Friday, 2014-03-21

« Thursday, 2014-03-20 Index Saturday, 2014-03-22 »
*** saschpe has quit IRC00:00
openstackgerritSamuel Merritt proposed a change to openstack/swift: Move patch_policies in obj server tests  https://review.openstack.org/8194500:02
*** saschpe has joined #openstack-swift00:07
*** RK has joined #openstack-swift00:08
*** zaitcev has quit IRC00:25
*** shri has quit IRC00:33
openstackgerritSamuel Merritt proposed a change to openstack/swift: tiny stylistic fixup  https://review.openstack.org/8195100:35
*** sungju- has quit IRC00:43
*** sungju_ has joined #openstack-swift00:44
openstackgerritA change was merged to openstack/python-swiftclient: Fix temporary pypy gate issue with setuptools  https://review.openstack.org/8147900:51
*** d89 has joined #openstack-swift00:56
*** changbl has joined #openstack-swift01:26
*** d89 has quit IRC01:44
*** haomaiwa_ has quit IRC01:59
openstackgerritSamuel Merritt proposed a change to openstack/swift: Block X-Backend* in gatekeeper  https://review.openstack.org/8196402:00
*** d89 has joined #openstack-swift02:04
*** RK has quit IRC02:31
*** saschpe has quit IRC02:31
*** saschpe has joined #openstack-swift02:33
openstackgerritMadhuri Kumari proposed a change to openstack/swift: Added swift-container-info tool.  https://review.openstack.org/8196802:36
*** d89 has quit IRC02:48
*** madhuri__ has joined #openstack-swift02:48
*** d89 has joined #openstack-swift02:53
openstackgerritMadhuri Kumari proposed a change to openstack/swift: Added swift-container-info tool.  https://review.openstack.org/8196803:02
*** krtaylor has joined #openstack-swift03:11
*** madhuri__ has quit IRC03:18
*** RockKuo_iPad has joined #openstack-swift03:23
*** RockKuo_iPad has quit IRC03:31
*** RockKuo_iPad has joined #openstack-swift03:31
*** RockKuo_iPad has quit IRC03:31
*** RockKuo has joined #openstack-swift03:35
*** saurabh_ has quit IRC04:13
Anju1what is the usage of cdmi ??04:25
*** haomaiwa_ has joined #openstack-swift04:32
*** d89 has quit IRC04:49
*** d89 has joined #openstack-swift04:57
*** ppai has joined #openstack-swift05:06
*** haomaiwa_ has quit IRC05:11
*** haomaiwa_ has joined #openstack-swift05:11
*** haomaiw__ has joined #openstack-swift05:22
*** haomaiwa_ has quit IRC05:25
*** nshaikh has joined #openstack-swift05:26
*** gyee has quit IRC05:30
*** d89 has quit IRC05:55
*** RockKuo has quit IRC06:11
*** d89 has joined #openstack-swift06:13
*** psharma has joined #openstack-swift06:20
*** saurabh_ has joined #openstack-swift06:22
*** saurabh_ has joined #openstack-swift06:22
openstackgerritA change was merged to openstack/swift: Make backend container requests use the same X-Timestamp  https://review.openstack.org/8115806:25
*** sungju_ has quit IRC06:31
*** d89 has quit IRC06:48
openstackgerritA change was merged to openstack/swift: Allow specification of object devices for audit  https://review.openstack.org/7588106:48
openstackgerritA change was merged to openstack/swift: Fix changing of storage policy index  https://review.openstack.org/7253606:54
*** d89 has joined #openstack-swift07:06
*** RockKuo_iPad has joined #openstack-swift07:15
*** RockKuo_iPad has quit IRC07:19
*** RockKuo has joined #openstack-swift07:38
openstackgerritJenkins proposed a change to openstack/swift: Updated from global requirements  https://review.openstack.org/8120707:52
*** mlipchuk has joined #openstack-swift08:08
openstackgerritA change was merged to openstack/swift: Purge self.policies from Swift  https://review.openstack.org/8194008:16
openstackgerritA change was merged to openstack/swift: tiny stylistic fixup  https://review.openstack.org/8195108:16
openstackgerritA change was merged to openstack/python-swiftclient: Copy Swift's .mailmap to swiftclient repo  https://review.openstack.org/8102708:16
openstackgerritA change was merged to openstack/swift: Move patch_policies in obj server tests  https://review.openstack.org/8194508:23
*** mmcardle has joined #openstack-swift08:33
*** RockKuo has quit IRC08:44
*** sphoorti has joined #openstack-swift08:50
*** nacim has joined #openstack-swift08:56
*** sphoorti has quit IRC08:58
*** sayalilunkad has joined #openstack-swift09:00
sayalilunkadHey, I want to upload a .jpeg file to a swift container using python swift client. The file gets uploaded but I don't see any contents in the file. I am using this command:  put_object(container, obj, contents, content_length=None)ΒΆ09:02
sayalilunkadSo in this the contents field I am not sure what I am supposed to use09:03
*** bvandenh has quit IRC09:08
*** bvandenh has joined #openstack-swift09:13
sayalilunkadchmouel: Hey09:26
sayalilunkadchmouel: Could you please help with the query posted09:27
tristanCsayalilunkad: if you put content_length = None and the contents is a file like interface (with a read method) it will perform a http chunked upload09:33
tristanC(ie: content=open("your_file.jpeg"), content_length=None)09:34
sayalilunkadtristanC: Thanks! it works:)09:35
*** sayalilunkad has quit IRC09:37
*** psharma has quit IRC09:39
openstackgerritTristan Cacqueray proposed a change to openstack/python-swiftclient: Add requests related unit-tests  https://review.openstack.org/7432809:41
*** sayalilunkad has joined #openstack-swift09:44
sayalilunkadtristanC: Also how can I download an .jpeg object after I get it using the get_object command?09:46
*** saju_m has joined #openstack-swift09:55
tristanCsayalilunkad: yes, using the same url/container name/object name09:56
sayalilunkadtristanC: Alright, will check that09:57
*** sayalilunkad has quit IRC10:00
*** tanee-away is now known as tanee10:17
*** tanee is now known as tanee-away10:17
*** RockKuo has joined #openstack-swift10:32
*** Gooch has quit IRC10:33
*** RockKuo has quit IRC11:00
*** saju_m has quit IRC11:22
*** Trixboxer has joined #openstack-swift11:30
*** d89 has quit IRC11:32
*** mlipchuk has quit IRC11:35
*** saju_m has joined #openstack-swift11:38
*** nshaikh has quit IRC12:12
*** dmsimard has joined #openstack-swift12:14
*** nshaikh has joined #openstack-swift12:24
*** JuanManuelOlle has joined #openstack-swift12:26
*** russellb is now known as rustlebee12:40
*** ppai has quit IRC12:51
*** saju_m has quit IRC12:57
*** mkollaro has joined #openstack-swift13:00
*** mmcardle has quit IRC13:04
*** saju_m has joined #openstack-swift13:06
*** tdasilva has joined #openstack-swift13:15
*** jasondotstar has joined #openstack-swift13:31
*** mmcardle has joined #openstack-swift13:43
*** byeager has joined #openstack-swift14:07
*** piyush has joined #openstack-swift14:31
*** mrsnivvel has quit IRC14:37
*** pberis has quit IRC14:50
*** mrsnivvel has joined #openstack-swift14:52
openstackgerritpaul luse proposed a change to openstack/swift: Container Confict on obj update with wrong Storage-Policy-Index  https://review.openstack.org/7973114:54
peluse^ rebase14:54
*** saju_m has quit IRC15:05
openstackgerritpaul luse proposed a change to openstack/swift: Add Storage Policy Support to Account HEAD  https://review.openstack.org/7374715:17
openstackgerritpaul luse proposed a change to openstack/swift: Container Confict on obj update with wrong Storage-Policy-Index  https://review.openstack.org/7973115:22
*** mrsnivvel has quit IRC15:24
openstackgerritpaul luse proposed a change to openstack/swift: Add Storage Policy Support to Account HEAD  https://review.openstack.org/7374715:24
pelusetorgomatic:  your test_set_storage_policy_index* tests were using bogus indexes (on purpose I assume) however with the two patches above they're checked via the pending file encoding that was added so I updated the tests to use valid patched policies.15:27
*** jasondotstar has quit IRC15:33
*** byeager has quit IRC15:35
*** byeager has joined #openstack-swift15:36
*** byeager has quit IRC15:40
*** nshaikh has quit IRC15:41
*** byeager has joined #openstack-swift15:44
*** gyee has joined #openstack-swift16:01
*** Longgeek_ has joined #openstack-swift16:21
*** Longgeek_ is now known as Longgeek16:27
notmynamegood morning, everyone16:30
pelusegood morning!16:30
portantenotmyname: mornin'16:31
notmynameclose my computer and don't open it until the next day at the office? that'll be 201 unread emails16:31
peluseit's a beautiful day!16:31
portantegood luck!16:31
*** pberis has joined #openstack-swift16:32
creihtmark all as read, the day is done :)16:32
*** mkollaro has quit IRC16:34
pelusemake the 202, just sent you an email :)16:34
notmynamepeluse: I'll send you a text to check IRC where I'll mention that I replied to the email16:35
*** jogo is now known as flashgordon16:39
*** jamie_h has joined #openstack-swift16:45
*** jasondotstar has joined #openstack-swift16:48
pelusenotmyname:  excellent and I'll tweet something out once I see it :)16:48
notmynameperfect.16:53
openstackgerritChuck Thier proposed a change to openstack/swift: Add "If-None-Match: *" support to PUT  https://review.openstack.org/8164616:55
*** zaitcev has joined #openstack-swift17:00
*** ChanServ sets mode: +v zaitcev17:00
*** byeager has quit IRC17:02
*** byeager has joined #openstack-swift17:02
*** byeager has quit IRC17:07
*** byeager has joined #openstack-swift17:11
*** nacim has quit IRC17:19
*** mkollaro has joined #openstack-swift17:22
*** piyush has quit IRC17:26
*** saju_m has joined #openstack-swift17:38
*** sriram has joined #openstack-swift17:47
*** saju_m has quit IRC17:58
*** judd7_ has joined #openstack-swift18:01
*** piyush has joined #openstack-swift18:15
*** piyush1 has joined #openstack-swift18:16
*** piyush has quit IRC18:19
*** jamie_h has quit IRC18:23
*** lpabon has joined #openstack-swift18:26
*** mmcardle has quit IRC18:31
openstackgerritpaul luse proposed a change to openstack/swift: Add Storage Policy Support to Account HEAD  https://review.openstack.org/7374718:37
openstackgerritSamuel Merritt proposed a change to openstack/swift: Merge branch 'master' into feature/ec  https://review.openstack.org/8216518:42
*** iccha has joined #openstack-swift18:48
icchahey i have some queries wrt python swiftclient18:49
zackmdavisiccha, what about it?18:51
icchazackmdavis: we upgraded to using python-swiftclient-2.0.3 and have been seeing WARNING requests.packages.urllib3.connectionpool [-] HttpConnectionPool is full, discarding connection18:51
icchathis is glance using swiftclient to talk to swift18:52
icchawondering if anything changed18:52
icchaor if there is a setting for this anywhere18:52
zackmdavisiccha, sorry, I haven't run into that and don't have anything useful to say18:53
*** dmsimard has quit IRC18:56
zackmdavisiccha, well, it looks like 2.0 (last month) switched to using Requests instead of httplib, so that would be the clear candidate for something-that-changed https://github.com/openstack/python-swiftclient/commit/19d7e1812a99d73785146667ae2f3a7156f0689818:57
icchazackmdavis: http://stackoverflow.com/questions/18466079/can-i-change-the-connection-pool-size-for-pythons-requests-module18:57
icchalooks like there is some setting in requests18:57
icchamaybe a patch in swift client to increase or customize pool size19:06
icchaadapter = requests.adapters.HTTPAdapter(pool_connections=100, pool_maxsize=100)19:06
openstackgerritGreg Lange proposed a change to openstack/swift: Fix account/container logging mismatch  https://review.openstack.org/8217419:16
claygpeluse: is there some change for feature/ec that's waiting to merge and isn't merged?19:23
*** NM1 has joined #openstack-swift19:23
claygnm, I think i found it19:23
NM1Cool!19:24
NM1But found what?19:24
peluseclayg:  only the 409 thing and the acct rollup that I know of19:24
claygpeluse: yessir19:24
claygpeluse: what about the un-.policies-ification?19:25
pelusethought I saw that go through.... maybe not19:25
* peluse looks19:25
peluseyeah, its merged19:26
pelusetorgotmatic has a few others though I think also, all reconcliator-ifier...19:26
icchazackmdavis: https://bugs.launchpad.net/python-swiftclient/+bug/129581219:27
claygstupid swiftclient 2.019:29
iccha+1 clayg19:29
icchalol19:29
claygtorgomatic: peluse: I'm trying to get the container conflict thingy merged - but I"m still getting unittest failures in proxy server setup that says prosrv.polices <- AttributeError19:33
claygfeature/ec does show purge self.polices is merged...19:33
claygi tried to rebase on that...19:33
peluseI think there's an active rebase msater-->feature EC right now19:33
claygso strange, I just checked out feature/ec and sure enough it says prosrv.policies?19:34
torgomaticpeluse: clayg: merge is in the gate now, and I think it fixes that19:34
pelusehttps://review.openstack.org/#/c/82165/19:34
claygweird19:35
peluseconfirmed it does address that change19:35
peluseclayg:  container conflict thingy - I assume the one I updated this morning right?  Did you look at the changes made based on your comments yet?  Note that it's rebased as of my last push19:37
claygpeluse: no you're talking about the account rollup yeah?19:38
claygi started to look at that - and it depends on the container-obj-put-409 change - which I'm trying to get merged/tested/approved!19:38
peluseclayg:  both actually, sorry.  I rebased the 409 deal also but didn't make addtl changes beyond that.  Yes wrt the comments I mean the acct rollup, sorry19:38
peluse:)19:38
claygyeah i'm still looking at the latest changes to acct-rollup - it did look like you addressed most of the early stuff I spotted - thanks - will continue review19:39
pelusegracias.  need me to do anything on the 409 patch?19:40
claygpeluse: don't think so - i rebased it on the pending merge and I'm giving the diff a final once over test run19:41
peluserock n roll!19:41
peluseclayg:  when you're ready we can talk about the policy index header replacement scheme, might have an idea for making that more efficient but will leave the patch alone for now til you review the other changes19:43
clayg.... which part now?  policy ohoh oh oh - that - yeah I mean i don't think it's gunna get worse :P19:44
claygpeluse: I think the acct-rollup change may need some other stuff anyway - and I"m not looking at it for another 10 mins or - so if you had something in mind you wanna try out you can push it in the diff and I'll poke at it19:44
claygI don't have anything in mind that I'm just *sure* woudl be better - but I was planning on playing with it and seeing if anything I like jumps out19:45
pelusecool, I'll dork wiht it a bit and see if anything pans out...19:45
claygno worries - we can also merge acct-rollup with that part like that and worry about it later, but it was ok before and sorta not great with the acct-rollup changes so I think it's worth a bit of our time19:46
claygummm there's a fixme in container-obj-put-409 I'm going to try and cleanup19:48
pelusethere is?  something you added??19:50
claygyeah just something i noticed playing around with tests19:51
*** NM1 has quit IRC19:51
peluseahhh, ok19:52
*** NM1 has joined #openstack-swift19:53
clayghugokuo: can you link me the standalone script that spings up keystone?  I've been wanting to take a stab at adding optional support for keystone to vagrant-swift-all-in-one...19:56
claygI mean I get by with my devstack setup, but I'm keen on having something backed into that tool, but a little leary of maintaining it in the face of updates to keystone19:57
claygi scripted a keystone setup once by hand - it wasn't that fun19:57
claygpeluse: thanks for writing StoragePolicyCollection.__iter__ it's just what I wanted!19:59
*** NM1 has joined #openstack-swift20:02
acorwinclayg: i think this is what you want https://github.com/swiftstack/keystone_install20:05
*** saju_m has joined #openstack-swift20:05
claygnice20:06
*** judd7_ has quit IRC20:12
*** seandrummond has joined #openstack-swift20:16
NM1Does anyone know how to set write permissions per container? I tried "curl -v -XPOST -H "X-Storage-Token: AUTH_xxx" -H "X-Container-Write: myuser" http://localhost:8081/v1/AUTH_swift/mycontainer/"20:16
*** shri has joined #openstack-swift20:16
seandrummondHey guys since release of swift version 1.11.0 I should not be seeing 'X-If-Delete-At and X-Delete-At do not match' Correct?20:16
peluseclayg: that would have been torgomatic for that particlar method20:16
clayggit blame says you - you'll just have to take credit20:17
claygNM1: that looks pretty reasonable - does the server accept the request?  what's the response code?20:17
NM1HTTP/1.1 204 No Content - But if I try to request a object with the user I got a 40320:18
claygseandrummond: you could see those anytime, but it would mean that someone has an ojbect that is having it's x-delete-at updated frequently, and the .expiring_objects container listing is racing with the deletes20:19
claygNM1: does the X-Container-Write header get set on the container metadata tho?20:19
claygand when you say "try to request a object" you mean try and PUT and object yeah?  read/write permissions are seperate - it's like ftp servers20:19
seandrummondclayg: once the object gets upload it is not changed. The objects are getting uploaded with a header x-delete-at set with a timestamp for 36 hours from the time it gets uploaded20:20
clayghrmm... then you shouldn't really be seeing those messages - and you getting a lot of them?20:21
seandrummondyes in fact it appears nothing is getting deleted and I am seeing the same issue reported in https://bugs.launchpad.net/swift/+bug/125733020:21
*** jasondotstar has quit IRC20:21
seandrummondhowever I have verified I am running 1.11.020:21
NM1The container has:20:22
NM1X-Container-Write: entretenimentocma20:22
NM1< X-Timestamp: 1395352461.8727620:22
NM1< X-Container-Read: entretenimentocma20:22
*** openstackgerrit has quit IRC20:22
*** openstackgerrit has joined #openstack-swift20:23
NM1I cant request the object using the GET method using curl20:23
*** tdasilva has left #openstack-swift20:23
*** samwright has joined #openstack-swift20:25
claygseandrummond: yeah fix should be in 1.11 - I've verified the expirer recently20:25
claygseandrummond: the patch is pretty trivial - you might track down one of the servers that returned the 4XX to the expirer and check swift/common/exceptions for a DiskFileExpired exception20:26
clayghttps://review.openstack.org/#/c/59811/4/swift/common/exceptions.py,unified <- should look like that seandrummond20:26
*** lpabon has quit IRC20:27
claygNM1: so either the token you're using with the GET request doesn't associate the name "entretenimentocma" with the groups for that GET request - or the auth system you're using doesn't support container acls?20:27
seandrummondyup that is what I verified and I just did again20:27
NM1The token is new. I'm using Temp Auth.20:29
claygso it's possible you have some dark matter left around while the bug was effecting you (see gholt's script for help cleanup, it never got fixed in the auditor, although we should probably still take a stab tat that)20:29
claygseandrummond: but that wouldn't explain you still seeing it in the logs after you restarted all the daemons/servers with the new code20:30
openstackgerritMichael Barton proposed a change to openstack/swift: Sendfile acceleration for GET requests  https://review.openstack.org/8218620:30
claygNM1: oh then you need to either use the account name or include the full name of the user (e.g. "test:tester" works, or just "test" for all users in that account, but you can't say just "tester")20:30
claygwell you can say just "tester" but then it doesn't work like you want it to ;)20:31
*** sriram has quit IRC20:31
seandrummondclayg: I can assure you they have all been restarted and I have my own clean up scripts that delete the data, similar to gholt's script20:31
seandrummond[root@host swift]# grep -c "\ 412\ " object-server.log20:31
seandrummond19816620:31
seandrummondwhile straceing I can see the reason for the 412 I cannot see if output in the log itself20:32
NM1clayg:  Do you think it mught be a bug from a old version? :( I'm using swift 1.720:33
claygNM1: no i'm pretty sure you have the wrong "user" in the header?  what's the full tempauth "acct:username" for that user?20:34
seandrummondcould it be obj/server.py:608 ? sorry clayg I realize you are in the middle of two conversations I can speak with you later if you prefer20:34
claygseandrummond: no worries NM1 will get it figured out20:34
NM1plataforma:entretenimentocma20:35
*** jasondotstar has joined #openstack-swift20:35
*** saju_m has quit IRC20:35
claygyeah i mean that's the time you *want* it to return 412 tho, like I said they could come up anytime - but they shouldn't if you're not updating an objects X-Delete-At20:35
claygNM1: so you need to use that in your x-container-write/read headers20:35
claygseandrummond: so i'm remembering now that the bug there was the object server returning 404 for objects that had expired and not acctually deleting the .data file on disk - which the expirer would just assume was fine and delete the record in the expiring object acct20:38
NM1clayg:  It worked like magic.20:38
openstackgerritA change was merged to openstack/swift: Merge branch 'master' into feature/ec  https://review.openstack.org/8216520:39
claygnot magic - apis20:39
NM1Thank you. I lost all my afternoon with that...20:39
seandrummondclayg: I am also getting many 404s but I was thinking that was from running multiple expirers20:39
claygseandrummond: ok that one makes sense then :P20:39
claygseandrummond: hrmm... the bug report specifically mentions the 412 - but that doens't make any sense in light of the patch?20:40
seandrummondand it seems nothing is getting cleaned up20:40
*** jasondotstar has quit IRC20:41
*** JuanManuelOlle has quit IRC20:48
*** Trixboxer has quit IRC20:54
peluseclayg:  so I think I have something better for the policy index replacement however I have a mtg now that will waste my time - err keep me occupied I mean - for a at least an hr20:55
peluseclayg/torgomatic:  sidenote it looks like there's a missing part from the 'purge policies' patch that was just merged.  _convert_policy_to_index() needs the proxy application class .policies member and even unit tests fail for me, how did it get merged? Or am I in need or more coffee (quite possibly)?20:57
torgomaticpeluse: could be that I missed a spot; what unit test fails? test/unit/proxy/test_server.py?20:57
seandrummondclayg:  I am going to continue to look into it I will touch base again later possibly Monday20:58
pelusetorgomatic:  whoa, looking on jenkins output for the patch it fails also...20:58
pelusehttp://logs.openstack.org/40/81940/1/check/gate-swift-python26/5539ae0/nose_results.html20:59
torgomaticpeluse: so the tests failed, but the change merged anyway?20:59
peluselooks like Jenkins says it passed even though the detail showed the failure, m aybe because it wasn't an explciit test? ex: ERROR: test suite for <module 'test.unit.proxy.test_server'20:59
torgomaticpeluse: okay, I have the same failure locally... let me see what I can do21:00
pelusetorgomatic: cool, thanks21:00
openstackgerritSamuel Merritt proposed a change to openstack/swift: Fix unit tests.  https://review.openstack.org/8220721:07
*** seandrummond has quit IRC21:09
*** samwright has left #openstack-swift21:13
openstackgerritSamuel Merritt proposed a change to openstack/swift: Allow overriding of storage policy index.  https://review.openstack.org/8221121:16
openstackgerritSamuel Merritt proposed a change to openstack/swift: Put X-Timestamp in object 404 responses  https://review.openstack.org/8072921:16
openstackgerritSamuel Merritt proposed a change to openstack/swift: Make the container replicator notice misplaced objects  https://review.openstack.org/8073121:16
openstackgerritSamuel Merritt proposed a change to openstack/swift: Add object-reconciler daemon  https://review.openstack.org/8073021:16
openstackgerritSamuel Merritt proposed a change to openstack/swift: Make object updater reconcile on 409s  https://review.openstack.org/8073221:16
*** byeager has quit IRC21:27
*** byeager has joined #openstack-swift21:27
*** piyush1 has quit IRC21:28
*** piyush has joined #openstack-swift21:28
*** piyush1 has joined #openstack-swift21:30
*** piyush has quit IRC21:32
* peluse thinks torgomatic must have a tommy gun that shoots patches :)21:34
torgomaticnow all I need is one that shoots quality :)21:34
*** byeager has quit IRC21:34
*** byeager has joined #openstack-swift21:35
*** byeager has quit IRC21:39
*** pberis has quit IRC21:48
*** mkollaro has quit IRC21:53
*** byeager has joined #openstack-swift21:57
*** piyush1 has left #openstack-swift21:59
portantenotmyname: what was the name of that flow tool your wrote?22:02
notmynameI wrote a flow tool?22:03
notmynametell me more22:03
notmyname(and I might remember it)22:03
portantelog flow, looks at swift logs and makes a graph?22:03
notmynameoh yeah22:04
portantegoflow or something?22:04
notmynameand look at that. I have a directory called "flow". that must be it :-)22:04
portantelogflow22:04
portantegot it22:04
portantehttps://github.com/notmyname/logflow22:04
notmynameportante: ya https://github.com/notmyname/logflow22:04
notmynameportante: you wanna push some code there? :-)22:05
portantewe want to use it to help somebody learn what is happening and why it is breaking22:05
notmynamecool22:05
notmynamenext steps I had for it were to somehow designate a particular set of log lines (maybe log file name?) as belonging to a particular server. that way you can actually see which servers (and how many) are involved in each task22:06
notmynameI'll do that in my free time22:06
portanteyes, that would be cool22:06
notmynameso maybe next August22:06
notmyname(next == 2015)22:07
notmynameportante: but I'd be happy to give you write access to that repo.22:07
notmynamenow that you reminded me of it (I seriously had forgotten), maybe I'll take another look22:07
notmynameit's a good "on the airplane" project22:08
portanteyes22:09
portanteand I am going to try to use it tonight to help illuminate a problem22:09
notmynamecool22:09
portanteboth as a teaching instrument and to help debug22:09
notmynamewas that yes to working on the airplane or yes to commit acces?22:09
portanteyes on the airplane, if I make changes, I'll post a pull request first, and then you can decide if I should have write access to the repo. ;)22:10
*** byeager has quit IRC22:10
notmynameheh. like you don't have commit access to other projects? you know, like swift? ;-)22:10
portante:)22:10
portantealready too dangerous22:11
notmynameheh. ok. I'll be happy to merge your PR22:11
openstackgerritMatthew Treinish proposed a change to openstack/swift: Stop using use_develop with tox  https://review.openstack.org/8222822:13
mtreinishnotmyname: ^^ that should fix the unit test issue22:13
notmynamehmm...seems like that project is not following the swift naming pattern. it should be swogflow. or slogflow. or just slow22:13
mtreinishclarkb: ^^^22:13
notmynamemtreinish: ah, cool22:13
clarkbmtreinish: is it usedevelop that breaks it?22:14
mtreinishclarkb: yeah I ran it locally with and without it22:14
notmynamehmm...added back in 200922:14
mtreinishI have no idea why22:14
clarkbinteresting. I wonder if that is a setuptools bug22:14
clarkbmtreinish: you should bug mordred about it22:14
mtreinishclarkb: I'll save that for next week22:14
notmynamemtreinish: https://gist.github.com/notmyname/e851249703427e3d9bdc22:15
openstackgerritpaul luse proposed a change to openstack/swift: Add Storage Policy Support to Account HEAD  https://review.openstack.org/7374722:18
openstackgerritpaul luse proposed a change to openstack/swift: Container Confict on obj update with wrong Storage-Policy-Index  https://review.openstack.org/7973122:18
peluseclayg:  the new acct rollup patch has a better way of doing the policy index/name thing but probably still needs another round... get/head for all requsts is no longer impacted and the gymnastics of fixing up the acct HEAD request are less22:19
openstackgerritMatthew Treinish proposed a change to openstack/swift: Stop using use_develop with tox  https://review.openstack.org/8222822:19
peluseclayg: and not that I didn't add unit test coverage for the new strip function pending your input on a better way to skin the cat22:21
claygpeluse: ah yeah cool - sorry I got all hung up making some of these container tests awesome22:30
pelusecool... BTW, not sure why the 409 patch got updated when I pushed the acct rollup change.. something to do with the dependency no doubt but wasn't expected22:33
claygpeluse: ok, but you didn't put any changes fixups there - it was just your general rebase mess?22:34
peluseI sure hope so :)  I didn't make any changes on purpose, no22:34
pelusechecking diff now...22:35
peluseclayg:  bunch of stuff I don't recognize so looks like it picked up the rebase from this morn22:38
peluseor its just all jacked to hell now :-o22:39
openstackgerritClay Gerrard proposed a change to openstack/swift: Container Confict on obj update with wrong Storage-Policy-Index  https://review.openstack.org/7973122:43
clayg^ that's all rebased up on fix unittests - but probably blew the account roll up one22:43
claygI need to checkout that branch to review it - so I'll do the rebase now if that's cool22:44
claygshit, nm that was the wrong branch :P22:45
* clayg really hates the branch names that you get with git review -d22:46
notmynameclayg: step 2 of `git review -d` is `git branch -m readable`22:46
*** jasondotstar has joined #openstack-swift22:47
pelusecool22:56
*** jasondotstar has quit IRC22:57
*** piousbox has joined #openstack-swift23:00
piousboxI have an architectural question regarding the Swift API23:00
piousboxWhy is the account in the URL?23:00
torgomaticwhere else would it be?23:00
piousboxHere, for example: http://docs.openstack.org/api/openstack-object-storage/1.0/content/GET_showAccountDetails_v1__account__storage_account_services.html23:00
piousboxWell, account information is passed in the headers.23:01
torgomaticdepending on your auth system, one user may have access to multiple Swift accounts23:01
torgomaticif nothing else, admins can access any account23:02
piousboxand if you want to prevent a used from accessing any account?23:02
piousboxYou still have to specify account information elsewhere, correct?23:02
claygtorgomatic: back away slowly23:02
piousboxused=user23:02
claygi think the best reason for having the fully qualified resource in the url is for shared access, and the fact that two users can have a container with the same name23:03
torgomaticso, the user sends the X-Auth-Token header, and the auth system determines if they have access to the named resource23:03
claygif you a) don't allow users to name their resources or b) don't ever intend for resources to be public - there's little reason to include the owner of the resources in the url - otherwise - you sort of have too23:03
torgomaticthe first is identity, the second is the resource23:04
piousboxAmazon S3 does not have account in their path, only bucket name.23:05
piousboxHaving containers with the same names and different users makes sense.23:05
claygthey can't have multiple uses with the same container/bucket name23:05
claygand knowning is half the battle23:05
piousboxHm okay, thanks.23:06
claygnp!23:07
*** jasondotstar has joined #openstack-swift23:11
*** seandrummond has joined #openstack-swift23:12
*** seandrummond has quit IRC23:14
*** NM1 has quit IRC23:36
*** jasondotstar has quit IRC23:43
openstackgerritClay Gerrard proposed a change to openstack/swift: Container Confict on obj update with wrong Storage-Policy-Index  https://review.openstack.org/7973123:47
openstackgerritClay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to Account HEAD  https://review.openstack.org/7374723:47
*** NM1 has joined #openstack-swift23:56
*** NM1 has quit IRC23:57
*** NM1 has joined #openstack-swift23:59
« Thursday, 2014-03-20 Index Saturday, 2014-03-22 »

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