Friday, 2014-09-26

*** bknudson has quit IRC00:01
*** oomichi has joined #openstack-keystone00:03
*** gokrokve has joined #openstack-keystone00:10
*** marcoemorais has quit IRC00:12
*** marcoemorais has joined #openstack-keystone00:12
*** gokrokve has quit IRC00:13
*** topol has quit IRC00:17
*** dims has quit IRC00:20
*** morgan_remote_ has quit IRC00:20
*** dims has joined #openstack-keystone00:21
*** richm has quit IRC00:24
*** dims has quit IRC00:25
*** andreaf has quit IRC00:26
*** andreaf has joined #openstack-keystone00:26
*** gokrokve has joined #openstack-keystone00:44
*** cjellick has quit IRC00:44
*** gokrokve has quit IRC00:45
*** gokrokve has joined #openstack-keystone00:45
*** cjellick has joined #openstack-keystone00:46
*** cjellick has quit IRC00:46
*** ncoghlan has joined #openstack-keystone00:52
*** _cjones_ has quit IRC00:55
*** _cjones_ has joined #openstack-keystone00:56
*** _cjones_ has quit IRC01:00
*** dims has joined #openstack-keystone01:07
*** dims_ has joined #openstack-keystone01:09
*** alex_xu has quit IRC01:11
*** dims has quit IRC01:13
*** marcoemorais has quit IRC01:14
openstackgerritBrant Knudson proposed a change to openstack/keystone: Tests raise exception if logging problem  https://review.openstack.org/11994601:15
openstackgerritBrant Knudson proposed a change to openstack/keystone: sys.exit mock cleanup  https://review.openstack.org/12424001:15
ayoungboltR, I was off today, should have shut down my IRC client, but...01:27
ayoungyes, you have it right.01:27
*** alex_xu has joined #openstack-keystone01:30
*** junhongl has quit IRC01:36
*** junhongl has joined #openstack-keystone01:36
openstackgerritNathan Kinder proposed a change to openstack/identity-api: Correct response status for HEAD requests  https://review.openstack.org/12424301:39
openstackgerritNathan Kinder proposed a change to openstack/keystone: Add expected status for all HEAD requests in tests  https://review.openstack.org/12424401:44
*** HenryG has joined #openstack-keystone01:49
*** andreaf has quit IRC01:52
*** r-daneel has quit IRC01:56
*** diegows has quit IRC02:02
*** zzzeek has quit IRC02:10
*** packet has joined #openstack-keystone02:23
*** packet has quit IRC02:23
*** packet has joined #openstack-keystone02:23
*** morgan_remote_ has joined #openstack-keystone02:26
morgan_remote_ayoung: nice token constraint description.02:26
*** gokrokve has quit IRC02:33
*** KanagarajM has joined #openstack-keystone02:37
*** harlowja is now known as harlowja_away02:38
*** marcoemorais has joined #openstack-keystone02:43
*** marcoemorais1 has joined #openstack-keystone02:45
*** amcrn has quit IRC02:47
*** wanghong has quit IRC02:48
*** marcoemorais has quit IRC02:48
*** _cjones_ has joined #openstack-keystone02:54
*** ncoghlan is now known as ncoghlan_afk02:55
*** wanghong has joined #openstack-keystone02:59
*** ncoghlan_afk is now known as ncoghlan03:04
*** KanagarajM has quit IRC03:05
*** KanagarajM has joined #openstack-keystone03:05
*** dims_ has quit IRC03:11
*** packet has quit IRC03:11
*** gyee has quit IRC03:24
*** KanagarajM has quit IRC03:24
*** KanagarajM has joined #openstack-keystone03:25
openstackgerritOpenStack Proposal Bot proposed a change to openstack/keystone: Updated from global requirements  https://review.openstack.org/12425903:54
*** zhiyan has quit IRC03:54
*** zhiyan has joined #openstack-keystone03:57
*** dims has joined #openstack-keystone04:12
*** dims has quit IRC04:17
*** arborism has joined #openstack-keystone04:25
*** KanagarajM has quit IRC04:49
*** KanagarajM has joined #openstack-keystone04:52
stevemarwe should use `git review --no-rebase` when the dependent change is in the gate right?04:54
morganfainbergstevemar, yes04:54
morganfainbergstevemar, or -R04:54
stevemarthanks morganfainberg04:54
morganfainbergnp04:55
morganfainbergliterally just got back from OpenStack LA meetup too :P04:55
stevemarjust double checking... it's been a while since i actually cared about something not getting bumped out04:55
stevemarhowd that go?04:55
morganfainbergwas a good meetup04:55
morganfainbergVXLAN talk04:55
morganfainbergstevemar, slloooowly slooooowly we're having things make their way through the gate.04:56
*** gokrokve has joined #openstack-keystone04:59
openstackgerritSteve Martinelli proposed a change to openstack/keystone: Use openstackclient examples in configuration documentation  https://review.openstack.org/12427005:00
*** morgan_remote_ has quit IRC05:00
stevemarmorganfainberg, got a quick q for you, style question05:03
morganfainbergstevemar, quick answer for you.05:04
morganfainberg4205:04
stevemarmorganfainberg, if i were to show keystoneclient examples here too: http://docs-draft.openstack.org/95/124095/1/check/gate-keystone-docs/2a219d8/doc/build/html/cli_examples.html05:04
stevemarshould I put it at the end or the docs, "like here's a quick summary in ksc CLI"05:04
stevemaror inline?05:04
morganfainbergi would make it a separate section unless you completely rewrite to to say "this document will explain the osc method and then the ksc method for accomplishing tasks"05:05
*** KanagarajM has quit IRC05:05
morganfainbergso you'd have OSC - projects - project create05:06
morganfainbergfor example05:06
morganfainbergand KSC - projects - project create05:06
stevemardo you have a preference for either?05:06
*** ncoghlan is now known as ncoghlan_afk05:06
morganfainbergnot really05:06
stevemarmorganfainberg, I was thinking under example, to just put: "Here is the equivalent command in KSC": blah05:09
stevemarbut the problem with that is, it may take different parameters05:09
morganfainbergit might be cleanest to make it separate sections05:10
stevemaryeah, a bit redundant, but yeah05:10
*** arborism has quit IRC05:10
*** KanagarajM has joined #openstack-keystone05:13
*** _cjones_ has quit IRC05:16
*** _cjones_ has joined #openstack-keystone05:17
*** KanagarajM has quit IRC05:18
*** fifieldt has joined #openstack-keystone05:21
*** _cjones_ has quit IRC05:21
*** _cjones_ has joined #openstack-keystone05:22
*** ncoghlan_afk is now known as ncoghlan05:24
openstackgerritSteve Martinelli proposed a change to openstack/keystone: Update the CLI examples to use openstackclient  https://review.openstack.org/12409505:25
*** ajayaa has joined #openstack-keystone05:26
*** KanagarajM has joined #openstack-keystone05:31
stevemarmorganfainberg, i realllllly love the new project-config repo05:33
stevemarit's so much less daunting now!05:33
morganfainbergright?!05:33
morganfainbergok i'm gonna go lie down05:33
morganfainbergugh. tirred05:33
morganfainbergstevemar, also https://review.openstack.org/#/c/124243/05:34
*** ncoghlan is now known as ncoghlan_afk05:34
*** amcrn has joined #openstack-keystone05:37
*** vdreamarkitex has quit IRC05:37
*** gokrokve_ has joined #openstack-keystone05:40
*** rushiagr_away is now known as rushiagr05:41
*** KanagarajM has quit IRC05:42
*** gokrokve has quit IRC05:43
*** gokrokve_ has quit IRC05:45
stevemarugh! gate restarted :(05:45
stevemarmorganfainberg, at least we're near the top05:45
*** afazekas has joined #openstack-keystone05:48
stevemarmorganfainberg, i -1'ed the keystone code05:49
*** KanagarajM has joined #openstack-keystone05:50
*** ajayaa has quit IRC05:56
openstackgerritOpenStack Proposal Bot proposed a change to openstack/keystone: Imported Translations from Transifex  https://review.openstack.org/12394106:05
*** ncoghlan_afk is now known as ncoghlan06:11
*** gokrokve has joined #openstack-keystone06:15
*** ajayaa has joined #openstack-keystone06:16
*** gokrokve has quit IRC06:20
*** stevemar has quit IRC06:26
*** uvirtbot has quit IRC06:28
*** swartulv has quit IRC06:29
*** KanagarajM has quit IRC06:29
*** swartulv has joined #openstack-keystone06:30
*** k4n0 has joined #openstack-keystone06:37
*** uvirtbot has joined #openstack-keystone06:43
*** lufix has joined #openstack-keystone06:49
openstackgerritA change was merged to openstack/keystone: Adding an index on token.user_id and token.trust_id  https://review.openstack.org/10204106:53
*** dims has joined #openstack-keystone07:02
viklundI'm planning on using linotp for 2-factor auth. And I'd like hear if my approach is sound.07:02
viklundWe're using a LDAP backend to keystone and my idea was to overide the authenticate method to not only try to auth against ldap07:03
viklundbut after a successful auth (or before) send  the OTP to linotp07:03
viklundI can't decide on whether I should make a new backend that inherits everything from the current LDAP and just implements the extra auth-logic07:04
viklundor if I should just make a small patch against the current ldap backend.07:05
*** dims has quit IRC07:07
*** jaosorior has joined #openstack-keystone07:08
*** marekd|away is now known as marekd07:08
*** BAKfr has joined #openstack-keystone07:10
*** f13o has quit IRC07:12
*** boris-42 has quit IRC07:15
*** swartulv has quit IRC07:15
*** gokrokve has joined #openstack-keystone07:16
*** swartulv has joined #openstack-keystone07:17
*** marcoemorais1 has quit IRC07:19
*** gokrokve has quit IRC07:21
*** alex_xu has quit IRC07:30
*** alex_xu has joined #openstack-keystone07:37
*** _cjones_ has quit IRC07:54
*** _cjones_ has joined #openstack-keystone07:54
*** ukalifon has joined #openstack-keystone07:55
openstackgerritA change was merged to openstack/keystone: Fix failure of delete domain group grant when identity is LDAP.  https://review.openstack.org/12358507:57
*** amerine has quit IRC07:58
*** _cjones_ has quit IRC07:59
*** bdossant_ has quit IRC08:01
*** bdossant has joined #openstack-keystone08:01
*** ncoghlan has quit IRC08:08
*** oomichi has quit IRC08:11
openstackgerritA change was merged to openstack/keystone: Clean up the Configuration documentation  https://review.openstack.org/12396008:13
openstackgerritA change was merged to openstack/keystone: New section for CLI examples in docs  https://review.openstack.org/12396908:13
openstackgerritA change was merged to openstack/keystone: Update architecture documentation  https://review.openstack.org/12393808:13
*** gokrokve has joined #openstack-keystone08:15
*** gokrokve has quit IRC08:17
*** gokrokve has joined #openstack-keystone08:17
*** gokrokve has quit IRC08:21
*** ajayaa has quit IRC08:21
*** f13o has joined #openstack-keystone08:28
*** amcrn has quit IRC08:32
*** ajayaa has joined #openstack-keystone08:34
*** jamiec has joined #openstack-keystone08:45
*** dvorak has quit IRC08:47
*** KanagarajM has joined #openstack-keystone08:47
*** aix has joined #openstack-keystone08:49
*** dvorak has joined #openstack-keystone08:50
*** Dafna has joined #openstack-keystone09:07
*** lsmola has joined #openstack-keystone09:15
*** gokrokve has joined #openstack-keystone09:16
*** gokrokve has quit IRC09:17
*** gokrokve has joined #openstack-keystone09:18
*** gokrokve has quit IRC09:22
*** andreaf has joined #openstack-keystone09:28
*** KanagarajM has quit IRC09:50
ekarlsohey guys, will ksclient requires be bumped to 0.11 ?10:07
*** gokrokve has joined #openstack-keystone10:16
*** gokrokve has quit IRC10:21
*** dims has joined #openstack-keystone10:37
*** ajayaa has quit IRC11:12
*** ajayaa has joined #openstack-keystone11:40
*** k4n0 has quit IRC11:46
*** diegows has joined #openstack-keystone11:47
openstackgerritAlvaro Lopez Garcia proposed a change to openstack/python-keystoneclient: auth_token: http_connect_timeout should be an int  https://review.openstack.org/11721311:50
openstackgerritAlvaro Lopez Garcia proposed a change to openstack/python-keystoneclient: DO NOT MERGE: debug failure for change I201c2dbe48e649ac302406dfff16aa482aa7cea2  https://review.openstack.org/12438311:56
*** htruta has joined #openstack-keystone12:10
*** vdreamarkitex has joined #openstack-keystone12:17
*** radez_g0n3 is now known as radez12:17
ayoungmorganfainberg, thanks12:26
openstackgerritIlya Pekelny proposed a change to openstack/keystone: Comparision of database models and migrations.  https://review.openstack.org/8063012:56
*** gordc has joined #openstack-keystone12:57
*** dims has quit IRC12:58
*** dims has joined #openstack-keystone12:59
*** joesavak has joined #openstack-keystone13:05
*** richm has joined #openstack-keystone13:12
*** dhellmann_ is now known as dhellmann13:13
*** gokrokve has joined #openstack-keystone13:16
*** gokrokve has quit IRC13:21
openstackgerritayoung proposed a change to openstack/python-keystoneclient-kerberos: kerberos client plugin  https://review.openstack.org/12361413:24
openstackgerritMarek Denis proposed a change to openstack/keystone: Set issuer value to CONF.saml.idp_entity_id.  https://review.openstack.org/12417613:25
openstackgerritayoung proposed a change to openstack/python-keystoneclient-kerberos: kerberos client plugin  https://review.openstack.org/12361413:27
*** bdossant_ has joined #openstack-keystone13:31
*** bdossant has quit IRC13:34
*** adrienverge has joined #openstack-keystone13:35
*** ajayaa has quit IRC13:37
adrienvergeHi all13:37
adrienvergeWhat would be the best way to test if a service exists (e.g. 'volumev2')? Knowing that the requesting user doesn't necessarily have the right to list services in policy.13:37
*** sigmavirus24_awa is now known as sigmavirus2413:39
*** rushiagr is now known as rushiagr_away13:43
*** ajayaa has joined #openstack-keystone13:43
*** bdossant_ has quit IRC13:43
*** bdossant has joined #openstack-keystone13:44
dstanekadrienverge: so you dont' have a catalog? what information do you have?13:46
adrienvergedstanek: I'm calling from Heat, in the Cinder client _create() method. I want to know what volumes services are available ('volume' and/or 'volumev2').13:49
adrienvergedstanek: I don't see any way to access the catalog without calling keystone.client.services.list() / find()13:50
adrienvergedstanek: Is it stored somewhere in the context?13:51
*** ajayaa has quit IRC13:52
dstanekadrienverge: if you have an instance of HTTPClient it may be in client.service_catalog13:52
dstanekadrienverge: i'm sorta guessing by piecing together what i know13:53
*** zzzeek has joined #openstack-keystone13:54
adrienvergedstanek: Thanks, I'll check that13:58
*** ayoung is now known as ayoung-afk14:15
*** andreaf has quit IRC14:16
*** andreaf has joined #openstack-keystone14:17
*** adrienverge has quit IRC14:17
*** dims has quit IRC14:19
*** dims has joined #openstack-keystone14:21
*** stevemar has joined #openstack-keystone14:22
*** sigmavirus24 is now known as sigmavirus24_awa14:24
*** dims has quit IRC14:25
*** david-lyle has joined #openstack-keystone14:25
*** andreaf has quit IRC14:26
*** andreaf has joined #openstack-keystone14:27
openstackgerritSteve Martinelli proposed a change to openstack/keystone: Use openstackclient examples in configuration documentation  https://review.openstack.org/12427014:27
openstackgerritSteve Martinelli proposed a change to openstack/keystone: Update the CLI examples to use openstackclient  https://review.openstack.org/12409514:27
*** sigmavirus24_awa is now known as sigmavirus2414:27
openstackgerritSteve Martinelli proposed a change to openstack/keystone: Update the CLI examples to use openstackclient  https://review.openstack.org/12409514:31
stevemarhenrynash, ^14:32
*** dims has joined #openstack-keystone14:32
*** dims has quit IRC14:32
*** adrienverge has joined #openstack-keystone14:33
*** ukalifon has quit IRC14:33
*** dims has joined #openstack-keystone14:33
*** dims has quit IRC14:34
*** dims has joined #openstack-keystone14:35
*** dims has quit IRC14:35
*** zoresvit has joined #openstack-keystone14:41
*** dims has joined #openstack-keystone14:42
*** dims has quit IRC14:42
*** ukalifon has joined #openstack-keystone14:42
*** dimsum_ has joined #openstack-keystone14:43
*** dimsum_ has quit IRC14:43
*** dimsum_ has joined #openstack-keystone14:45
*** jorge_munoz has joined #openstack-keystone14:49
lbragstadbknudson1: you had some comments here that didn't get addressed before the change was merged https://review.openstack.org/#/c/119452/3114:50
*** vhoward has left #openstack-keystone14:50
lbragstadshould another patch be pushed to master before attempting to backport that?14:50
bknudson1lbragstad: that's pretty common... it takes me a long time to do reviews14:50
lbragstadAFAICT, they are valid comments though14:50
bknudson1backport?14:51
bknudson1to stable/icehouse?14:51
lbragstadbknudson1: yes14:51
lbragstadit fixed this bug https://bugs.launchpad.net/keystone/+bug/136044614:51
uvirtbotLaunchpad bug 1360446 in keystone "client connection leak to memcached under eventlet due to threadlocal" [Medium,Fix committed]14:51
bknudson1lbragstad: I haven't been able to make the time to propose the changes myself.14:53
bknudson1also, I don't know what the response should be to some of them14:53
bknudson1lbragstad: too busy with PSIRTs14:53
lbragstad:)14:54
lbragstadnew hire?14:54
bknudson1no, a guy in shanghai... he's not as reliable as we'd hope.14:55
lbragstadbknudson1: I can see if I can push something up to master, fixing your comments, and it can be iterated there I guess?14:55
bknudson1and they've got several week-long national holidays14:55
bknudson1lbragstad: if you've got the time to make the changes, go ahead.14:57
nkinder_stevemar: thanks for pointing out that head() checks for 204 in test_v3 already14:57
nkinder_stevemar: I totally missed that!14:57
lbragstadbknudson1: I have a few minutes right now, I'll see if I can address some of your comments.14:58
nkinder_stevemar: I added you to a related API doc change - https://review.openstack.org/#/c/124243/14:58
stevemarnkinder_, np! i actually have that open in a browser tab now :)14:58
openstackgerritJulien Danjou proposed a change to openstack/keystonemiddleware: Switch to oslo.utils  https://review.openstack.org/12443515:02
openstackgerritJulien Danjou proposed a change to openstack/keystonemiddleware: Switch to oslo.serialization  https://review.openstack.org/12443615:02
*** bdossant has quit IRC15:02
*** radez is now known as radez_g0n315:05
*** adrienverge has quit IRC15:05
*** _cjones_ has joined #openstack-keystone15:10
*** gokrokve has joined #openstack-keystone15:16
openstackgerritSteve Martinelli proposed a change to openstack/keystone: Use openstackclient examples in configuration documentation  https://review.openstack.org/12427015:18
openstackgerritSteve Martinelli proposed a change to openstack/keystone: Update the CLI examples to use openstackclient  https://review.openstack.org/12409515:18
*** thedodd has joined #openstack-keystone15:19
*** afazekas has quit IRC15:19
*** gokrokve has quit IRC15:20
*** anteaya has quit IRC15:23
dstaneklbragstad: old-style classes..yuck15:24
lbragstaddstanek: :)15:24
lbragstadbknudson1: note, I'm not real sure what to do about some of those comments for the doc strings, so I might have to defer those to someone with a little more knowledge on the patch15:25
bknudson1lbragstad: so nobody knows how the class is supposed to work?15:26
lbragstadbknudson1: about this comment, https://review.openstack.org/#/c/119452/31/keystone/tests/core.py you want this in our common log implementation?15:26
morganfainbergmornin15:26
lbragstadbknudson1: speak of the devil ^ that's probably the guy we're looking for15:27
morganfainberglbragstad, i don't think that belonds in oslo.15:27
bknudson1lbragstad: I think we've got similar code somewehere that affects server.15:27
morganfainbergbknudson1,lbragstad, but it is something we can probably override somehow15:27
dstaneklbragstad: which comments are you unsure of?15:27
lbragstadbknudson1: doing a quick grep and I didn't see it15:28
lbragstaddstanek: some of the doc string questions15:28
lbragstadhttps://review.openstack.org/#/c/119452/31/keystone/common/cache/_memcache_pool.py15:29
bknudson1lbragstad: http://git.openstack.org/cgit/openstack/keystone/tree/keystone/config.py#n2915:29
bknudson1it's called from keystone-all and probably in the wsgi server too15:30
morganfainbergbknudson1, a lot of your comments are valid on this and actually should be cleaned up in the next phase which is "make this it's own library" not something we need to carry in keystone15:30
*** cjellick has joined #openstack-keystone15:30
dstaneklbragstad: i'll add some additonal commentary15:30
morganfainbergthe only reason this is in keystone is because we are beyond dep freeze :(15:30
bknudson1morganfainberg: it's shared with middleware?15:30
morganfainbergbknudson1, yeah. thats why it needs to be split out15:31
bknudson1to... keystoneclient?15:31
bknudson1that would be a bad place15:31
morganfainbergkeystoneclient we didn't port to, we probably should.15:31
bknudson1if it needs to be in keystoneclient anyways then could just leave it there15:31
morganfainbergno, the idea is it should stand alone something not carrlied by keystone packages15:31
morganfainbergsince it should be pluggable into dogpile itself (in the case of ekystone)15:32
morganfainbergbut there is more than just work on the keystone side to get there.15:32
openstackgerritLance Bragstad proposed a change to openstack/keystone: Address some late comments for memcache clients  https://review.openstack.org/12444315:33
lbragstaddstanek: ^15:33
lbragstadyou can iterate over that review15:33
morganfainbergalso the queue object is weird because in non-eventlet it's an old-style class, and under eventlet it's a new style class :(15:33
bknudson1looks like it belongs in dogpile.cache.backends15:33
openstackgerritSteve Martinelli proposed a change to openstack/keystone: Update the CLI examples to also use openstackclient  https://review.openstack.org/12409515:33
dstaneklbragstad: you can't use super15:33
morganfainbergbknudson1, the idea is dogpile will get configurable pools.15:33
*** lufix has quit IRC15:33
openstackgerritNathan Kinder proposed a change to openstack/keystone: Improve documentation of RBAC policy behavior  https://review.openstack.org/12386215:33
morganfainbergbknudson1, so it'll not be in dogpile directly15:34
morganfainbergbknudson1, or if it is it'll be different than dogpile.cache.backends15:34
* lbragstad brb15:34
morganfainbergbknudson1, talked with zzzeek a lot on this and thats the direction we're hoping to take it.15:34
morganfainbergbknudson1, zzzeek is the dogpile maintainer :)15:34
bknudson1he's busy15:34
morganfainbergyep15:34
*** gokrokve has joined #openstack-keystone15:36
nkinder_morganfainberg: this change can be abandoned - https://review.openstack.org/#/c/111088/15:39
nkinder_morganfainberg: I don't have the ability to do so, but I've talked with the reporter (and a customer who requested that change)15:39
morganfainbergnkinder_, ok15:39
morganfainbergnkinder_, will do so15:39
nkinder_morganfainberg: cool, thx15:40
*** andreaf has quit IRC15:41
*** andreaf has joined #openstack-keystone15:42
*** BAKfr has quit IRC15:42
*** samuelmz has quit IRC15:44
openstackgerritDavid Stanek proposed a change to openstack/keystone: Address some late comments for memcache clients  https://review.openstack.org/12444315:50
dstaneklbragstad: ^ i removed your use of super() and added a few docstrings15:50
morganfainbergdstanek, mind running tox -esample_config on that?15:56
morganfainbergi mean... since config.py was updated.15:56
dstaneksure15:57
morganfainbergand do we want to land that in K1? ( dstanek, lbragstad, bknudson1)? or try and sneak it in during RC? It looks mostly like stuff we could drop in K1 and look to backport (mostly harmless)15:57
morganfainbergi'm leaning towards K1 just based on how backed up the gate is.15:57
morganfainbergbut if overwhelminly this needs to land in J, i'd consider it15:58
dstanekthat would say k1 because it's mostly cosmetic in nature15:58
morganfainbergdstanek, ++ ok i'm goint to toss a -2 on it with a comment for that then.15:59
*** gyee has joined #openstack-keystone16:03
morganfainbergbknudson1, that requirements update looks like it might be the last thing we need (after the k2k is expirimental and marek's "add version to saml") patches16:06
morganfainbergbknudson1, am i correct in that?16:06
openstackgerritDavid Stanek proposed a change to openstack/keystone: Address some late comments for memcache clients  https://review.openstack.org/12444316:07
bknudson1morganfainberg: I can't think of anything that's really required.16:09
morganfainbergbknudson1, cool and i assume that requirements (kombu) update is just putting us in line with everyone else?16:09
bknudson1morganfainberg: it was accepted into global requirements16:09
morganfainbergah and we don't have a juno branch yet on reqs16:10
morganfainbergok16:10
morganfainbergyep16:10
*** jaosorior has quit IRC16:12
*** __TheDodd__ has joined #openstack-keystone16:16
*** thedodd has quit IRC16:17
*** marcoemorais has joined #openstack-keystone16:17
*** __TheDodd__ has quit IRC16:20
henrynashstevemar: commented on the new openstack cli examples…there appears to be no mention of domains in any of the apis?  seems odd…16:21
stevemarhenrynash, I just wanted parity with what's there. I can add a new section (v3 vs v2 for OSC) and show domains/groups there16:22
henrynashstevemar: ok, maybe add a follow on patch, maybe to add that?16:22
henrynashstevemar: or if you want to do it in the same patch, that;s fine too16:22
stevemarhenrynash, OK just getting your comments now - so for V2 project create and user create we don't need domain (and like I said, just wanted parity)16:23
henrynashstevemar: I assume the openstack client defautls to the default domain if you don’t specify?16:23
stevemarhenrynash, of course16:23
*** marcoemorais has quit IRC16:23
stevemarhenrynash, i'll add a follow on patch, i don't like bogging down the reviews with 100s of lines to review16:24
*** marcoemorais has joined #openstack-keystone16:24
henrynashstevemar: agreed16:24
*** marcoemorais has quit IRC16:24
*** marcoemorais has joined #openstack-keystone16:24
*** marcoemorais has quit IRC16:25
*** marcoemorais has joined #openstack-keystone16:26
openstackgerritAndre Aranha proposed a change to openstack/keystone: Creating a policy sample  https://review.openstack.org/12350916:26
*** henrynash has quit IRC16:28
*** wwriverrat has joined #openstack-keystone16:29
*** anteaya has joined #openstack-keystone16:34
*** zzzeek has quit IRC16:39
*** thedodd has joined #openstack-keystone16:40
*** zzzeek has joined #openstack-keystone16:40
marekdmhu: hey16:46
*** ayoung-afk is now known as ayoung16:48
*** thedodd has quit IRC16:51
*** Dafna has quit IRC16:53
mhumarekd: hellp16:54
mhuhello16:54
marekdi am now playing with your patch for OSC (auth plugins)16:55
marekdlooks very nice.16:55
*** ukalifon1 has joined #openstack-keystone16:55
marekdI am wondering if you think we need some kind a wrapper for v3unscopedsaml and v3scopedsaml or you are planning to cover it in osc?16:55
*** andreaf has quit IRC16:55
*** andreaf has joined #openstack-keystone16:56
marekdmhu: in other words: do you think https://review.openstack.org/#/c/106751/ will be useful16:56
marekdstevemar: ^^ question for you too..16:56
mhumarekd, I think your wrapper is the way to go16:56
mhumaking it available at the library level makes more sense I think16:57
*** packet has joined #openstack-keystone16:57
marekdstevemar: mhu i think we might want it in keystoneclient, so osc doesn;t need to resolve any dependencies or worse, error handling. I was also thinking to use the wrapper to use either  shibboleth or adfs plugin for unscoped tokens and handle tthis somewhat behind the scenes.16:57
marekdmhu: ok, so we have consensus on this.16:57
mhuyup, you got my +1 on that !16:57
marekdmhu: thank you!16:58
*** ukalifon has quit IRC16:58
*** marcoemorais has quit IRC16:58
*** henrynash has joined #openstack-keystone16:58
*** marcoemorais has joined #openstack-keystone16:59
*** thedodd has joined #openstack-keystone16:59
*** wwriverrat has left #openstack-keystone16:59
*** marcoemorais has quit IRC16:59
*** marcoemorais has joined #openstack-keystone17:00
stevemarmarekd, mhu well that was resolved quickly... but yeah, keep it in KSC for the reasons marek suggests17:01
marekdstevemar: ++17:03
henrynashstevemar: so what determins, for openstack client, ehther ‘project create’ is a v2 or v3 command…the url set in —os-url ?17:03
*** f13o has quit IRC17:04
henrynashstevemar: if so, I’m not sure we ever actually say that in configuration.rst17:05
*** marcoemorais has quit IRC17:05
*** marcoemorais has joined #openstack-keystone17:06
*** sigmavirus24 is now known as sigmavirus24_awa17:08
*** henrynash has quit IRC17:14
*** henrynash has joined #openstack-keystone17:17
*** dhellmann is now known as dhellmann_17:18
openstackgerritA change was merged to openstack/keystone: Mark k2k as experimental  https://review.openstack.org/12410717:25
stevemarhenrynash, the variable OS_IDENTITY_API_VERSION17:25
stevemarhenrynash, it defaults to v2.0 for now, i'll mention it in another patch17:26
henrynashstevemar: ok, got it, thanks17:26
marekdkorek17:29
*** marcoemorais has quit IRC17:30
*** marcoemorais has joined #openstack-keystone17:31
*** praneshp has joined #openstack-keystone17:31
*** __TheDodd__ has joined #openstack-keystone17:32
*** thedodd has quit IRC17:32
*** thedodd has joined #openstack-keystone17:35
*** dhellmann_ is now known as dhellmann17:36
*** henrynash has quit IRC17:38
*** __TheDodd__ has quit IRC17:38
*** __TheDodd__ has joined #openstack-keystone17:41
*** thedodd has quit IRC17:44
*** __TheDodd__ has quit IRC17:50
*** thedodd has joined #openstack-keystone17:57
*** _cjones_ has quit IRC17:58
*** _cjones_ has joined #openstack-keystone17:59
*** harlowja_away is now known as harlowja17:59
*** gokrokve_ has joined #openstack-keystone18:01
*** sigmavirus24_awa is now known as sigmavirus2418:02
*** _cjones_ has quit IRC18:03
*** gokrokve_ has quit IRC18:05
*** gokrokve has quit IRC18:05
*** gokrokve has joined #openstack-keystone18:05
*** morgan_remote_ has joined #openstack-keystone18:07
*** gokrokve has quit IRC18:10
gordcstevemar: is federation implemented in juno?18:14
rodrigodsgordc, k2k you mean?18:15
rodrigodsor just regular federation?18:15
gordcwhat's the use case/definition of k2k18:15
gordcrodrigods: ^18:16
rodrigodsk2k is keystone to keystone federation, the ability to federate multiple keystones18:16
rodrigodswhich, is implemented in Juno as experimental18:16
rodrigodsand regular federation is available since icehouse18:16
gordcrodrigods: cool cool. wanted to make sure it wasn't some super fancy keystone term.18:17
stevemarrodrigods is doing a fantastic job in my absence18:17
gordcstevemar: go away.lol18:17
stevemargordc, as you wish!18:17
gordcnooo!!!18:17
gordcok. so it's experimental... and i assume requires v3 api enabled in all projects?18:18
rodrigodsgordc, yes, it does18:19
gordcwhen we say experimental, does that mean "it works, but might break" or "we've no idea if it works"18:19
rodrigodsstevemar, right?18:19
rodrigods=)18:19
gordceither of you know which projects actually use v3 or have the ability to use v3?18:19
rodrigodsi think that it requires v3 only for keystone, i might be wrong, though18:20
rodrigodsgordc, about being experimental: https://review.openstack.org/#/c/124107/2/doc/source/configure_federation.rst18:22
stevemargordc, i think just heat as of today, still18:22
stevemargordc, experimental == stuff might break, but we plan on continuing development anyway18:23
gordcstevemar: cool cool. i'll take a look at heat. i think we have v3 code in ceilometer but we don't actually use it.18:23
stevemarlikely18:23
gordcrodrigods: stevemar: thanks... will come back with more tedious questions later.18:24
*** radez_g0n3 is now known as radez18:24
rodrigodsstevemar, btw, if you need some extra hands to help with it, i'm available =)18:28
stevemarrodrigods, just testing it all out :(18:28
stevemarrodrigods, manual test, tempest test, functional tests, anything really18:28
stevemarwe need federation to be more stable18:28
rodrigodsstevemar, i absolutely can do that18:29
gordcstevemar: i might have time to help...18:29
rodrigodswe have a "regular" federation deployment here18:29
stevemarrodrigods, nice18:29
gordcstevemar: if you give me back my vpn access.18:29
stevemarrodrigods, how painful was it to set up? how beneficial has it been?18:29
stevemargordc, send you the cert later18:29
rodrigodsstevemar, biggest issues were related about shibboleth setup18:30
gordcstevemar: cool cool. all about trusts.18:30
rodrigodsand we did it only to investigate about it's status, check if it's nice and stable18:30
stevemarrodrigods, i figured that would be the pain point. what about the mapping and authentication18:30
morgan_remote_lbragstad, bknudson1: we might want to get the memcache pool log override in for Juno.18:31
ayoungstevemar, morgan_remote_ I have a better idea for K2K18:31
rodrigodsstevemar, this was easier, no issues about that as I remember now18:31
morgan_remote_Or tag it for a Juno stable back port18:31
*** gokrokve has joined #openstack-keystone18:31
bknudson1keystonemiddleware doesn't even have support for v3 auth.18:33
bknudson1morgan_remote_: is there a review?18:33
morgan_remote_bknudson1: no not yet. I'll make sure we have a bug and tag it for stable back port potential.18:33
gordcbknudson1: so k2k is not possible or does it not matter?18:33
rodrigodsstevemar, bknudson1 gordc i'm planning to make a test deployment here (k2k), is it possible?18:34
rodrigodseven with the known issues18:34
bknudson1I didn't know there were any known issues with k2k.18:34
ayoungbknudson1, it only does Identity18:35
ayoungI have a plan for assignments...18:35
bknudson1what more do you need?18:35
gordcrodrigods: i'd be interested in results.18:36
morgan_remote_ayoung: isn't that the whole reason we were doing the local token? How does local keystone know the remote's structure / project / roles?18:36
ayoungmorgan_remote_, that is my plan...18:36
ayoungmorgan_remote_, "shared domains"18:36
ayoungso  K1 creates the domain, and exports the data to K218:36
*** andreaf has quit IRC18:36
*** gokrokve has quit IRC18:36
stevemarrodrigods, i think marekd found 2 bugs that are being merged / gating right now18:36
morgan_remote_I'll be upfront and say I'm skeptical due to data sync needs but willing to see what you're proposing18:36
ayoungK2 says "Ok, K1 can sign for only tokens scoped to these domains or proejcts in them"18:37
*** andreaf has joined #openstack-keystone18:37
ayoungmorgan_remote_, it involves Multiple signers...all sorts of good sh....stuff18:37
morgan_remote_Not opposed to the concept of we can make it robust and not-too-onerous to manage18:37
stevemarrodrigods, one is gating, the other is this one: https://review.openstack.org/#/c/124176/18:37
morgan_remote_S/of/if18:38
ayoungmorgan_remote_, I think it won't be too bad.  still crafting the message18:38
morgan_remote_ayoung: sure. Looking forward to more details then.18:39
ayoungmorgan_remote_, in keystone.egg-info/PKG-INFO  I see the value Version: 2014.2.dev154.g1af2428.  I know the last part is based on the git hash.  Any idea what dev154  is?18:40
morgan_remote_stevemar: if I read the bug correctly as long as the id would match the conf value it works. It's a config issue vs a never could work issue. And the fix makes the ux more expected.18:41
openstackgerritSteve Martinelli proposed a change to openstack/keystone: Add v3 openstackclient CLI examples  https://review.openstack.org/12448918:41
stevemarmorgan_remote_, i think for the sanity of the deployer it should be merged18:42
morgan_remote_ayoung: hmm. I'd have to look into pbr and setuptools (dark magic). Dev Is either "installed from git" or pip install -e / setup develop18:42
morgan_remote_The number I don't know18:42
ayoungmorgan_remote_, I'm digging through PBR now18:43
morgan_remote_stevemar: but my assessment is correct right?  Not disagreeing that it should be merged. Looking at "is this never going to work" or is it "bad us/non-intuitive behavior"18:43
morgan_remote_S/us/ux18:44
morgan_remote_ayoung: you likely could just ask lifeless or mordred.18:44
morgan_remote_Or clarkb18:44
ayoungmorgan_remote_, yeah.18:44
ayoungand I suspec that this is somethi pbr inherits, not generates18:45
stevemarmorgan_remote_, yes your assessment is correct18:45
morgan_remote_I *still* want to circle back and make sample config generated on "install"18:45
ayoungmorgan_remote_, and I want to make an RPM not have to wipe out the PBR data18:45
morgan_remote_Which is going to take a pbr fix afaict18:45
morgan_remote_ah that sounds like worth fixing!18:46
ayoungmorgan_remote_, yeah....so close18:46
ayoungmorgan_remote_, I think that, if I can get the RPM and setup.py version in sync, the pbr issue goes away18:46
ayoungright now, if I leave pbr in, when I go to build the rpm I get18:46
morgan_remote_stevemar: ok. Sounds good. Let's look at where we sit and start classifying thing a as "must get in for rc" and stable Juno potential.18:47
ayoung+ /usr/bin/python setup.py build18:47
ayoungerror in setup command: Error parsing /home/ayoung/rpmbuild/BUILD/keystone-2014.2.b3/setup.cfg:Exception: Versioning for this project requires either an sdist tarball, or access to an upstream git repository. Are you sure that git is installed?18:47
morgan_remote_ayoung: ugh. That isn't fun! :(18:47
stevemarmorgan_remote_, my vote is that it's 'must get in for rc'18:47
ayoungmorgan_remote_, no,  but its made me angry enough that HULK SMASH!18:47
stevemarmorgan_remote_, were you able to reproduce that bug on 12.04 last night?18:48
morgan_remote_stevemar: I'll be back at my computer soon. (Finishing my breakfast/lunch/coffee). Hard to see a complete view from my phone (though the big screen does help make it less icky)18:48
morgan_remote_stevemar: have to do t today. Was at openstack la meetup till like 10. Too tired to do VM wrangling last night.18:49
ayoungmorgan_remote_, oooh...so speaking of  generating the config file options, it turns out we modify the sample conf on build,  with a patch....18:50
* ayoung needs to talk to apevec18:51
ayoungmorgan_remote_, new rule of thumb...if you are carrying long term patches in your distro package specification, you are in a state of sin18:52
*** ukalifon1 has quit IRC18:53
morgan_remote_ayoung: can't disagree.18:53
*** rm_work is now known as rm_work|away18:57
*** dhellmann is now known as dhellmann_18:57
*** rm_work|away is now known as rm_work18:57
*** jorge_munoz has quit IRC18:58
*** rushiagr_away is now known as rushiagr19:11
dstaneklbragstad: you beat me to the userId/username review :P19:13
lbragstaddstanek: :)19:13
lbragstadyou can still review it if you want, I tested the logic19:13
dstaneklbragstad: i think there may be tests missing so i'm going to post a follow up, but it's not an RC blocker19:13
lbragstadok19:13
*** _cjones_ has joined #openstack-keystone19:18
*** gordc has quit IRC19:19
*** zzzeek has quit IRC19:21
*** andreaf has quit IRC19:21
*** andreaf has joined #openstack-keystone19:22
*** boris-42 has joined #openstack-keystone19:27
*** saipandi has joined #openstack-keystone19:27
*** rushiagr is now known as rushiagr_away19:29
*** gokrokve has joined #openstack-keystone19:32
*** nkinder_ has quit IRC19:33
*** nkinder has joined #openstack-keystone19:33
morganfainbergstevemar, ok looking at that migration issue here now.19:35
*** gokrokve has quit IRC19:37
morganfainbergstevemar, haha looks like someone took topol's name on IRC just sent a message to the wrong person with some weird anti-spam thing19:41
morganfainbergstevemar, he should register it and enable nick protection :P19:42
*** andreaf has quit IRC19:44
*** andreaf has joined #openstack-keystone19:45
*** amcrn has joined #openstack-keystone19:50
*** hurgleburgler has joined #openstack-keystone19:51
morganfainbergoh boy: https://bugs.launchpad.net/keystone/+bug/137449719:52
uvirtbotLaunchpad bug 1374497 in oslo.db "change in oslo.db "ping" handling is causing issues in projects that are not using transactions" [High,In progress]19:52
*** dhellmann_ is now known as dhellmann19:55
*** gordc has joined #openstack-keystone19:55
dstanekmorganfainberg: wow, that's not good19:57
morganfainbergdstanek, that ... is potentially an ugly fix if we need to roll that in now19:59
lbragstadI remember seeing a review come through for that a long time ago19:59
dstanekmorganfainberg: is that a real problem that people are seeing or is it a theortical problem?20:00
lbragstadisn't it part of the reason we are seeing the "MySQL server has gone away" bug?20:00
lbragstadhttps://bugs.launchpad.net/keystone/+bug/136137820:00
uvirtbotLaunchpad bug 1361378 in oslo.db ""MySQL server has gone away" again" [Undecided,Incomplete]20:00
dstaneklbragstad: hmmm...yes i do remember something about that20:00
lbragstadI thought I saw a review a while back where someone went through everything and changed that,20:01
lbragstadwhere we were relying on the autocommit stuff20:01
lbragstadbut, I don't remember... my review memory has a max depth of about 5 reviews.20:02
morganfainberglbragstad, i think so.20:04
morganfainberglbragstad, that looks like the root of it20:04
morganfainbergwhich... i'd much rather see another oslo.db release...but i don't see that happening20:05
stevemarmorganfainberg, hehe, topol is out today20:07
morganfainbergi know i hit him up on twitter about it20:08
*** hurgleburgler has left #openstack-keystone20:11
stevemarwow, i get to vote, apparently i contributed to cinder20:12
stevemarmust have been a small bug >.>20:12
morganfainbergstevemar, in either Icehouse or Juno20:12
stevemarprobably juno20:12
*** zzzeek has joined #openstack-keystone20:13
morganfainbergwhy ubuntu why do you make it so hard to install openssh server20:19
*** topol_ has joined #openstack-keystone20:21
*** aix has quit IRC20:21
lbragstadmorganfainberg: 14.04?20:22
morganfainberglbragstad, 12.0420:22
dstanekmorganfainberg: i have recently switched to Fedora on a trial bases20:26
dstanekor basis20:26
morganfainbergdstanek, except i am trying to duplicate a 12.04 bug20:26
morganfainberg;)20:26
lbragstadah.. that makes sense20:27
*** rwsu has quit IRC20:31
*** gokrokve has joined #openstack-keystone20:32
*** zohar has joined #openstack-keystone20:34
zoharHi all20:35
zoharI am trying to do a simple thing, I have a single node which I want to use as an authentication endpoint20:35
zoharon ubuntu1420:35
zohari do apt-get install keystone20:35
zoharconfigure /etc/keystone/keystone.conf20:35
zoharservice keystone restart20:36
*** gordc has quit IRC20:36
*** NM1 has joined #openstack-keystone20:36
zoharassign some env vars like OS_USERNAME/TENANT/PASSWORD/AUTH_URL20:36
zoharthen a keystone tenant-list returns this:20:36
*** morganfainberg is now known as morgan20:36
*** morgan is now known as morganfainberg20:37
*** gokrokve has quit IRC20:37
zoharAuthorization Failed: Unable to establish connection to http://127.0.0.1:35357/v2.0/tokens20:37
zoharwhat am i missing?20:37
*** raildo has quit IRC20:39
nkinderzohar: what is your OS_AUTH_URL set to?20:39
morganfainbergdstanek, ok https://bugs.launchpad.net/keystone/+bug/1374497 talked w/ dhellmann and zzzeek in #openstack-oslo this is something we need to fix in kilo, but oslo.db should get fixed for us for the Juno timeframe20:40
uvirtbotLaunchpad bug 1374497 in oslo.db "change in oslo.db "ping" handling is causing issues in projects that are not using transactions" [High,In progress]20:40
dstanekmorganfainberg: cool, that works for me20:41
*** thedodd has quit IRC20:45
morganfainbergayoung, is this still relevant? https://bugs.launchpad.net/keystone/+bug/1208588 or is it not based on your new kerberos work?20:46
uvirtbotLaunchpad bug 1208588 in keystone "Support getting Auth attributes from Kerberos PAC" [Wishlist,New]20:46
ayoungKILL THAT20:46
ayounger...20:46
ayoungnah, I guess it is still valid20:46
ayoungit would be a Federation type use case20:46
morganfainbergis it a bug?20:46
ayoungWishlist20:47
morganfainbergor is it a -spec? or is it something else?20:47
morganfainbergok so.. we *want* it?20:47
ayoungmorganfainberg, I think, maybe, we can close it.  the PAC would be unpacked by Apache20:47
morganfainbergok i'll close it for now, and we can re-open if needed20:47
ayoungso nothing that Keystone would have to do but treat it like another Federation case20:47
morganfainbergor attach it where we need it.20:47
morganfainbergk20:47
ayoungYeah, close20:48
morganfainbergooh we're close to having no untriaged (new state) bugs in keystone :)20:48
*** thedodd has joined #openstack-keystone20:49
*** topol_ is now known as bradtopol20:50
*** radez is now known as radez_g0n320:50
zoharnkinder, sorry for late response, it's just http://127.0.0.1:35357/v2.0/20:53
zoharsince i want to run it on the same node20:53
*** dimsum_ has quit IRC20:56
zohari think i am missing an the actual apache server to receive requests at that address20:56
*** dimsum_ has joined #openstack-keystone20:56
zoharwhat would be the easiest way to setup a working keystone authentication endpoint on a fresh node?21:00
*** dimsum_ has quit IRC21:00
*** marcoemorais has quit IRC21:07
*** marcoemorais has joined #openstack-keystone21:07
nkinderzohar: is keystone running?  (ps -ef | grep keystone-all)21:10
nkinderzohar: that should show your keystone server (unless you've deployed it in httpd w/mod_wsgi)21:11
*** amcrn_ has joined #openstack-keystone21:11
*** amcrn has quit IRC21:12
zoharyes its running21:12
zoharnkinder, it's running. ive been looking now into this: https://github.com/swiftstack/keystone_install21:12
zoharit would be nice to go through all the steps myself though to get better understanding of setting this up21:13
nkinderzohar: it's odd that you get a connection error though.  Does 'netstat -an | grep 35357' show that it's listening on the right interfaces?21:15
zoharnkinder, ill get back to you on this in a second, let me recreate this on another node, ive run the keystone_install thing on my previous one21:16
zoharnkinder, without changing keystone.conf, i get this from netstat: tcp        0      0 0.0.0.0:35357           0.0.0.0:*               LISTEN21:19
*** bradtopol is now known as btopol21:19
*** btopol is now known as topol21:20
zoharoh wow, and after changing keystone.conf and doing service keystone restart, it is not listening anymore21:20
*** stevemar has quit IRC21:20
*** morgan_remote_ has quit IRC21:20
zoharoh oops21:20
*** dimsum_ has joined #openstack-keystone21:21
zoharnkinder, i fixed the conf, now netstat returns this: tcp        0      0 <correct_ip>:35357       0.0.0.0:*               LISTEN21:22
nkinderzohar: ok, so can you do a 'keystone tenant-list' now?21:24
nkinderzohar: with all of your proper ENV vars set...21:24
zoharnope, same error, unable to establish connection21:24
zoharalso, the keystone_install "autoinstall" shell script didnt work :P21:24
*** marcoemorais has quit IRC21:25
zohareither way i rather do it by hand21:26
nkinderzohar: so <correct_ip> is a real IP, or is it 127.0.0.1?21:27
zoharreal ip21:27
zoharis that the reason?21:27
nkinderzohar: if it's not 127.0.0.1 or 0.0.0.0, you're not listening on the loopback21:27
zoharoh wow21:28
nkinderzohar: you can connect over the real IP by setting OS_AUTH_URL to use it21:28
zoharyes i just did21:28
zoharnow next issue21:28
zoharCould not find user, admin. (HTTP 401)21:29
zohar:P21:29
zohari guess its time to create a user :D21:29
nkinderzohar: it you want to listen on all addresses, set admin_bind_host/public_bind_host to 0.0.0.0 in keystone.conf21:29
nkinderzohar: yeah, you can use the admin_token in keystone.conf to connect and set up all of your users/roles/projects21:30
zoharok thank you21:30
zoharnow here is my big picture task21:31
zoharim trying to set up a minimal test node, which will point to an object-storage-only cluster21:31
*** andreaf has quit IRC21:31
zoharand run some tests on the test node side over swift-api21:31
*** andreaf has joined #openstack-keystone21:32
*** thedodd has quit IRC21:32
zoharso im trying to put together everything necessary for this test enviornment, and keystone is the first point21:32
*** gokrokve has joined #openstack-keystone21:32
zoharon the remote end, there is a working swift api and keystone authentication21:33
*** gokrokve has quit IRC21:33
zoharso once i get keystone auth server working on the test node, i will add its tenants to the remote cluster and point to it via service endpoint21:33
*** gokrokve has joined #openstack-keystone21:34
*** gokrokve has quit IRC21:38
*** packet has quit IRC21:39
*** joesavak has quit IRC21:40
*** marcoemorais has joined #openstack-keystone21:44
*** marcoemorais has quit IRC21:46
*** marcoemorais1 has joined #openstack-keystone21:46
*** david-lyle has quit IRC21:56
*** marcoemorais1 has quit IRC21:57
*** rkofman has quit IRC21:58
*** rkofman has joined #openstack-keystone21:58
morganfainberghttps://bugs.launchpad.net/keystone/+bug/1371620 is SQLite specific21:58
uvirtbotLaunchpad bug 1371620 in keystone "Setting up database schema with db_sync fails with OperationalError: (OperationalError) database is locked u'DELETE FROM user_project_metadata' ()" [Undecided,Confirmed]21:58
morganfainbergI believe this is a similar issue to what we had with 042 .21:59
*** marcoemorais has joined #openstack-keystone22:02
*** imkarrer has joined #openstack-keystone22:02
*** marcoemorais has quit IRC22:02
*** marcoemorais has joined #openstack-keystone22:04
imkarrerHey, I am looking for information on extensions within Keystone IceHouse release.  The documentation states that keystone ships with extensions in keystone/contrib/extensions.  I would like to use the OS-KSADM but it is not clear how to install it or which releases support this extension.  Could anyone point me in the right direction?22:05
morganfainberghm22:10
morganfainbergdstanek, might have a question for you22:10
dstaneki may have an answer for you22:11
morganfainbergdstanek, so migration 3922:11
morganfainbergdstanek, https://github.com/openstack/keystone/blob/master/keystone/common/sql/migrate_repo/versions/039_grant_to_assignment.py#L85 this fails in SQLite22:11
morganfainbergi think because of table locks22:11
morganfainbergi'm not *sure* why we're needing to delete all the rows, the next migration we drop the tables.22:11
morganfainbergthats the root cause of https://bugs.launchpad.net/keystone/+bug/1371620 and it *only* affects SQLite22:12
uvirtbotLaunchpad bug 1371620 in keystone "Setting up database schema with db_sync fails with OperationalError: (OperationalError) database is locked u'DELETE FROM user_project_metadata' ()" [Undecided,Confirmed]22:12
morganfainbergi'm not sure how we're passing unit tests tbh22:12
dstanekmorganfainberg: is it a sqlite version thing?22:13
*** sigmavirus24 is now known as sigmavirus24_awa22:13
morganfainbergas in sqlite version? or schema version22:13
dstaneksqlite version22:13
morganfainbergi think we'd run into this earlier if it was sqlite specific22:13
morganfainbergerm version specific22:13
morganfainbergbecause we did test on 12.04 for a long time22:13
*** NM1 has quit IRC22:14
morganfainbergunless a minor version introduced a regression?22:14
*** topol has quit IRC22:14
dstanekthe only two things i've ever seen cause a locked issue was messing with tables on an ancient version of sqlite and when the DB file was actually on a share instead of local disk22:15
morganfainbergthis is on local disk22:16
morganfainbergand this is a "clean" 12.04 install22:16
morganfainbergas in i just spun it up and installed everything22:16
dstanekso you can reproduce?22:16
morganfainbergdstanek, yep consistently22:16
*** marcoemorais has quit IRC22:16
morganfainbergworks just fine in relation dbs.22:16
morganfainbergyou know.. let me distupgrade here and 2x check22:17
*** david-lyle has joined #openstack-keystone22:17
morganfainbergbut SQLite seems special.22:17
dstanekwhat version of sqlite is installed?22:18
*** andreaf has quit IRC22:19
*** andreaf has joined #openstack-keystone22:19
dstanekmorganfainberg: ok, i did a phone a friend to get a better answer. when we were using that old version the real problem was that we had 2 processing doing stuff in the DB and that caused the issue. we upgraded to fix.22:19
morganfainbergdstanek, ii  libsqlite3-0                                                       3.7.9-2ubuntu1.1                                                   SQLite 3 shared library22:19
morganfainbergdstanek, this is the packaged version of sqlite with 12.0422:20
morganfainberglatest version22:20
*** david-lyle has quit IRC22:21
morganfainbergi'm thinking it's a list operation that hasn't closed the session action yet.22:21
morganfainbergso looking a bit more closely at the migration22:21
morganfainbergbut... again i'm surprised this works in our unit tests / worked on 12.0422:22
morganfainbergooh, i think i see it.22:22
morganfainbergwe're not "Really" using a transaction but using some autocommit22:22
dstanekmorganfainberg: if that's the case i would have expected this to have come up before; very wierd22:23
morganfainbergyeah22:23
morganfainbergthis is very strange22:23
dstaneki'm trying this on a newer box22:25
morganfainbergk22:26
morganfainbergi don't have a 14.04 instance ready atm or i'd be trying it there22:26
morganfainbergwould need to rebuild.22:26
dstanekyeah, wow - fails for me too22:30
morganfainbergso, simplest solution is "don't delete the rows, we're dropping the tables in the next migration"22:31
morganfainbergbut what bugs me is that is passes unit tests obstensibly22:31
morganfainbergmaybe... this is a disk-based vs in-memory sqlite issue22:31
morganfainbergwe use in-memory iirc for everything now.22:31
morganfainbergwould be a strange differentiation imo22:32
dstaneki think i fixed it locally22:32
*** gokrokve has joined #openstack-keystone22:32
morganfainbergdstanek, how?22:32
dstanekmorganfainberg: http://paste.openstack.org/show/115749/22:33
morganfainbergdstanek, sure.22:34
morganfainbergdstanek, ok i'm good with that.22:34
dstanekmorganfainberg: what you said about the operation not being closed gave me the idea22:34
morganfainbergdstanek, right.22:34
morganfainbergthis feels like something that we should get into RC.22:34
morganfainbergand backport to icehouse.22:35
dstaneki have no idea why this hasn't failed before, but meh22:35
dstanekdo you want me to commit or do you want the honors?22:35
morganfainbergdstanek, it does fail using our docs on "setting up keystone" as outlined in the bug22:35
morganfainberggo for it, i'll review.22:35
morganfainbergtagging this to RC. and for icehouse backport.22:36
morganfainbergdstanek, thanks22:36
*** leonchio__ has joined #openstack-keystone22:37
morganfainbergdstanek, *doh* RC bugs trickling in! (at least we're finding them sooner vs later)22:37
*** gokrokve has quit IRC22:37
*** rm_work is now known as rm_work|away22:40
*** NM1 has joined #openstack-keystone22:40
*** dhellmann is now known as dhellmann_22:42
*** dimsum_ has quit IRC22:43
*** henrynash has joined #openstack-keystone22:43
*** dimsum_ has joined #openstack-keystone22:44
dstanekmorganfainberg: pushing now - this feel like a hack because sqlite shouldn't be breaking like this22:45
morganfainbergi agree.22:45
morganfainbergi'll poke zzzeek to review this.22:46
morganfainberghe might have better insight22:46
openstackgerritDavid Stanek proposed a change to openstack/keystone: Adds a commit in a migration between SQA calls  https://review.openstack.org/12453322:46
*** marcoemorais has joined #openstack-keystone22:46
morganfainbergzzzeek, ^ if you don't mind.22:47
*** rwsu has joined #openstack-keystone22:47
zzzeekmorganfainberg: how come22:47
morganfainbergzzzeek, something is weird with SQLite in this case.22:47
morganfainbergit's failing in 039 migration and it looks like something is being held open before the delete.22:47
morganfainbergthe commit closes it out and make the migration work22:48
zzzeekwell youre mixing engines and sessions that arent linked together again, it seems :)22:48
morganfainbergbut this feels weird, because theorestically this should have failed unit tests.22:48
morganfainbergoh22:48
* zzzeek sings the happy alembic doesnt have this problem dance22:48
morganfainbergzzzeek, can has?22:48
morganfainbergzzzeek, :)22:48
zzzeekat the top of the whole script:  with engine.begin as conn:22:48
*** dimsum_ has quit IRC22:48
zzzeekthen “conn” is it, if you make a Session, say, Session(bind=conn)22:48
zzzeekeverything on that conn22:49
zzzeekengine.begin() that is22:49
zzzeekone connection for the whole thing22:49
morganfainbergah, so we should just engine.begin and it should also solve it?22:49
morganfainbergand use it for the session22:49
morganfainbergok let me try that. i like that better than the commits22:50
morganfainbergzzzeek hope we can move to alembic for keystone in Kilo. would make me happy22:51
zzzeekmorganfainberg: OK but if we just do decent connectivity in scripts, that’ll be almost as good :)22:51
morganfainbergfair enough22:52
dstanekzzzeek: thanks, i am trying that out22:54
morganfainbergdstanek, i know why it doesn't fail in tests22:56
morganfainbergdstanek, we explicitly pass sessions22:56
morganfainbergand connections22:56
praneshpdolphm: I see you assigned https://bugs.launchpad.net/cinder/+bug/1306559?comments=all to me. I’ll work on it, but was curious why you picked me out.23:02
uvirtbotLaunchpad bug 1306559 in keystone "Fix python26 compatibility for RFCSysLogHandler" [Low,Confirmed]23:02
*** marcoemorais has quit IRC23:05
*** dimsum_ has joined #openstack-keystone23:07
*** praneshp has quit IRC23:10
morganfainberglbragstad, ping23:13
*** _cjones_ has quit IRC23:14
morganfainberglbragstad, is this solved with your validation patch? https://bugs.launchpad.net/keystone/+bug/1244423 or .. am i mis-reading this?23:14
uvirtbotLaunchpad bug 1244423 in keystone "Inconsistency in the keystone api "enabled" field" [Undecided,Triaged]23:14
morganfainberghm23:14
*** _cjones_ has joined #openstack-keystone23:15
*** _cjones_ has quit IRC23:19
*** marcoemorais has joined #openstack-keystone23:20
*** _cjones_ has joined #openstack-keystone23:24
dstanekzzzeek: i just tested in my env and the one reported in the bug - all i had to do to get around the error was to the the existing session instead of the session and engine23:27
zzzeekum OK23:27
zzzeekdidnt look too closely but when i see “migrate_engine” and “session” mixed up i know somtehings off23:28
morganfainbergdstanek, lbragstad, ayoung, are we supporting V3 constructs in keystoneclient, i think we determined we arent: https://bugs.launchpad.net/python-keystoneclient/+bug/136786823:29
uvirtbotLaunchpad bug 1367868 in python-keystoneclient "List inherited role assignments for domains available on keystone API but not on client" [Undecided,In progress]23:29
morganfainberggyee, dolphm, ^23:29
*** rwsu has quit IRC23:30
morganfainbergbknudson1, ^23:30
dstanekmorganfainberg: did we determine we didn't want to do that? i voted a +1 or +2 or a related review i think23:30
morganfainbergdstanek, i'm fine if we do want to, but trying to get the bugs that are in limbo dealt with in LP23:31
morganfainbergthis is keystonelclient fix that hasn't gone anywhere23:31
morganfainbergs/fix/bug23:31
morganfainbergreport23:31
openstackgerritDavid Stanek proposed a change to openstack/keystone: Uses session in migration to stop DB locking  https://review.openstack.org/12453323:32
gyeemorganfainberg, client SDK or CLI23:32
gyee?23:32
* zzzeek hopes dstanek is in PDT time23:32
dstanekzzzeek: can you take a quick peek and see if i'm crazy ^23:32
zzzeekdstanek: sure ! :)23:32
dstanekzzzeek: nope, EST here23:32
zzzeekreallly23:32
zzzeekdude23:32
*** gokrokve has joined #openstack-keystone23:32
gyeewe are not doing anymore V3 CLI stuff in keystoneclient23:32
zzzeek7:30 friday23:32
morganfainbergzzzeek, dstanek is crazy like that.23:32
zzzeekdrinky23:32
dstanek...maybe i've already started...23:33
zzzeekdstanek: yeah thats OK session.execute() will use the ongoing context23:33
zzzeekdstanek: if it works, you’re golden23:34
zzzeekgotta go out to meet the wife23:34
dstanekzzzeek: i really need to learn more about SQA. i just know the basics because i've been traditionally a django guy23:34
dstanekzzzeek: thanks for you help23:34
dstanekdinner time for me now23:35
*** gokrokve has quit IRC23:36
*** NM1 has quit IRC23:37
*** zzzeek has quit IRC23:44

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