of laster request is older than first request, we can get 409.05:12
*** tab___ has quit IRC05:12
*** silor has joined #openstack-swift05:31
*** bill_az has quit IRC05:41
*** bill_az has joined #openstack-swift05:43
*** SkyRocknRoll has joined #openstack-swift05:44
*** bill_az has quit IRC05:45
*** bill_az has joined #openstack-swift05:46
*** janonymous has joined #openstack-swift05:47
*** haomaiwang has quit IRC05:52
*** haomaiwang has joined #openstack-swift05:52
*** esker has quit IRC06:01
*** esker has joined #openstack-swift06:03
*** hezhiqiang has joined #openstack-swift06:10
*** breitz has joined #openstack-swift06:29
*** breitz has quit IRC06:33
*** marzif_ has quit IRC06:35
*** pcaruana has joined #openstack-swift06:36
*** breitz has joined #openstack-swift06:37
*** SkyRocknRoll_ has joined #openstack-swift06:37
*** SkyRocknRoll has quit IRC06:39
*** breitz has quit IRC06:41
openstackgerritClay Gerrard proposed openstack/swift: Fix TypeError if backend response doesn't have expected headers  https://review.openstack.org/20937006:50
*** marzif_ has joined #openstack-swift07:13
*** marzif__ has joined #openstack-swift07:16
*** marzif_ has quit IRC07:18
*** silor has quit IRC07:18
*** geaaru has joined #openstack-swift07:22
*** ppai has joined #openstack-swift07:28
*** geaaru has quit IRC07:50
*** SkyRocknRoll_ has quit IRC07:52
*** bill_az has quit IRC07:55
*** tsubic has joined #openstack-swift08:01
*** jordanP has joined #openstack-swift08:02
*** joeljwright has joined #openstack-swift08:08
*** ChanServ sets mode: +v joeljwright08:08
*** jistr has joined #openstack-swift08:14
*** geaaru has joined #openstack-swift08:16
*** jordanP has quit IRC08:20
*** jordanP has joined #openstack-swift08:21
*** acoles_away is now known as acoles08:23
*** barra204 has quit IRC08:25
acolesgood morning08:26
joeljwrightmornin08:27
acoleshi joeljwright you going to make it to austin next week?08:27
joeljwrightno, sadly not08:27
joeljwrightI was chatting to people and it seemed that swiftclient stuff was unlikely to be high priority in Austin08:28
acolesok nm.08:28
acolesthat sounds like a reasonable assumption08:28
*** km has quit IRC08:40
*** remix_tj has joined #openstack-swift08:41
*** sanchitmalhotra has joined #openstack-swift08:51
*** kei_yama has quit IRC08:51
*** sanchitmalhotra1 has quit IRC08:52
openstackgerritHisashi Osanai proposed openstack/swift: Remove _keystone_identity method  https://review.openstack.org/20405009:01
*** mahatic has quit IRC09:24
*** hezhiqiang has quit IRC09:24
*** sanchitmalhotra1 has joined #openstack-swift09:30
*** sanchitmalhotra has quit IRC09:31
*** sanchitmalhotra has joined #openstack-swift09:47
*** sanchitmalhotra1 has quit IRC09:49
*** mahatic has joined #openstack-swift10:11
*** ho has quit IRC10:23
openstackgerritKota Tsuyuzaki proposed openstack/swift-specs: Add global ec cluster improvement  https://review.openstack.org/20944710:35
*** mahatic has quit IRC10:38
*** jasondotstar has joined #openstack-swift10:47
*** h00327910__ has quit IRC10:58
*** johnsonejezie has joined #openstack-swift10:59
*** hezhiqiang has joined #openstack-swift11:04
*** mahatic has joined #openstack-swift11:06
*** johnsonejezie_ has joined #openstack-swift11:13
*** johnsonejezie has quit IRC11:13
*** johnsonejezie_ is now known as johnsonejezie11:13
*** marzif__ has quit IRC11:15
*** marzif__ has joined #openstack-swift11:16
*** sanchitmalhotra1 has joined #openstack-swift11:31
*** sanchitmalhotra has quit IRC11:33
*** johnsonejezie has quit IRC11:36
*** johnsonejezie has joined #openstack-swift11:38
johnsonejezieHi everyone. getting this error The operation couldn’t be completed. (com.google.HTTPStatus error 400. with google GIDSignIn. It only happens when user has google+ app installed in the device. And if I remove the [GIDSignIn sharedInstance].setServerClientID, it works. Any idea why this is so? Thanks11:40
*** joeljwright has quit IRC11:41
*** sanchitmalhotra has joined #openstack-swift11:42
*** sanchitmalhotra1 has quit IRC11:45
tdasilvajohnsonejezie: sounds like you have a question for apple swift11:54
johnsonejezietdasilva hi I already asked the question. Did you see it?11:55
*** jistr has quit IRC11:56
tdasilvajohnsonejezie: sorry, what i meant to say is that I think you are asking the question in the wrong place. This channel is for openstack-swift11:57
*** johnsonejezie has left #openstack-swift11:59
*** mahatic_ has joined #openstack-swift12:04
*** mahatic has quit IRC12:04
*** mahatic_ has quit IRC12:11
*** mahatic has joined #openstack-swift12:11
*** jistr has joined #openstack-swift12:18
*** jistr is now known as jistr|biab12:19
*** alejandrito has joined #openstack-swift12:19
*** robefran has joined #openstack-swift12:33
*** hezhiqiang has quit IRC12:37
*** hrou has joined #openstack-swift12:37
*** esker has quit IRC12:38
*** marzif__ has quit IRC12:43
*** marzif__ has joined #openstack-swift12:44
*** marcusvrn has joined #openstack-swift12:49
*** bkopilov has quit IRC12:51
*** bapalm has joined #openstack-swift12:52
*** bapalm_ has joined #openstack-swift12:53
*** joeljwright has joined #openstack-swift12:56
*** ChanServ sets mode: +v joeljwright12:56
*** bapalm has quit IRC12:57
*** thumpba has joined #openstack-swift13:01
*** dustins has joined #openstack-swift13:02
*** petertr7_away is now known as petertr713:03
*** thumpba has quit IRC13:04
*** zul has joined #openstack-swift13:05
*** jistr|biab is now known as jistr13:06
*** esker has joined #openstack-swift13:07
*** bill_az has joined #openstack-swift13:09
*** alejandrito has quit IRC13:10
*** esker has quit IRC13:12
*** mahatic has quit IRC13:21
*** sanchitmalhotra has quit IRC13:28
*** sanchitmalhotra has joined #openstack-swift13:29
*** silor has joined #openstack-swift13:30
*** petertr7 is now known as petertr7_away13:34
*** mahatic has joined #openstack-swift13:36
*** petertr7_away is now known as petertr713:38
*** ppai has quit IRC13:45
*** sc68cal_ has joined #openstack-swift13:46
*** esker has joined #openstack-swift13:47
*** breitz has joined #openstack-swift13:51
*** briancline has quit IRC13:52
*** sc68cal has quit IRC13:52
*** ccavanna has joined #openstack-swift13:52
*** jrichli has joined #openstack-swift13:53
*** blmartin has joined #openstack-swift13:53
*** briancline has joined #openstack-swift13:54
*** esker has quit IRC14:04
*** zigo has quit IRC14:05
*** esker has joined #openstack-swift14:05
*** sanchitmalhotra has quit IRC14:06
*** zigo has joined #openstack-swift14:06
*** NM has joined #openstack-swift14:07
*** thumpba has joined #openstack-swift14:07
openstackgerritCharles Hsu proposed openstack/swift: Add extra_header_count to document and config.  https://review.openstack.org/20952514:10
*** breitz has quit IRC14:11
*** esker has quit IRC14:12
*** esker has joined #openstack-swift14:14
*** jlhinson has joined #openstack-swift14:15
mahaticjrichli: I put down some notes for probe tests on etherpad. FYI14:16
*** delattec has quit IRC14:22
*** annegentle has joined #openstack-swift14:23
*** mordred has quit IRC14:25
jrichlimahatic: Thanks.  I made some comments as well.14:26
openstackgerritBen Martin proposed openstack/swift: +Document method to avoid rsync filling root drive  https://review.openstack.org/19920314:28
openstackgerritBen Martin proposed openstack/swift: +Document method to avoid rsync filling root drive  https://review.openstack.org/19920314:29
mahaticjrichli: sure, looking14:30
*** bapalm_ has quit IRC14:38
*** esker has quit IRC14:39
*** NM has quit IRC14:39
*** esker has joined #openstack-swift14:40
ccavannaHi. Good moring everyone. Question about Jerkins tests. I noticed that the cluster test was re-added. However, I also noticed that when it was misconfigured last week (there are some setup errors)14:41
ccavannaI tried rerunning posting a "recheck" comment. However, that reruns the Jenkins check only. Is there another way to re-run the cluster tests?14:41
ccavannaOr should I just issue a git review again.14:42
*** jistr has quit IRC14:44
*** jistr has joined #openstack-swift14:44
*** delattec has joined #openstack-swift14:45
*** bapalm_ has joined #openstack-swift14:46
*** amoturi has joined #openstack-swift14:49
*** robefran has quit IRC14:49
*** blmartin has quit IRC14:51
*** blmartin has joined #openstack-swift14:52
tdasilvaccavanna: is the 'cluster tests' the same as the swiftstack tests?14:59
tdasilvaccavanna: I think charz can manually re-run them14:59
ccavannaI think they are the same.14:59
ccavannaThe full name of the tests is "SwiftStack Cluster CI check"14:59
tdasilvadon't know of an automated way besides another patchset14:59
ccavannaI can just post an update on the commit message, then.15:00
ccavannaThat should trigger the re-check.15:00
tdasilvayep15:00
*** esker has quit IRC15:00
charzccavanna: or you can add comment "recheck".15:01
ccavannarecheck didn't work.15:01
*** mahatic has quit IRC15:01
ccavannaIt only re-rean Jenkins tests.15:01
charzccavanna: which patchset.15:01
*** esker has joined #openstack-swift15:02
*** esker has quit IRC15:02
ccavannahttps://review.openstack.org/#/c/202657/15:02
*** breitz has joined #openstack-swift15:02
ccavannalatest patset and the previous one.15:02
*** esker has joined #openstack-swift15:02
*** mahatic has joined #openstack-swift15:02
charzccavanna: one second, let me check the jenkins15:02
*** minwoob has joined #openstack-swift15:03
*** zaitcev has joined #openstack-swift15:04
*** ChanServ sets mode: +v zaitcev15:04
hrouThose aren't gating ccavanna right ?  I think jenkins gave you a +1 ? I think someone mentioned there were issues with those ?15:12
ccavannaI saw those tests executing properly in another review. I guess they were fixed.15:13
charzccavanna: That's our code issue, I'll fix that. I'm going to re-trigger it for your patchset manually.15:13
ccavannacharz: great. thanks.15:13
*** marzif__ has quit IRC15:15
*** marzif__ has joined #openstack-swift15:16
*** esker has quit IRC15:22
*** esker has joined #openstack-swift15:23
*** esker has quit IRC15:23
*** esker has joined #openstack-swift15:25
*** esker has quit IRC15:30
*** resker has joined #openstack-swift15:30
*** esker has joined #openstack-swift15:30
*** bapalm_ has quit IRC15:34
*** petertr7 is now known as petertr7_away15:35
*** shakamunyi has joined #openstack-swift15:35
*** gyee has joined #openstack-swift15:36
*** nadeem has joined #openstack-swift15:39
*** acoles is now known as acoles_away15:40
*** esker has quit IRC15:40
*** nadeem has quit IRC15:44
*** nadeem has joined #openstack-swift15:44
*** esker has joined #openstack-swift15:48
*** robefran has joined #openstack-swift15:55
*** bapalm_ has joined #openstack-swift15:57
*** esker has quit IRC15:57
*** robefran_ has joined #openstack-swift15:58
*** esker has joined #openstack-swift15:59
*** robefran_ has quit IRC15:59
robefranI cloned a new SAIO env (2.3.1.dev244) and I hit an error linking libJerasure.so16:00
robefranliberasurecode_backend_open: dynamic linking error libJerasure.so: cannot open shared object file: No such file or directory16:00
robefranam I missing some SAIO setup?16:01
charzccavanna: I fixed that, it seems we ignore "recehck" in comment, but "recheck no bug" is work.16:01
ccavannado you want me to post a "recheck no bug" to test it?16:01
charzccavanna: "recheck" is work now, and your patchset are already in jenkins queue.16:02
ccavannaoh, great.16:02
ccavannaThank you!16:02
charzccavanna: you're welcome.16:02
robefranfound a solution, apt-get install libjerasure-dev (should this be added to SAIO instructions?)16:09
*** doffm has joined #openstack-swift16:12
pelusetsg coming....16:13
*** tsg has joined #openstack-swift16:13
*** ccavanna has quit IRC16:17
*** ccavanna has joined #openstack-swift16:17
*** jordanP has quit IRC16:19
minwoobHas anyone seen unit test errors=29 after pulling in the latest changes from master?16:21
*** breitz has quit IRC16:22
*** jistr has quit IRC16:23
*** esker has quit IRC16:23
*** jistr has joined #openstack-swift16:23
*** jistr has quit IRC16:24
*** pcaruana has quit IRC16:26
*** janonymous_ has joined #openstack-swift16:27
*** esker has joined #openstack-swift16:27
openstackgerritpaul luse proposed openstack/swift: Add ability for GET path to see/select alternate frag archs  https://review.openstack.org/20716516:27
tsgrobefran: missed the  pyeclib/liberasurecode discussion earlier - mind repeating what problem were you having and what did you have to do to fix it?16:29
peluseminwoob, nope, did you update your test requirements though?16:29
*** petertr7_away is now known as petertr716:29
robefranI cloned out a new SAIO env (2.3.1.dev244) and hit16:29
*** openstackgerrit_ has joined #openstack-swift16:30
robefranliberasurecode_backend_open: dynamic linking error libJerasure.so: cannot open shared object file: No such file or directory16:30
robefranapt-get install python-pyeclib libjerasure-dev, solved my problem16:30
robefranI'll look into adding this to the SAIO instructions16:31
tsgrobefran: that makes sense - we have had jerasure as the default backend for Swift16:31
tsgrobefran: although - the "sudo pip install -r requirements.txt" step should have taken care of jerasure install16:33
tsgrobefran: can you check if you didn't do that as part of SAIO bring up?  Or if the pip install step failed?16:34
peluseyeah, I just built a new VM from the SAIO directions last week and didn't have any issues.  I'll fire that VM up real quick and make sure EC works there16:34
mahaticI've had issues whil installing cryptography though (when I ran sudo pip install -r requirements.txt). I had to manually install OpenSSL dev package and then run that step again16:36
mahaticI ran this "sudo apt-get install libssl-dev" for OpenSSL16:37
*** acoles_away is now known as acoles16:37
hrouHey mahatic, yep me too i.e. 'sudo apt-get install libssl-dev', but I think the reason for that is building swift, it needs some C libraries for AES.16:37
peluserobefran, so my fresh VM following current SAIO has no issues w/EC and installed the dependencies w/o issue.  Maybe something went wrong with your isntall that you didn't notice (failure lost in scroll by or something)?16:38
tsgyeah, perhaps a network/build issue16:39
mahatichrou: shouldn't we be including that in the SAIO instructions?16:39
*** esker has quit IRC16:39
pelusemahatic, yeah, I think I had to do that too but that's a different topic than the EC stuff (in case anyone thought this was one conversation).  I think the error there is a missing .h file until you do that step.  That's something to fix on feature/crypto16:39
hroumahatic, I think we should find a way to include it through requirements.txt or what not if at all possible.16:40
mahatichrou: part of requirements.txt16:40
mahaticyeah16:40
robefranpeluse,tsg: I'll check back through the scroll to see if requirements.txt failed16:40
tsgthanks16:40
mahaticpeluse: ah16:41
peluserobefran, cool - I'm sure I missed an error in the ton of stuff that flies by during the first install of the EC requirments at least once in the past...16:41
minwoobpeluse: Yeah, that was the issue. Thanks.16:41
mahaticpeluse: although i didn't think this was to do with EC stuff16:42
peluseminwoob, OK, what is the error?16:42
hrourobefran,  I also just recently set things up, and noticed I already had libjerasure-dev pulled down, which I assume comes from PyECLib, yea maybe we can repeat the SAIO install and see what error happened.16:42
robefranpeluse: scroll doesn't go back far enough, I'll try again with a clean VM, capture the output and see if requirements.txt failed16:43
pelusethat'd be great.  thanks16:44
mahaticrobefran: you could enable that in terminal settings I believe, scrolling unlimited16:44
mahaticyeah a capture would work too :)16:44
*** bill_az has quit IRC16:46
*** bill_az has joined #openstack-swift16:47
*** dmorita has joined #openstack-swift16:48
notmynamegood morning16:56
mahaticmorning16:59
*** bapalm_ has quit IRC17:00
*** openstackgerrit_ has quit IRC17:05
*** esker has joined #openstack-swift17:09
*** esker has quit IRC17:09
*** esker has joined #openstack-swift17:10
*** tongli has joined #openstack-swift17:12
*** SkyRocknRoll has joined #openstack-swift17:16
*** esker has quit IRC17:16
*** esker has joined #openstack-swift17:20
*** esker has quit IRC17:21
*** openstackgerrit_ has joined #openstack-swift17:24
mahatichrou: here17:27
mahatichrou: yeah there was a discussion about probe tests, it's on the etherpad17:27
mahatichttps://etherpad.openstack.org/p/swift_encryption_issues17:28
hroumahatic, great I'll check that out17:28
*** chsc has joined #openstack-swift17:28
pelusecharz, still around?17:29
charzpeluse: here17:29
mahatichrou: I'm currently looking into that17:30
pelusecool, hey can you send me the exact cmd line you use to start ssbench in your repro (i'm no a big ssbench user)17:30
hroumahatic, thanks, I see there's still some debate about the solution.17:30
hroumahatic, oh ok that sounds great!17:30
charzpeluse: sure17:31
pelusecharz, just paste here in IRC is fine :)  thanks17:31
peluseoh wait, looks like you did that already in the gist.  geeze17:32
pelusecharz, wait, the bug says you added it but I dont see it17:33
charzpeluse: yeah, just want to post gist link to here again.17:33
*** zhill has joined #openstack-swift17:33
charzpeluse: https://gist.github.com/charz/5e53c3f735f736bc6ced#file-1-ssbench-result-md search "ssbench-master"17:34
*** marzif__ has quit IRC17:34
peluseahh, see it now.  thanks.  will try17:34
*** openstackgerrit_ has quit IRC17:34
peluseOK, runs now.  thanks :)17:35
*** SkyRocknRoll has quit IRC17:35
*** SkyRocknRoll_ has joined #openstack-swift17:35
*** mahatic has quit IRC17:38
tdasilvaseriously??? http://ur1.ca/nd1vh17:41
hroutdasilva, : ) I ran into this that's an old version of hacking (H402 no longer exists), the new test-requirments.txt specifies a newer version of hacking (and you need to rebuild tox env with tox -r ...)17:43
tdasilvahrou: phew, thanks! trying out now17:43
hroutdasilva, sure thing, that should do the trick (make sure to do the  pip install -r test-requirements.txt step first)17:44
tdasilvahrou: cool, will do, thx :-)17:45
*** petertr7 is now known as petertr7_away17:46
*** dustins has quit IRC17:46
*** dustins has joined #openstack-swift17:52
*** bkopilov has joined #openstack-swift17:54
zaitcevFedora is backwards: onelines MUST NOT have the ending period. Just because.17:54
zaitcevYou can see it enforced in RPMs.17:55
*** annegentle has quit IRC17:56
*** petertr7_away is now known as petertr717:56
*** bapalm_ has joined #openstack-swift18:00
*** thurloat_isgone is now known as thurloat18:03
*** lpabon has joined #openstack-swift18:04
*** tsg has quit IRC18:06
*** bapalm_ has quit IRC18:06
*** tsg has joined #openstack-swift18:12
*** esker has joined #openstack-swift18:13
*** acoles is now known as acoles_away18:18
*** aagrawal has joined #openstack-swift18:29
*** zhill has quit IRC18:32
*** annegentle has joined #openstack-swift18:35
*** zhill has joined #openstack-swift18:41
*** lpabon has quit IRC18:43
*** openstackgerrit has quit IRC18:46
*** openstackgerrit has joined #openstack-swift18:46
robefranpeluse: I cloned a SAIO in a new VM, and hit the same libJerasure.so error. I also captured the log off my SAIO setup/clone. It successfully installed PyECLib-1.0.718:48
peluserobefran, not sure I understand, same issue but it successfully installed??  Which is it, did it work ot not? :)18:52
robefranpeluse: my SAIO setup is scripted, but I don't see any diffs in the apt-gets from the SAIO instructions. My setup did not apt-get install libjerasure-dev18:52
robefranpeluse: the SAIO instructions I'm looking at are: http://docs.openstack.org/developer/swift/development_saio.html18:53
*** tsg has quit IRC18:54
*** tsg has joined #openstack-swift18:54
robefranpeluse: It did not work out. The startmain fails with the missing libJerasure.so error. So far I have not manually done the apt-get install libjerasure-dev.18:55
tdasilvarobefran, peluse: so it is installing pyECLib-1.0.7, but not libjerausre-dev?18:55
robefrantdasilva: yes, for me, it is installing PyECLIb-1.0.7, but not libjerasure-dev18:55
peluseI'm going to be offline for a bit but I'd prefer this be filled as a bug and not a doc patch.  AFAIK right now it is supposed to just work (and for some reason did for me but now that I think about it I may have had older versions installed on that VM)18:56
pelusetsg, can you confirm that pip install pyeclin is indeed suppsoed to install libjerasure?18:57
tsgrobefran: can you check if you have libJerasure.so on your system?18:58
tsgrobefran: the next step would be to run "sudo ldconfig" and try again18:58
peluseoh yeah, I forgot I had to do that too!  its in the last output line after the isntall so if scripted you might not see it18:59
peluseif that's the problem then yes *that* should be in the doc18:59
robefrantsg, peluse: I did not do the ldconfig, I'll give it a try18:59
pelusecharz, still there?19:01
*** bhakta has quit IRC19:03
*** ajiang has quit IRC19:03
*** ajiang has joined #openstack-swift19:04
*** bhakta has joined #openstack-swift19:05
openstackgerritThiago da Silva proposed openstack/swift: versioned writes middleware  https://review.openstack.org/13434719:29
openstackgerritjanonymous proposed openstack/swift: pep8 fix: assertEquals -> assertEqual  https://review.openstack.org/20964219:29
openstackgerritjanonymous proposed openstack/swift: pep8 fix: assertEquals -> assertEqual  https://review.openstack.org/20964319:30
openstackgerritjanonymous proposed openstack/swift: pep8 fix: assertEquals -> assertEqual  https://review.openstack.org/20964519:31
openstackgerritjanonymous proposed openstack/swift: pep8 fix: assertEquals -> assertEqual  https://review.openstack.org/20964619:32
robefrantsg: ldconfig did not help. I did not find a libJerasure.so on my system. I'm looking in /usr/lib/x86_64-linux-gnu/19:33
openstackgerritjanonymous proposed openstack/swift: pep8 fix: assertEquals -> assertEqual  https://review.openstack.org/20964719:33
*** silor has quit IRC19:34
tsgrobefran: can you capture the output of "sudo pip install -r requirements.txt" and share a gist?19:35
janonymous_Please find the patches for assert checks for review  ,i have uploaded the changes based on directories to avoid large patch set and easy reviews :)19:37
*** zhill has quit IRC19:42
*** bkopilov has quit IRC19:43
*** NM has joined #openstack-swift19:45
*** bkopilov has joined #openstack-swift19:45
*** zhill has joined #openstack-swift19:51
openstackgerritBill Huber proposed openstack/swift: pep8 fix: assertEquals -> assertEqual  https://review.openstack.org/20964919:53
*** proteusguy has quit IRC19:54
*** esker has quit IRC19:55
robefrantsg: the captured sudo pip install -r requirements.txt stated Requirement already satisfied19:55
robefrantsg: I tried sudo pip install --upgrade --force-reinstall PyECLib (after the ldconfig), and this did the trick. My startmain now works.19:56
*** janonymous_ has left #openstack-swift19:56
robefrantsg: should I update the SAIO instructions to run a ldconfig before the sudo pip install -r requirements.txt ?19:57
*** thumpba has quit IRC19:58
tsgrobefran: after pip install makes sense - not sure why you would suggest running it before? (unless you were install jerasure by hand)19:59
tsgrobefran: that is, you were installing jerasure from source before running "pip install" ..19:59
openstackgerritBill Huber proposed openstack/swift: pep8 fix: assertEquals -> assertEqual  https://review.openstack.org/20965220:00
peluserobefran, so this is a pyeclib thing (not a swift thing) but we do need to make note of it since its kind of important and clearly the msg that spits out at the end of the pyeclib isntall can easilybe missed20:01
robefrantsg: I did the following20:01
robefrantsg: sudo ldconfig20:01
robefrantsg: sudo pip install --upgrade --force-reinstall PyECLib20:01
robefrantsg: the ldconfig by itself did not solve the problem20:02
peluserobefran, however its not just the SAIO doc... please submit a patch with a small but clear note for the user docs as well as SAIO (maybe in the form of a footnote that calls out this step as being needed by one of the swift requiements)20:02
pelusetsg, remind me again why this can't be part of the install of pyeclib (this last step)?20:02
*** bkopilov has quit IRC20:04
tsgpeluse: it can be - although with v1.0.8 we don't have jerasure integrated with pyeclib anymore20:04
tsgso soon enough jerasure install will have to be a separate step anyway20:04
peluseso the step isn't needed > 1.0.8?  At that point in time do we then need to add it as its own requirement or do we need to then add doc to explain how to install one of the available EC libs?20:05
peluse>=20:05
peluseoh you already answered before I finished typing :)20:05
peluseOK so robefran don't submit a patch right now it will just make it even more confusing20:06
tsgpeluse: we'll switch to "liberasurecode_rs_vand" for default EC scheme for Swift (I have a patch I need to submit for review)20:06
*** proteusguy has joined #openstack-swift20:06
pelusetsg, OK, cool.  that's the answer then20:06
tsgpeluse: so jerasure will not be a must-have for SAIO20:06
robefrantsg, peluse: ok, I won't submit anything20:06
pelusewhen?20:06
peluse1.0.8 that is20:06
tsg1.0.8 has already been tagged - zigo built .debs last night20:07
peluseand at that time lets be *sure* to document where to find directions on installing both ISA-L and jerasure after the fact, we don't have that now20:07
tsgwe need to push a pypi package .. should happen today20:07
pelusegreat!20:07
tsgpeluse: makes sense20:07
*** jasondot_ has joined #openstack-swift20:07
notmynametsg: but isn't there an issue with the packaging that will allow a 1.0.9 quickly?20:08
notmynameor was that only "for next time" sort of thing20:08
tsgpeluse: we are holding off pypi release due to the integrated-liberasurecode thingy, where we may just move to 1.0.920:08
tsgnotmyname: those changes have already been pushed20:08
notmynameok20:08
notmynamegreat20:08
*** jasondotstar has quit IRC20:08
tsgso liberasurecode-1.0.8, pyeclib-1.0.9 is what we are thinking20:08
pelusenotmyname, do you prefer we include instructions on how to install jerasure and ISAL in the swift docs or just include links to other sites that explain how to install those?20:09
tsgpeluse: that's a great idea IMO20:09
notmynamepeluse: in swift20:09
pelusenotmyname, agreed but wanted to make sure you don't have an issue with it20:09
notmynamepeluse: nope. not right now20:09
pelusetsg, can you include the doc change w/the req change so we know they land at the same time?20:10
tsgpeluse: will do20:10
pelusegracias20:10
peluserobefran, sorry for the confusion :)20:10
notmynametsg: and when you push to pypi you'll push as 1.0.9 with the changes zigo said?20:11
notmynametsg: and it will be tagged as such in the repo?20:11
tsgnotmyname: yes20:11
notmynamegreat20:11
notmynamesame version everywhere20:11
pelusereobefan, and thanks for asking the questions that help drive making this thing better20:11
tsgnotmyname: the only relevant change there for us will be making the liberasurecode install optional20:11
robefranpeluse: it's amazing how I trip over these things just before they are solved. I must remember to remove that 'kick me' sign from my back ;-)20:11
tsg(automated install that is)20:11
peluseheh, OK I'm outta here.... later20:12
tsgrobefran: it will still be good to get a log from your "sudo pip install -r requirements.txt" on a clean VM20:12
tsgrobefran: just in case we still need to amend the docs ..20:12
tsgrobefran: can you send me the complete log (with "screen" or similar)20:13
*** sc68cal_ is now known as sc68cal20:13
robefrantsg: I have to run, but I have the log saved, and will get it to you tomorrow20:13
*** bapalm has joined #openstack-swift20:16
*** bkopilov has joined #openstack-swift20:18
*** amoturi has quit IRC20:25
*** robefran has quit IRC20:27
*** bill_az has quit IRC20:28
*** bill_az has joined #openstack-swift20:30
*** marzif__ has joined #openstack-swift20:38
*** changbl has quit IRC20:39
*** changbl has joined #openstack-swift20:40
*** NM has quit IRC20:44
*** itlinux has joined #openstack-swift20:46
itlinuxhi all20:46
itlinuxdoes anyone have a ref by chance of the usage of Swift as a photo storage20:47
itlinuxI know shutterfly uses object but not sure which one. Thanks20:47
notmynameitlinux: all of the wikimedia images (ie all the pictures you see on wikipedia) are stored in swift20:47
itlinuxthanks I forgot about that.20:48
tdasilvaitlinux: http://blog.wikimedia.org/2012/02/09/scaling-media-storage-at-wikimedia-with-swift/20:48
*** zhill has quit IRC20:48
notmynameand yes, one of those big ones (like shutterfly) uses swift, but I'm not sure how public that is20:49
itlinuxthank you20:49
openstackgerritBill Huber proposed openstack/swift: pep8 fix: assertEquals -> assertEqual  https://review.openstack.org/20964920:50
notmynameitlinux: oh, ok. I found some public references. snapfish uses swift20:50
*** aerwin has joined #openstack-swift20:53
*** ho has joined #openstack-swift20:54
*** acoles_away is now known as acoles20:57
mattoliverauMorning20:57
notmynamegood morning20:57
*** kota_ has joined #openstack-swift20:57
*** ChanServ sets mode: +v kota_20:57
*** zhill has joined #openstack-swift20:57
joeljwrightmorning20:57
kota_morning20:57
hogood morning20:59
notmynameok, it's meeting time. wheeee20:59
*** lcurtis has joined #openstack-swift21:02
*** blmartin has quit IRC21:03
*** breitz has joined #openstack-swift21:05
-openstackstatus- NOTICE: Zuul has been restarted to resolve a reconfiguration failure: previously running jobs have been reenqueued but change events between 19:50-20:54 UTC have been lost and will need to be rechecked or their approvals reapplied to trigger testing.21:06
notmynameMooingLemur: you should fork cowsay21:06
*** bapalm has quit IRC21:06
MooingLemurhah21:07
*** petertr7 is now known as petertr7_away21:10
-openstackstatus- NOTICE: Correction: change events between 20:50-20:54 UTC (during the restart only) have been lost and will need to be rechecked or their approvals reapplied to trigger testing.21:11
*** chlong has quit IRC21:16
*** kcaj has left #openstack-swift21:28
*** chlong has joined #openstack-swift21:29
openstackgerritPete Zaitcev proposed openstack/swift: Having said H, I, J, we ought to say K  https://review.openstack.org/20967721:34
*** breitz1 has joined #openstack-swift21:37
*** breitz has quit IRC21:39
*** marzif__ has quit IRC21:39
*** bill_az has quit IRC21:39
*** haomaiwang has quit IRC21:42
notmynamezaitcev: nice easy patches :-)21:42
zaitcevnotmyname: sorry. A guy at work asked for Swift 101 and one after another things happened.21:43
notmynamezaitcev: no, I think it's great21:43
zaitcevWe hired someone to care for Manila, so at least that's not on me anymore.21:44
notmynamecool21:44
*** haomaiwang has joined #openstack-swift21:44
* mattoliverau is off to find breakfast21:44
*** zul has quit IRC21:45
kota_tdasilva: added my thought to versions_writes one.21:46
timburkenotmyname: at what point should start dropping releases from that doc?21:47
*** kota_ has left #openstack-swift21:47
*** kota_ has joined #openstack-swift21:48
*** ChanServ sets mode: +v kota_21:48
notmynametimburke: when the existing links break? honestly, I don't want to spend any time worrying about it ;-)21:49
tdasilvakota_: just responded :)21:55
tdasilvakota_: I need to drive home, but will check-in again tonight21:55
kota_tdasilva: ok, let me think.21:55
tdasilvakota_: thanks!21:55
honotmyname: as for the allow_account_management, i checked the situation. http://paste.openstack.org/show/409723/ . if doc is correct, we need to change bahavior No.2?21:56
*** acoles is now known as acoles_away21:57
notmynameho: allow_account_management is not an ACL. it enables or disables the ability for the proxy to respond to account-level PUT or DELETE requests (ie RBAC cannot override it)21:57
notmynameho: seems that (5) needs to be updated (the Juno installation guide)21:58
honotmyname: i know but in keystone env, the value is FALSE, there is no way to delete unnecessary accounts. it might be a bug.21:59
*** dustins has quit IRC21:59
*** hrou has quit IRC21:59
notmynameI'm not saying it should always be false. just that the default IMO should be false. you don't want a swift proxy server that is connected to the public internet to allow for account deletion22:00
notmynameperhaps there need to be more/better docs? "you should deploy a proxy server that is firewalled so that only keystone can talk to it and this one allows account management and is used to provision and delete accounts"22:01
honotmyname: i see. we need to have double checking to create/delete accounts for security reason. unn... i need to have more time to think more deeper thanks! :)22:06
notmynameho: yeah, I think it comes down to figuring out what the best deployment decision is and then documenting that in swift (including sample configs). let's make sure that deployers get all the right info22:07
*** lcurtis has quit IRC22:09
*** zhill has quit IRC22:14
*** jrichli has quit IRC22:14
*** kota_ has quit IRC22:16
-ho- notmyname: in keystone env, deployer needs to set TRUE for allow_account_management because they need to delete account.22:19
-ho- notmyname: account can be created via put container (auto_create=TRUE) and it's worked with RBAC only.22:19
-ho- notmyname: means allow_account_management(FALSE) doesn't work in keystone env. the conclusion is keystone should be deployed in safe place? maybe??22:19
*** delattec has quit IRC22:21
*** bill_az has joined #openstack-swift22:21
*** zhill has joined #openstack-swift22:25
*** tongli has quit IRC22:28
*** bill_az has quit IRC22:29
*** bill_az has joined #openstack-swift22:34
*** marcusvrn has quit IRC22:43
*** hrou has joined #openstack-swift22:45
*** thurloat is now known as thurloat_isgone22:45
*** nadeem has quit IRC22:47
*** jlhinson has quit IRC22:48
*** tsg has quit IRC22:50
*** kota_ has joined #openstack-swift22:51
*** ChanServ sets mode: +v kota_22:51
*** mordred has joined #openstack-swift22:55
*** xnox has quit IRC22:57
honotmyname: thanks! i think i understand. spec is if allow_account_management is false, put/delete accounts are not allowed even if they have right role to do it. if this is correct understanding, i will make a patch for "account can be created via put container (auto_create=TRUE)"22:59
*** xnox has joined #openstack-swift22:59
*** km has joined #openstack-swift23:01
*** tsg has joined #openstack-swift23:01
*** esker has joined #openstack-swift23:06
*** esker has quit IRC23:22
*** esker has joined #openstack-swift23:23
*** kei_yama has joined #openstack-swift23:27
*** joeljwright has quit IRC23:27
*** geaaru has quit IRC23:28
*** esker has quit IRC23:39
*** esker has joined #openstack-swift23:41
*** zhill has quit IRC23:42
notmynameho: correct23:43
*** esker has quit IRC23:46
*** breitz1 has quit IRC23:51

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