Tuesday, 2014-05-20

*** igor_ has joined #openstack-swift00:00
*** igor_ has quit IRC00:04
*** kenhui has quit IRC00:14
*** fbo has quit IRC00:18
*** kenhui has joined #openstack-swift00:20
*** fbo has joined #openstack-swift00:28
*** tdasilva has joined #openstack-swift00:33
*** occupant has joined #openstack-swift00:34
*** mmcardle has joined #openstack-swift00:47
*** hipster has joined #openstack-swift00:47
*** shri has quit IRC00:49
*** mmcardle has quit IRC00:51
*** hipster has quit IRC00:51
*** igor_ has joined #openstack-swift01:00
*** Diddi has quit IRC01:02
*** saschpe has quit IRC01:04
*** Diddi has joined #openstack-swift01:05
*** igor_ has quit IRC01:05
*** saschpe has joined #openstack-swift01:06
*** kenhui has quit IRC01:22
*** mattoliverau has quit IRC01:30
*** mattoliverau has joined #openstack-swift01:42
*** nosnos has joined #openstack-swift01:47
*** hipster has joined #openstack-swift01:47
*** mmcardle has joined #openstack-swift01:48
*** foexle_ has joined #openstack-swift01:49
*** hipster has quit IRC01:51
*** foexle has quit IRC01:52
*** mmcardle has quit IRC01:52
*** igor_ has joined #openstack-swift02:01
*** saschpe has quit IRC02:01
*** saschpe has joined #openstack-swift02:03
*** igor_ has quit IRC02:05
*** erlon has quit IRC02:09
*** kenhui has joined #openstack-swift02:22
*** bill_az has quit IRC02:27
*** kenhui has quit IRC02:31
*** hipster has joined #openstack-swift02:39
*** foexle_ has quit IRC02:40
*** hipster has quit IRC02:44
*** hipster_ has joined #openstack-swift02:46
*** mmcardle has joined #openstack-swift02:49
*** mmcardle has quit IRC02:53
*** igor_ has joined #openstack-swift03:02
*** nosnos has quit IRC03:07
*** igor_ has quit IRC03:07
*** hipster_ has quit IRC03:23
*** kenhui has joined #openstack-swift03:27
*** madhuri has joined #openstack-swift03:31
*** kenhui has quit IRC03:31
*** hipster has joined #openstack-swift03:32
madhuriwhy is the ring data kept in gzip format? Is there any specific reason for it?03:33
*** bill_az has joined #openstack-swift03:36
*** bill_az has quit IRC03:42
*** nosnos has joined #openstack-swift03:47
*** mmcardle has joined #openstack-swift03:50
*** gyee has quit IRC03:54
*** mmcardle has quit IRC03:54
*** hipster has quit IRC03:57
*** kenhui has joined #openstack-swift03:58
*** nosnos has quit IRC04:01
*** kenhui has quit IRC04:02
*** ppai has joined #openstack-swift04:11
*** kenhui has joined #openstack-swift04:12
*** hipster has joined #openstack-swift04:27
*** hipster_ has joined #openstack-swift04:29
*** nosnos has joined #openstack-swift04:30
*** kenhui has quit IRC04:31
*** hipster has quit IRC04:33
*** hipster_ has quit IRC04:33
*** odyssey4me has quit IRC04:33
*** psharma has joined #openstack-swift04:35
*** haomaiwang has joined #openstack-swift04:41
*** zaitcev has quit IRC04:47
*** mmcardle has joined #openstack-swift04:52
*** mmcardle has quit IRC04:57
*** nshaikh has joined #openstack-swift05:04
*** igor_ has joined #openstack-swift05:04
*** hipster has joined #openstack-swift05:07
*** Midnightmyth has joined #openstack-swift05:07
*** igor_ has quit IRC05:08
*** hipster has quit IRC05:12
*** hipster has joined #openstack-swift05:47
*** hipster has quit IRC05:52
*** mmcardle has joined #openstack-swift05:53
*** mmcardle has quit IRC05:57
*** igor has joined #openstack-swift06:04
*** kenhui has joined #openstack-swift06:06
*** igor has quit IRC06:10
*** kenhui has quit IRC06:11
*** igor has joined #openstack-swift06:24
*** igor has quit IRC06:26
*** Midnightmyth has quit IRC06:36
*** hipster has joined #openstack-swift06:47
*** hipster has quit IRC06:52
*** mmcardle has joined #openstack-swift06:54
*** mmcardle has quit IRC06:59
*** h6w has joined #openstack-swift07:00
*** kenhui has joined #openstack-swift07:07
*** kenhui has quit IRC07:11
*** foexle has joined #openstack-swift07:12
*** linuxbaby has joined #openstack-swift07:12
linuxbabyhi.. i am backing up a siwft node for backup and restore in tripleo method of installation07:13
linuxbabymy siwft node is a bare metal server .. what should i backup to restore the node in case it breaks07:13
linuxbabyfrom a swift node07:13
*** odyssey4me has joined #openstack-swift07:19
*** madhuri has quit IRC07:20
*** igor has joined #openstack-swift07:28
*** sungju has quit IRC07:31
*** igor has quit IRC07:32
*** miqui__ has joined #openstack-swift07:32
*** mmcardle has joined #openstack-swift07:36
*** miqui_ has quit IRC07:37
*** omame has quit IRC07:37
*** odyssey4me has quit IRC07:40
*** omame has joined #openstack-swift07:43
*** mlipchuk has joined #openstack-swift07:58
*** kenhui has joined #openstack-swift08:07
*** kenhui has quit IRC08:12
*** bvandenh has joined #openstack-swift08:16
*** MooingLemur has quit IRC08:18
*** MooingLemur has joined #openstack-swift08:19
*** mkollaro has joined #openstack-swift08:19
*** ccorrigan has quit IRC08:24
*** jamie_h has joined #openstack-swift08:27
hugokuolinuxbaby: hmm... fstab, /etc/swift/* , rsync configurations ,08:28
hugokuolinuxbaby: I think those are all you need ....08:28
*** igor has joined #openstack-swift08:28
linuxbabyhugokuo, yes i got this list : /etc/rsyncd.conf /etc/swift08:30
linuxbabyhugokuo, what about memcached ?08:30
hugokuolinuxbaby: if you are running it with default setting, then you can ignore it. Unless there's some custom changes in your memcached configurations...08:31
linuxbabyhugokuo, okies.. thank you..08:32
*** igor has quit IRC08:33
*** mandarine has joined #openstack-swift08:35
mandarineHello everyone, it's been a while :)08:35
mandarineIt seems I have a problem I did not have before : swift services take a looong time before stopping08:37
mandarineAnyone has had the same problem before ?08:37
hugokuohmm... using strace ?08:39
*** mmcardle has quit IRC08:40
*** acoles_away is now known as acoles08:43
*** mmcardle has joined #openstack-swift08:45
*** ccorrigan has joined #openstack-swift08:48
*** kenhui has joined #openstack-swift09:08
*** mkollaro has quit IRC09:11
*** kenhui has quit IRC09:12
*** mkollaro has joined #openstack-swift09:14
acolescschwede_: ping09:22
*** igor has joined #openstack-swift09:24
*** igor has quit IRC09:28
openstackgerritChristian Berendt proposed a change to openstack/swift: replace string format arguments with function parameters  https://review.openstack.org/9433109:40
*** hipster has joined #openstack-swift09:47
*** hipster has quit IRC09:51
*** igor has joined #openstack-swift09:59
pconstantinewhere do I file bug for check-swift-dsvm-functional tests?10:02
cschwede_acoles: Hi Alistair!10:05
acolescschwede_: hi - recovered from jet lag?10:05
cschwede_acoles: still trying ;) And you? Had a safe flight back?10:05
acolescschwede_: yes, not too bad thanks10:06
acolescschwede_: got a question re swiftclient tests...10:06
cschwede_acoles: on the python3 functional testing?10:06
acolescschwede_: no, a different one...:) when func tests were added and unit tests were moved under tests/unit, test_shell.py got left in tests/ - was that deliberate?10:07
acolescschwede_: they don't run with tox10:07
cschwede_no, that wasn’t intended. hmm10:08
acolescschwede_: ok. i'll fix.10:09
acolescschwede_: they do pass py33 when i move them :)10:09
cschwede_acoles: thanks, and I’ll fix the py3 functional testing10:09
cschwede_acoles: that’s great :)10:09
acolescschwede_: thx. be great to have those func tests working too10:11
openstackgerritAlistair Coles proposed a change to openstack/python-swiftclient: Move test_shell.py under tests/unit/  https://review.openstack.org/9433510:25
acolescschwede_: ^^10:25
cschwede_acoles: thanks, looking!10:26
*** kenhui has joined #openstack-swift10:32
*** kenhui has quit IRC10:36
*** haomaiwang has quit IRC10:39
*** haomaiwang has joined #openstack-swift10:39
*** haomaiw__ has joined #openstack-swift10:43
cschwede_acoles: just run tox without and with your patch, works as expected - added my +2. thanks for fixing this!10:45
*** haomaiwang has quit IRC10:45
acolescschwede_: no prob.10:45
*** hipster has joined #openstack-swift10:47
openstackgerritAlistair Coles proposed a change to openstack/python-swiftclient: Add keystone v3 auth support  https://review.openstack.org/9178810:51
*** hipster has quit IRC10:52
*** haomaiwang has joined #openstack-swift10:54
*** haomaiw__ has quit IRC10:55
*** bvandenh has quit IRC11:02
*** nmap911 has joined #openstack-swift11:03
nmap911Hi guys. maybe I am being dumb. Do you delete swift accounts (when swift is authenticating vs keystone) in the same manor as listed here? - https://www.swiftstack.com/docs/cookbooks/swift_usage/superusers.html#delete-a-swift-account11:06
ctennisyes nmap911, that's the way to do it regardless of authentication method11:14
nmap911thanks ctennis, when issueing the delete request, is there a different user I must authenticate with when deleting the account? other than the account I am deleting, I mean11:15
nmap911like, keystone admin account I take it?11:16
ctennisyes, someone who has admin priviledges within swift itself11:16
ctennisbe that a keystone admin, if so configured, or another administrative account11:16
nmap911Thanks a stack!11:18
*** dmorita has quit IRC11:22
*** kenhui has joined #openstack-swift11:32
*** kenhui has quit IRC11:37
*** nacim has quit IRC11:40
*** matsuhashi has quit IRC11:41
psharmahi, is there any tool/test to check amazon S3 API compatibility with my openstack-swift cluster, i used the swift3 middleware , and i was able to start my swift services11:43
*** nacim has joined #openstack-swift11:53
*** ophuk has joined #openstack-swift12:01
*** mkollaro1 has joined #openstack-swift12:04
*** mkollaro has quit IRC12:04
*** psharma has quit IRC12:05
*** tdasilva has left #openstack-swift12:14
openstackgerritChristian Schwede proposed a change to openstack/python-swiftclient: Fix Python3 bugs  https://review.openstack.org/9434712:22
cschwede_acoles: ^^ this should fix the python3 functional testing problems12:24
acolescschwede_: ok i'll take a look12:29
cschwede_acoles: thanks a lot!12:29
*** kenhui has joined #openstack-swift12:33
*** kenhui has quit IRC12:38
openstackgerritAlistair Coles proposed a change to openstack/python-swiftclient: Relax requirement for tenant_name in get_auth()  https://review.openstack.org/9435312:42
acoleschmouel: ^^12:42
*** mmcardle has quit IRC12:44
chmouelfyi: the gate is borked atm for the postgresql   check12:51
*** prassee has joined #openstack-swift12:58
prasseecould anyone give some insight about the back end implementation of swift12:59
*** nosnos has quit IRC13:02
glangewhat part ?13:04
*** pconstantine has quit IRC13:04
*** ppai has quit IRC13:04
*** miqui__ is now known as miqui_13:07
*** pconstantine_ has joined #openstack-swift13:07
*** mmcardle has joined #openstack-swift13:08
*** Ju has quit IRC13:09
*** nacim has quit IRC13:13
*** prassee has quit IRC13:15
*** tdasilva has joined #openstack-swift13:23
*** nacim has joined #openstack-swift13:26
*** kenhui has joined #openstack-swift13:28
*** hipster has joined #openstack-swift13:29
*** mkollaro has joined #openstack-swift13:31
*** hipster has quit IRC13:32
*** miqui has quit IRC13:32
*** mkollaro1 has quit IRC13:35
portanteprassee: when you return, I can help13:36
openstackgerritAlistair Coles proposed a change to openstack/python-swiftclient: Relax requirement for tenant_name in get_auth()  https://review.openstack.org/9435313:41
*** mrsnivvel has quit IRC13:42
*** kenhui has quit IRC13:44
*** lpabon has joined #openstack-swift13:45
*** Trixboxer has joined #openstack-swift14:03
*** nmap911 has quit IRC14:04
*** bill_az has joined #openstack-swift14:04
*** mattoliverau has quit IRC14:06
*** miqui_ is now known as miqui14:06
*** byeager has joined #openstack-swift14:07
*** bvandenh has joined #openstack-swift14:14
*** mkerrin has quit IRC14:15
gholtnotmyname: If you get the chance, can you help Constantine out here: https://review.openstack.org/#/c/72157/14:18
gholtnotmyname: Nevermind, glange just informed me of https://twitter.com/notmyname/status/468574627049267201 Ow! Rest up.14:24
omamea broken collar bone is kinda fun to play with for the first couple of days14:25
*** mattoliverau has joined #openstack-swift14:29
*** hipster has joined #openstack-swift14:39
cschwede_notmyname: oh my, get well soon!14:40
notmynamegholt: cschwede_: thanks14:44
notmynamegood morning world14:44
*** Midnightmyth has joined #openstack-swift14:44
ccorrigan:notmyname - get well soon, sorry to hear about your accident14:47
notmynamealso, ALWAYS WEAR A BIKE HELMET!!!14:47
gholtYeah, what's that about? How's your head?14:47
notmynamemy head hit the ground pretty hard, but since I was wearing a helmet nothing bad happened14:49
notmynamehad I not been...well...let's not think of that14:49
gholtOh, good!14:49
gholtYeah really.14:49
notmynameso today I want to try to get the specs repo finished setting up. and then write down the planning for the storage policy merge14:51
notmynamegholt: does RAX get next monday off?14:51
gholtHmm, uhm.14:51
*** Jarok has joined #openstack-swift14:52
glangeyes we do14:52
notmynameglange: no, I meant people who normally do work ;-)14:53
glangeouch, I'm glad you are still alive! :)14:53
notmyname /burn14:53
gholtI think it all depends on if you use your floating holiday for that. They converted most everything to floating holidays since we have folks all over the world. But I could be wrong for that particular holiday.14:53
notmynamecouldn't resist :-)14:53
notmynameah ok14:53
glangegholt: no, in this case it's a company holiday14:53
glangegholt: link in other channel :)14:53
notmynameclayg is at gluecon today, but I think he'll have the patch chain for storage policies up by the end of the week14:54
notmynameand that will make the freeze next week14:54
*** mattoliverau has quit IRC14:55
notmynameI'll write this down later today and send it out14:55
openstackgerritChristian Schwede proposed a change to openstack/python-swiftclient: Remove testtools.main() call from tests  https://review.openstack.org/9439114:56
notmynamechmouel: what are gate-swift-specs-docs in https://review.openstack.org/#/c/94140/ ?14:57
notmynamechmouel: what tests are those?14:57
creihtnotmyname: shouldn't you be resting or something?14:58
creihtnotmyname: glad you are ok btw :)14:58
notmynamecreiht: thanks. I'm staying home today. percocet and advil to keep me company. mostly it just feels like the worst krick in your neck you've ever had14:58
chmouelnotmyname: build the sphinx doc of the specs15:00
chmoueli.e: see the cookiecutter repo of the *-specs https://github.com/mtreinish/specs-cookiecutter15:00
notmynamechmouel: I haven't looked at the other specs repos yet, but what sort of docs do they have.....ah ok. looking15:00
chmoueli think it supposed to be uploaded to specs.openstack.org15:01
*** mkerrin has joined #openstack-swift15:01
chmouelbut that does not seems up15:01
notmynamehmm...that seems like a _lot_ of boilerplate15:01
chmouelnotmyname: it just so that the specs are pretty (i.e: html)15:02
notmynamewhy does it need setup.py? or anything .py? is someone installing specs?15:02
notmyname(isn't the code the installed version of specs? ;-) )15:02
creihtthey also have unit tests :)15:02
mtreinishchmouel: that's up on openstack-dev git now: http://git.openstack.org/cgit/openstack-dev/specs-cookiecutter/15:03
creihtit is a little crazy15:03
chmoueli don't think so,15:03
creihtI vote for just a simple template that we all follow15:03
chmouelthose are checks if the specs conform properly15:03
chmouelinstead of having reviewer doign it15:03
chmouel(i.e: humans)15:03
notmynamechmouel: ie that the proposed specs follow a template?15:03
chmoueli believe so15:03
chmouelor like spellchecking and such15:04
creihtyeah but it seems a bit overboard15:04
chmouelsee mtreinish link earlier15:04
gholtHmm, anybody have any tips for these profile-required tests? I installed python-profiler but I'm still getting unit test failures and errors on Lucid.15:05
notmynamechmouel: I understand that it's good to check that proposed specs have the bases covered. but (1) I want a different template than that and (2) humans still gotta read it too15:06
gholtOh, maybe it's because I don't have the newly required middleware, heheh.15:06
creihtyeah it isn't the same as code, and it seems to me that if a spec is missing stuff it should be obvious15:07
gholtOh wait, that doesn't make sense, these are unit tests. Damnit.15:08
creihtgholt: ugh15:08
notmynamecreiht: chmouel: I'd be ok with some checker that simply ensure that a few sections are there, but my first thought was something a little less.15:08
*** sandywalsh_ has quit IRC15:08
creihtgholt: I don't have a vm handy, are you coming in today?15:08
*** sandywalsh has joined #openstack-swift15:08
chmouelcreiht: yeah sure but take for example misspells what do we want to do have jenkins automatically checks for them or ignore them or wait that the reviewer checks for them?15:08
gholtcreiht: Yeah, I was trying to review your patch, but I can't get master's unit tests to pass... so...15:09
notmynamechmouel: has the swift-specs repo been created yet? or just the -infra pieces that were proposed15:09
creihtchmouel: I guess some people care about that more than I do15:09
chmouelnothing has been created15:09
notmynamechmouel: ok, thanks15:09
creihtI care a lot more about the content, not the sundries15:09
*** lpabon has quit IRC15:10
*** PVince81 has joined #openstack-swift15:10
notmynamecreiht: I propose that we not worry about a template until we actually create the repo. then we make it how we want based on what works best for us15:10
gholtMaybe I'll submit a patch to make the profile unit tests optional.15:10
creihtgholt: can you paste the errors you are seeing?15:11
gholtYeah, hold on there's a few.15:11
*** zul has quit IRC15:11
*** zul has joined #openstack-swift15:11
gholtcreiht: https://gist.github.com/gholt/95c7a8c4067991c66eb315:12
chmouelnotmyname, sure, i don't care about it much i just think it would make reviewer life easier to add those jenkins checks15:13
notmynamechmouel: ok, thanks15:13
acolesnotmyname: just catching up here - ouch! take it easy, and +1 for the helmet!15:13
portantenotmyname: glad to hear you are okay, broke mine a number of years ago, sleeping was never fun15:14
chmouelI should post pics of my bicycle accident as well :)15:14
notmynameportante: ya, it worked better that I thought it would last night15:15
*** kevinc has joined #openstack-swift15:19
*** foexle has quit IRC15:22
*** Jarok has left #openstack-swift15:23
creihtgholt: I'm trying to figure out why that is happening15:24
creihtbut it isn't obvious :/15:24
creihtI imagine it is some subtle py2.6 incompatible thing, but I haven't been able to find it yet15:28
ophukwhen setting up a single SAIO node - it uses the tempAuth authentication. Thats version 1.0 authentication right? And keystone is version 2.0?15:28
creihtophuk: correct15:29
creihtnotmyname: is there no way for infra to do 2.6 testing?15:29
ophukcreiht: ok, making sure. What is version 1.1?15:30
creihtI know you mentioned that there wasn't any 2.6 testing currently at the summit15:30
creihtophuk: version 1.1?15:30
notmynamecreiht: AFAIK, no. and they haven't in a while. last time they did was with with a maverick or natty image (both EOL)15:31
*** mattoliverau has joined #openstack-swift15:31
openstackgerritgholt proposed a change to openstack/swift: Make the new xprofile tests optional.  https://review.openstack.org/9440415:31
creihtnotmyname: then the, openstack will no longer support python 2.6 at the k release is a bit disengenuous15:32
creihtit already doesn't support it :/15:32
*** mattoliverau has quit IRC15:35
gholtI was told a while ago by mordred that Lucid isn't really supported anymore. It still has almost a year left of Canonical support and folks have it in production environments, but what can you do if the org doesn't feel like it? Makes me wonder how early Precise will get dropped though.15:35
openstackgerritChristian Schwede proposed a change to openstack/python-swiftclient: Fix Python3 bugs  https://review.openstack.org/9434715:35
ophukcreiht: not sure - I'm trying to use libcloud to connect to either of my openstack installs and they have a version 1.1 to authenticate against. They say it's deprecated though but here is the link, http://libcloud.readthedocs.org/en/latest/compute/drivers/openstack.html15:36
notmynamecschwede_: hmm...now that swiftclient does the py3 thing, I should cut a release of it15:36
cschwede_notmyname: i think my last patch needs to be merged first. I discovered issues with post_* using py315:37
notmynamecschwede_: ack15:37
creihtophuk: oh I think that was a a rev of the 1.0, but I can't remember what that added15:38
ophukcreiht: no worries. Thanks!15:38
creihtmaybe the list of services?15:38
cschwede_notmyname: I added you into the review loop, thus you get the info when it got merged (the py3 fix)15:39
*** byeager has quit IRC15:41
*** mkollaro has quit IRC15:44
notmynamewow. I just took a look at my helmet. it's cracked where I hit it yesterday.15:47
*** mattoliverau has joined #openstack-swift15:48
creihtnotmyname: so what sweet jumps were you trying to do when this occured? :)15:49
notmynamecreiht: nothing crazy. there was a drawbridge. and fire. and monster trucks. you know, the usual15:50
redboI was hoping for that clip15:51
notmynamechmouel: thanks for pushing the specs noop check15:51
notmynamepeluse_: heh, you kinda stepped in it with the ratelimit in /info patch. "here do this paul, it will be easy." all the service providers, "RAWWWR!"15:53
*** mlipchuk has quit IRC15:54
*** gyee has joined #openstack-swift15:54
gholtWell, on the plus side, it makes me happy folks all looking at posted patches. :D15:55
creihtgholt: I wonder if you are hitting this? http://bugs.python.org/issue737216:03
creihtunfortunately it doesn't say specifically which version it was fixed in16:04
creihtpython -c "import16:04
*** mattoliverau has quit IRC16:05
creihtoh I guess you need a profile file16:05
*** bvandenh has quit IRC16:05
creihtsee: http://bugs.python.org/issue131516:06
*** nacim has quit IRC16:06
*** hipster_ has joined #openstack-swift16:07
*** cds has joined #openstack-swift16:08
*** mlipchuk has joined #openstack-swift16:08
*** hipster has quit IRC16:11
*** byeager has joined #openstack-swift16:12
creihtgholt: yeah looks like you are hitting that python bug :(16:13
*** kevinc has quit IRC16:14
*** acoles is now known as acoles_away16:17
*** kevinc has joined #openstack-swift16:17
chmouelgholt: is constantine here?16:17
creihtchmouel: pconstantine_ I think16:18
*** nshaikh has quit IRC16:18
chmouelpconstantine_:  i can reproduce the functional tests error with you account copy object patch on a devstack with keystone16:20
chmouelcreiht: tks16:20
*** mwstorer has joined #openstack-swift16:20
chmouelpconstantine_: it does not seem a bug with the gate job16:20
creihtchmouel: thanks for looking into that16:20
*** byeager has quit IRC16:20
pconstantine_chmouel: are the errors the same as on dvsm?16:22
*** sneezewort has joined #openstack-swift16:22
chmouelpconstantine_: yes and without your patch no errors16:22
chmouelpconstantine_: are you testing with tempauth or with keystoneauth?16:23
pconstantine_chmouel: tempauth16:23
chmouelpconstantine_: that may be the issues16:23
pconstantine_but that means there is a bug in keystone16:23
pconstantine_or tempauth16:23
chmouelor in your patch :)16:23
*** mkerrin has quit IRC16:23
pconstantine_because behavior should be the same and it is not16:23
chmouelit could be that the keystoneauth middleware is lacking behind tempauth16:24
chmoueland that should be a bug if it doesn't match tempauth behavior16:24
openstackgerritChristian Schwede proposed a change to openstack/python-swiftclient: Fix Python3 bugs  https://review.openstack.org/9434716:24
*** haomaiwang has quit IRC16:24
pconstantine_my patch does not touch auth, it uses it as is, as it should obviously work with any auth16:25
*** mattoliverau has joined #openstack-swift16:25
pconstantine_if keystone is not working it's very very fishy16:25
chmouelpconstantine_: you are misunderstanding keystone and keystoneauth16:25
pconstantine_when I say "keystone" I mean keystoneauth obviously, or even swift.authorize() from kaystoneauth16:26
chmouelpconstantine_: I am not sure we can do much about it,16:26
chmouelpconstantine_: so yeah people add new stuff to tempauth and keystoneauth does not implement it16:27
chmoueli am not sure if thats called fishy16:27
pconstantine_I'm sure I was using keystoneauth somewhere and it did pass16:30
*** linuxbaby has quit IRC16:34
*** openstackgerrit has quit IRC16:35
*** openstackgerrit has joined #openstack-swift16:35
pconstantine_chmouel: what version of keystone is used in your devstack?16:38
chmouelpconstantine_: i use a clean devstack so using HEAD16:39
pconstantine_ok, thanks16:39
chmouelto reproduce it's literally just doing a devstack with this set in your local.conf ENABLED_SERVICES=key,mysql,s-proxy,s-account,s-account,s-container,swift16:41
pconstantine_chmouel: can you check in logs, maybe there is an explicit deny reason for 403, like "tenant mismatch" or something?16:41
chmouelthis would setup the file /etc/swift/test.conf for keystone16:41
chmouelpconstantine_: let me see if i can give you that vm access16:42
pconstantine_I cannot run devstack right now because internet in Rackspace Israel office blocks access to pypi ....16:42
chmouelpconstantine_: can you point me to your ssh key ?16:42
chmouelthere is a Rackspace Israel? woooot16:42
pconstantine_yep, there is :)16:42
chmouelso yeah if you give me your public key i can give you access to that vm16:43
chmouelpublic ssh key16:43
*** mattoliverau has quit IRC16:44
pconstantine_chmouel: https://gist.github.com/pkit/e6cdc90636bd8c14099616:44
*** mlipchuk has quit IRC16:44
chmouelpconstantine_: ok you can ssh to stack@devstack2.chmouel.com16:45
*** florentflament has joined #openstack-swift16:45
chmoueland be able to screen -x16:45
chmouelif you go to ~/swift16:45
chmoueland do a ./.functest16:45
chmouelyou'll reproduce the error16:45
peluse_notmyname:  wrt /info, yeah but its all good :)  bummer on the accident man!  last time I broke mine was actually at Univ of Colo at a party, was jumping from one balcony to the other and... well you can guess what happened next16:49
*** florentflament has quit IRC16:51
*** mkerrin has joined #openstack-swift16:52
*** baojg has joined #openstack-swift16:59
*** baojg has quit IRC17:04
*** mattoliverau has joined #openstack-swift17:06
*** mmcardle has quit IRC17:06
pconstantine_chmouel: it gets fishier and fishier, when running for example testCopyFromAccountHeader I get 403 in test, but in logs there is no 403 anywhere17:07
*** byeager has joined #openstack-swift17:07
pconstantine_chmouel: oh, God, finally, found it: proxy-server: tenant mismatch: AUTH_5f215c92d3bb44f6b710e6809824b171 != 8aca16965b6f443aa0f8ba6288918af8 (txn: tx82acb476f2404efdb9c91-00537b8bda) (client_ip:
*** byeager has quit IRC17:11
sneezewortAny good docs on backup up Cinder to Swift?17:15
*** kevinc has quit IRC17:24
*** kenhui has joined #openstack-swift17:26
*** Midnightmyth has quit IRC17:30
*** igor has quit IRC17:34
*** igor_ has joined #openstack-swift17:35
pconstantine_chmouel: found it, the ACLs for X-Container-Read/Write are not compatible between auth1.0 and auth 2.0, and swift_test_client.py does not attempt to address it, because no ACLs are ever checked in any func tests17:37
notmynamesneezewort: I think you should go ask in #openstack-cinder. I'd bug jgriffith about it17:37
pconstantine_chmouel: and newer test_object.py func tests do address it properly17:41
pconstantine_that's why all the tests pass there17:42
*** krtaylor has quit IRC17:42
*** lpabon has joined #openstack-swift17:52
*** kenhui has quit IRC17:54
*** PVince81 has quit IRC17:59
*** shague_ has joined #openstack-swift18:00
*** hipster_ has quit IRC18:06
*** mmcardle has joined #openstack-swift18:07
*** jamie_h has quit IRC18:07
*** mmcardle has quit IRC18:11
*** shri has joined #openstack-swift18:20
shague_I have a jenkins check-swift-dsvm-functional failure in test.functional.tests.TestSloTempurlUTF8blocking a devstack commit18:25
shague_I can't correlate it to a recehck bug so was wondering if anyone else has seen somthign similar?18:25
creihtshague_: not that I am aware of18:27
creihtshague_: but if you could point us to some more specific details, that might help18:27
creihtif someone not at RAX could review https://review.openstack.org/#/c/92933/18:27
creihtI would appreciate it :)18:27
shague_here is the log: http://logs.openstack.org/23/92123/1/check/check-swift-dsvm-functional/247711b/console.html18:27
*** kenhui has joined #openstack-swift18:29
creihtshague_: for whatever reason, looks like it got a 401 (unauthorized) from keystone18:29
*** lpabon has quit IRC18:30
creihtRESP BODY: {"access": {"token": {"issued_at": "2014-05-20T00:10:02.663827", "expires": "2014-05-20T01:10:02Z"18:30
creihtlooks like the token expired from keystone18:30
shague_yeah, so it seems random. My code change was to a neutron file. All other tests pass except that one18:30
creihtevery time?18:31
*** kenhui1 has joined #openstack-swift18:31
creihtshague_: I'm looking at some recent swift reviews, which those tests seem to pass18:32
shague_not sure. Haven't reran the test yet. Wasn't sure of the process to run the test.18:32
shague_is it OK to just add a comment "recheck no bug" to rerun the test or is there something different to do?18:33
creihtI don't think they let you do that any more18:33
*** kenhui has quit IRC18:33
*** bvandenh has joined #openstack-swift18:34
creihtholy cow18:34
creihtoh nm... was looking at the wrong field :)18:34
shague_so to rerrun I need a bug and then recheck18:35
creihtyeah I guess so18:35
creihtI'm trying to see if there is anything else that would indicate why that happened18:35
dfg"We’re cloud specialists,18:37
dfgso you don’t have to be"18:37
dfgwhat's wrong with having a "." at the end?18:37
creihtdfg would like to recall that last message18:38
dfgtotally on board with the no period.  completely useless18:38
*** kenhui has joined #openstack-swift18:38
*** baojg has joined #openstack-swift18:38
*** krtaylor has joined #openstack-swift18:40
*** kenhui1 has quit IRC18:41
*** kenhui2 has joined #openstack-swift18:41
creihtshague_: yeah not sure what caused that18:42
shague_Had same conclusion - but that doesn't mean much since I know zip about swift...18:42
shague_Thank you for the help.18:43
creihtshague_: yeah if it keeps coming up, let us know18:43
shague_will do18:43
*** kenhui1 has joined #openstack-swift18:44
*** kenhui has quit IRC18:44
*** byeager has joined #openstack-swift18:44
*** kenhui2 has quit IRC18:46
*** kenhui1 has quit IRC18:51
*** acoles_away is now known as acoles18:55
*** mmcardle has joined #openstack-swift19:07
openstackgerritChristian Schwede proposed a change to openstack/python-swiftclient: Fix Python3 bugs  https://review.openstack.org/9434719:11
*** mmcardle has quit IRC19:11
*** bvandenh has quit IRC19:13
*** baojg has quit IRC19:15
*** kevinc_ has joined #openstack-swift19:30
*** hipster has joined #openstack-swift19:33
*** hipster has quit IRC19:33
openstackgerritDavid Goetz proposed a change to openstack/swift: Add ability to remove subsections from /info  https://review.openstack.org/9445819:34
*** hipster has joined #openstack-swift19:34
*** mmcardle has joined #openstack-swift19:36
*** gyee has quit IRC19:36
*** annegent^ has quit IRC19:37
*** mmcardle1 has joined #openstack-swift19:38
*** mmcardle has quit IRC19:39
*** annegentle_ has joined #openstack-swift19:40
*** mmcardle1 has quit IRC19:43
*** byeager has quit IRC19:46
openstackgerritAlistair Coles proposed a change to openstack/python-swiftclient: Relax requirement for tenant_name in get_auth()  https://review.openstack.org/9435319:53
*** baojg has joined #openstack-swift19:55
*** jeblair has quit IRC19:55
*** jeblair has joined #openstack-swift19:55
*** baojg has quit IRC20:00
*** zaitcev has joined #openstack-swift20:16
*** ChanServ sets mode: +v zaitcev20:16
*** baojg has joined #openstack-swift20:16
openstackgerritgholt proposed a change to openstack/swift: Container sync no longer sending swift_bytes value  https://review.openstack.org/9446520:23
*** judd7 has joined #openstack-swift20:26
*** Trixboxer has quit IRC20:29
*** krtaylor has quit IRC20:29
*** kevinc_ has quit IRC20:33
*** elambert has joined #openstack-swift20:34
*** mmcardle has joined #openstack-swift20:39
*** mkollaro has joined #openstack-swift20:40
*** baojg has quit IRC20:42
*** mmcardle has quit IRC20:43
*** shakamunyi has joined #openstack-swift20:45
*** elambert has quit IRC20:50
portantepconstantine_: so where are you at with the account-to-account patch?20:51
portantedo you want help with that?20:51
openstackgerritDavid Goetz proposed a change to openstack/swift: Add ability to remove subsections from /info  https://review.openstack.org/9445820:53
*** kevinc_ has joined #openstack-swift20:55
openstackgerritA change was merged to openstack/python-swiftclient: Move test_shell.py under tests/unit/  https://review.openstack.org/9433520:58
*** judd7 has quit IRC21:01
*** byeager has joined #openstack-swift21:04
portantecreiht: on my SAIO, I ran the unit tests for your swift-init patch, and I get a failing unit test for internal client21:11
* notmyname is back online after sleeping off some pain meds21:11
*** acoles is now known as acoles_away21:11
portantenotmyname: hope your recover is quick21:12
notmynamethanks. me too. flying to CO will be "fun" ;-)21:12
portanteoh yes21:12
notmynameI don't guess I'll be wearing a 20lbs backpack for that trip21:13
* peluse_ heads out for vacation for the rest of the week...21:16
portantepeluse_: have a good one!21:17
notmynamepeluse_: thanks for the well-put email to eric21:17
notmynamepeluse_: have a good vacation!21:17
*** anticw has joined #openstack-swift21:19
*** byeager has quit IRC21:19
notmynamejuno schedule, in general https://wiki.openstack.org/wiki/Juno_Release_Schedule21:23
*** csd has joined #openstack-swift21:25
*** kevinc_ has quit IRC21:27
*** byeager has joined #openstack-swift21:29
*** krtaylor has joined #openstack-swift21:31
ophukhow come with i put a bind_port in my proxy-server.conf and account-server.conf swift-init won't start?21:33
portanteare you running them on the same system with the same bind port number?21:34
portanteophuk: is this an SAIO?21:35
ophukportante: yeah...good point. I'll run one at 8080 and the other at 8081. I'm getting 503 errors while trying to create a container. No, this is a semi-full set up with keystone. My SAIO set up is on a different box21:35
*** foexle has joined #openstack-swift21:44
*** jergerber has joined #openstack-swift21:49
*** lpabon has joined #openstack-swift21:50
*** lpabon has quit IRC21:55
*** kevinc_ has joined #openstack-swift21:56
notmynamechmouel: FYI, per the discussion/vote in the project-wide meeting just now, the -specs repo will be named object-specs or objectstorage-specs (rather than swift-specs)21:56
notmynamesince it will be a single specs repo for the entire openstack program (swift, swiftclient, and swift-bench)21:57
creihtnotmyname: is nova changing to compute-specs?21:57
notmynamecreiht: ya, I think so. all will change21:57
creihtI think that is wrong21:57
Alex_GaynorI do too, but I have no energy to care :-)21:57
creihtbut whatever21:57
creihtbecaue you are defining specs for a specific project called swift21:58
notmynamebut also swift-bench and python-swiftclient projects21:58
creihtif we have them for swiftclient they should be in a separate repo21:58
creihtand there absolutely doesn't need to be any specs for swift-bench :)21:58
creihtit will be a bit confusing21:58
Alex_GaynorBoth of which have swift in their name. Mostly though, swift-specs is less of a mouthful21:58
*** cihhan has left #openstack-swift21:58
notmynameAlex_Gaynor: ya, I agree with that21:58
creihtswiftclient-specs would be fine with me21:59
*** byeager has quit IRC21:59
creihtfor swiftclient specific stuff21:59
notmynameIMO consistency > {project}-specs > {program}-specs21:59
Alex_Gaynorthat's fair21:59
creihthow is having a {project}-specs any less consistent then {program}-specs?22:00
creihtseems like {project}-specs is much more explicit22:00
creihtif we need specs for something like python-swiftclient, then it deserves its own specs repo22:00
*** gyee has joined #openstack-swift22:01
creihtbut whatever22:01
notmynameIMO because there should be one place to go to propose and review stuff dealing with object storage, as managed under openstack.22:02
notmynameI completely agree that it would work to have different specs repos22:02
notmynamebut mostly I think it's our specs repo that we can use as we want :-)22:03
creihtexcept where everyone else says we can't22:05
creihtit just seems to be a lot more obvious to find swift specs at openstack/swift-specs rather than openstack/object-storage-specs22:07
creihtnotmyname: and you aren't allowed to go to any more meetings while on drugs ;)22:08
notmynamecreiht: lol22:09
notmynamecreiht: so all those places where we're told we can't do something or should do something else, as I said last week, it turns out that there's far from much commonality across the openstack projects :-)22:10
notmyname-specs is ours, so we can do what we want. we can't be a slave to our tools. make the tools work for us22:10
notmynamewe should have our own template, our own whatever (or not, eg specs docs? specs tests? that's just weird)22:11
creihtuntil we can't because openstack22:11
notmynameand maybe the simplest answer is to have 3 top level directories in -specs for the projects22:12
*** dvas has joined #openstack-swift22:12
creihtI shouldn't let the little things get me so frustrated22:13
creihtbut it seems like it is the little things that always leads to the bigger things22:13
creihtI look forward to the day when all the openstack spec repos are continously integrated22:14
notmynamecreiht: if you drank I would say you need to go get a beer and relax ;-)22:17
*** kenhui has joined #openstack-swift22:20
claygdfg: what's the use-case for "remove sections from /info" - maybe they shouldn't be there in the first place?22:23
notmynameclayg: stuff that's useful to have, but eg public service providers don't want to expose22:24
notmynameclayg: the specific case under discussion now are the ratelimit params22:24
claygnotmyname: yeah but... wait - aren't you like brokent in half right now?22:24
*** mkollaro has quit IRC22:24
notmynameclayg: I'm sorta glued together a little. well there's a sling. and pain meds22:24
claygoic, yeah - i thought sometime someone decided it's better for people to know why they're slow then trying to up their concurrency and just getting confused?22:25
* notmyname is not currently on pain meds22:25
notmynameclayg: did you see peluse_'s patch to add ratelimiting to /info?22:25
notmynamethere was some discussion there between RAX and HP on what to expose or not22:25
notmynamevs eg our customers in a provate cloud setting22:26
claygi must have missed that discussion22:26
notmynameluisbg: https://review.openstack.org/#/c/88633/22:26
notmynameclayg: https://review.openstack.org/#/c/88633/22:26
notmynameluisbg: whoops22:26
dfgacoles_away: you cool with me just adding a comment saying not top having dict keys with "."s in them instead of walking all the dicts to look for "."'s ? i just don't see it as being a problem (its in source code) and don't want to complicate things.22:30
dfghe'll see that right?22:30
dfgoh i'll just put it in the ticket22:31
ophukIf I set a different bind_port for proxy and account and make sure all storage nodes have the same account bind_port swift doesn't appear to work. When I take out the bind_port in account-server.conf I can do swift stat but can't create containers due to 503 errors or insufficient space. Ideas on what to look at?22:31
*** kenhui has quit IRC22:31
claygophuk: all the servers need *different* bind ports if they're running on the same machine22:32
claygsay.. 8080 for proxy 6000, 6010, 6020 for object/container/account22:32
*** kenhui has joined #openstack-swift22:33
ophukclayg: different machines. Proxy and account server is on one machine. oh..when i use swift-ring-builder. The port in there is the port that needs to be in account-server.conf - well the account one22:34
ophukclayg: but yes each of those are different ports, 6000-600222:34
*** byeager has joined #openstack-swift22:41
*** csd has quit IRC22:41
*** csd has joined #openstack-swift22:41
claygso as long as the service bind ports match the ring entries it should work22:42
claygaccount stat can be tricky with autocreate account's fake-y responses22:42
*** csd has quit IRC22:42
claygyou could try just adding some metadata to the account to force the account PUT (i.e. swift post -m 'color:blue')22:42
*** csd has joined #openstack-swift22:43
ophukclayg: is there a way to pull that meta data back, swift get -m ? the post didn't return an error. All the rings ports match the ports in *-server.conf on all nodes yet I'm still getting insufficient storage errors while trying to create a container22:44
notmynameclayg: did you get to gluecon? (glucon? gloocon?)22:45
*** kenhui has quit IRC22:46
*** sneezewort has quit IRC22:52
notmynamemost unhelpful review comment ever22:56
notmynamemine on https://review.openstack.org/#/c/94458/22:56
*** bill_az has quit IRC23:01
ophukclayg: oh I see the meta data now - yes it did work. Creating a container still doesn't though I've noticed that it's stating that the device isn't mounted though it is23:01
*** byeager has quit IRC23:09
*** byeager has joined #openstack-swift23:10
*** dvas has quit IRC23:16
*** byeager has quit IRC23:17
*** kevinc_ has quit IRC23:21
*** tdasilva has left #openstack-swift23:23
*** sungju has joined #openstack-swift23:28
h6wnotmyname: Canada Geese honk to encourage the leader to keep going.  I've taken to just using "*honk*" in review comments like that. :-p23:28
* h6w tends to overcapitalise to compensate for lack of coffee. :-p23:30
*** kenhui has joined #openstack-swift23:30
*** kenhui has quit IRC23:35
*** kenhui has joined #openstack-swift23:35
*** csd has quit IRC23:40
*** csd has joined #openstack-swift23:40
*** kevinc_ has joined #openstack-swift23:43
*** kenhui has quit IRC23:49
*** foexle has quit IRC23:51
h6wWhat's the default logfile if it's not specified?   Dependent on your log handler?  e.g. /var/log/syslog | /var/log/messages23:53
h6wThere doesn't seem to be a /var/log/swift, so I'm presuming it goes global.23:53
*** kevinc_ has quit IRC23:54
notmynameh6w: right. it uses the default syslog handler you have configured23:54
notmynameh6w: http://docs.openstack.org/developer/swift/development_saio.html#optional-setting-up-rsyslog-for-individual-logging  <-- eg23:54
notmynameh6w: I prefer the syslog-ng config setup better, but we documented rsyslogd just because that's what's there by default23:55
h6wYeah. /me sticking with defaults to keep things more vanilla and less complicated.23:56
*** mwstorer has quit IRC23:58

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