| *** troytoman is now known as troytoman-away | 00:01 | |
| *** flaper87 has quit IRC | 00:01 | |
| *** colinmcnamara has quit IRC | 00:07 | |
| *** stevebaker has quit IRC | 00:11 | |
| *** rpedde is now known as rpedde_away | 00:11 | |
| *** stevebaker has joined #openstack-dev | 00:12 | |
| *** tomoe_ has quit IRC | 00:14 | |
| *** datsun180b has quit IRC | 00:16 | |
| *** stevebaker has quit IRC | 00:19 | |
| *** cloudchimp has quit IRC | 00:23 | |
| topol | ayoung, so I am debugging the devstack ldap integration. I noticed something nasty with the existing code that adds tenants to projects. Objectclass groupOfNames requires a member and a desc. When I run from devstack, description (which maps to desc is None) and member isn't a listed attribute. How did this work before with mapping tenants to something with an ObjectClass groupofNames. I... | 00:27 |
|---|---|---|
| topol | ...was hoping to get this working by just changing values in the keystone.conf file but I fear I may need to update the mappings in the ldap/core.py file. Am I missing something? | 00:27 |
| *** stevebaker has joined #openstack-dev | 00:27 | |
| *** dontalton has quit IRC | 00:27 | |
| Ryan_Lane | description isn't a required attribute of groupofnames | 00:28 |
| *** donaldh has quit IRC | 00:28 | |
| Ryan_Lane | member also isn't required | 00:28 |
| Ryan_Lane | only cn is | 00:28 |
| topol | Ryan_Lane here is the error I get when I run: OBJECT_CLASS_VIOLATION: {'info': "object class 'groupOfNames' requires attribute 'member'", 'desc': 'Object class violation'} | 00:29 |
| *** pabelanger has quit IRC | 00:30 | |
| Ryan_Lane | what directory server are you using? | 00:30 |
| Ryan_Lane | that makes no sense. how would you have empty groups? | 00:30 |
| *** alszar has quit IRC | 00:30 | |
| topol | openldap | 00:31 |
| Ryan_Lane | well, holy crap, the rfc says its required | 00:31 |
| topol | :-) | 00:31 |
| *** roampune1 has joined #openstack-dev | 00:33 | |
| Ryan_Lane | I definitely have groups with no member, though | 00:33 |
| *** eharney has quit IRC | 00:33 | |
| *** roampune has quit IRC | 00:35 | |
| *** bencherian has joined #openstack-dev | 00:36 | |
| Ryan_Lane | hah. opendj's directory schema only defines cn as must. that seems non-rfc standard to me :D | 00:42 |
| *** jcmartin has joined #openstack-dev | 00:42 | |
| *** anteaya has quit IRC | 00:43 | |
| Ryan_Lane | http://lists.forgerock.org/pipermail/opendj/2012-November/002354.html \o/ | 00:43 |
| openstackgerrit | A change was merged to openstack/keystone: Workaround Migration issue with PostgreSQL https://review.openstack.org/22215 | 00:46 |
| openstackgerrit | A change was merged to openstack/keystone: Add pysqlite as explicit test dep https://review.openstack.org/22066 | 00:46 |
| openstackgerrit | A change was merged to openstack/keystone: project membership to role conversion https://review.openstack.org/21327 | 00:47 |
| *** digitalsanctum has quit IRC | 00:49 | |
| *** rnirmal has quit IRC | 00:50 | |
| *** sacharya has joined #openstack-dev | 00:52 | |
| *** jergerber has quit IRC | 00:53 | |
| *** roampune1 has quit IRC | 00:53 | |
| *** yamahata has quit IRC | 00:56 | |
| *** apmelton has quit IRC | 01:00 | |
| adam_g | is the glance_api_servers flag in nova.conf still relevant, or can that be set to resolve via the keystone catalog? | 01:01 |
| openstackgerrit | A change was merged to openstack/keystone: Remove usage of UserRoleAssociation.id in LDAP https://review.openstack.org/19815 | 01:01 |
| *** sthaha has joined #openstack-dev | 01:02 | |
| openstackgerrit | A change was merged to openstack/nova: remove intermediate libvirt downloaded images https://review.openstack.org/21955 | 01:02 |
| *** winston-d_ has joined #openstack-dev | 01:02 | |
| openstackgerrit | A change was merged to openstack/nova: Add libvirt XML schema support for resource tuning parameters https://review.openstack.org/22103 | 01:02 |
| *** jaybuff1 has joined #openstack-dev | 01:03 | |
| *** jaybuff has quit IRC | 01:05 | |
| *** jaybuff1 has quit IRC | 01:07 | |
| *** mrodden has quit IRC | 01:09 | |
| *** shang has joined #openstack-dev | 01:09 | |
| *** yamahata has joined #openstack-dev | 01:09 | |
| *** bknudson has quit IRC | 01:11 | |
| *** darjeeling has quit IRC | 01:12 | |
| *** stevebaker has quit IRC | 01:12 | |
| *** stevebaker has joined #openstack-dev | 01:12 | |
| *** imsplitbit has quit IRC | 01:15 | |
| *** yaguang has joined #openstack-dev | 01:18 | |
| *** jog0_ has joined #openstack-dev | 01:19 | |
| *** jog0_ has quit IRC | 01:19 | |
| *** pixelbeat has quit IRC | 01:19 | |
| *** adalbas has quit IRC | 01:22 | |
| *** hemna has joined #openstack-dev | 01:22 | |
| *** jog0 has quit IRC | 01:23 | |
| *** markwash has quit IRC | 01:23 | |
| *** pixelbeat has joined #openstack-dev | 01:24 | |
| *** bdpayne has quit IRC | 01:26 | |
| ayoung | topol, so, we can drop that aspect anyway. | 01:31 |
| ayoung | we won;'t have users as memberOf anymore | 01:31 |
| ayoung | instead, it will all be done via roles | 01:31 |
| ayoung | topol just merged https://review.openstack.org/#/c/21327/ | 01:32 |
| *** devoid has quit IRC | 01:33 | |
| *** thingee is now known as thingee_zzz | 01:33 | |
| ayoung | topol, Ryan_Lane So we can stop using groupOfNames, even, but it seems to me that memberOf can be empty, no? | 01:33 |
| *** otherwiseguy has quit IRC | 01:39 | |
| openstackgerrit | A change was merged to openstack/oslo-incubator: blueprint amqp-rpc-fast-reply-queue https://review.openstack.org/19721 | 01:40 |
| *** mrodden has joined #openstack-dev | 01:40 | |
| *** Tross has quit IRC | 01:40 | |
| *** Tross has joined #openstack-dev | 01:42 | |
| *** danwent has joined #openstack-dev | 01:47 | |
| *** ek6 has quit IRC | 01:47 | |
| openstackgerrit | A change was merged to openstack/oslo-config: Make sure to install the oslo package as well. https://review.openstack.org/22250 | 01:51 |
| *** dprince has joined #openstack-dev | 01:51 | |
| *** zyluo has joined #openstack-dev | 01:52 | |
| *** jcmartin has quit IRC | 01:53 | |
| *** sthaha has quit IRC | 01:55 | |
| *** salv-orlando has quit IRC | 01:56 | |
| *** alexxu has joined #openstack-dev | 01:58 | |
| *** doude has quit IRC | 01:58 | |
| *** jcmartin has joined #openstack-dev | 01:58 | |
| zyluo | gongysh, ping | 01:58 |
| gongysh | zyluo: hi | 01:59 |
| zyluo | gongysh, do you have spare time after lunch for a patch review | 02:00 |
| gongysh | zyluo: what is it? | 02:00 |
| *** bing_bu has joined #openstack-dev | 02:00 | |
| gongysh | zyluo: wadl? | 02:00 |
| zyluo | yes that one | 02:00 |
| gongysh | zuluo: sure, I will contact u then. | 02:01 |
| zyluo | gongysh, https://review.openstack.org/#/c/21038/4..6/tools/wadl/extract_attrs.py | 02:01 |
| zyluo | gongysh, this will help you to see the history thanks! | 02:01 |
| *** pixelbeat has quit IRC | 02:02 | |
| openstackgerrit | A change was merged to openstack/nova: Add image to request_spec during resize https://review.openstack.org/22020 | 02:02 |
| *** njoy1__ has joined #openstack-dev | 02:06 | |
| *** njoy_ has quit IRC | 02:06 | |
| *** cloudchimp has joined #openstack-dev | 02:07 | |
| *** darjeeling has joined #openstack-dev | 02:11 | |
| *** cloudchimp has quit IRC | 02:12 | |
| *** pcm_ has joined #openstack-dev | 02:12 | |
| *** sthaha has joined #openstack-dev | 02:12 | |
| *** hemna has quit IRC | 02:14 | |
| *** sacharya1 has joined #openstack-dev | 02:19 | |
| *** johnpur has joined #openstack-dev | 02:19 | |
| *** sacharya has quit IRC | 02:22 | |
| *** dprince has quit IRC | 02:25 | |
| *** pcm_ has quit IRC | 02:25 | |
| *** jcmartin has quit IRC | 02:26 | |
| *** cloudchimp has joined #openstack-dev | 02:27 | |
| topol | ayoung, Im back from dinner :-). So if we drop groupOfNames then do we need to use something else? Can you fpaste what you thought the ldap structures would look like? | 02:29 |
| topol | ayoung, I will download the new code | 02:31 |
| topol | looks like a big change in structure | 02:31 |
| *** thickskin has joined #openstack-dev | 02:32 | |
| *** gary has joined #openstack-dev | 02:33 | |
| *** gary is now known as Guest87128 | 02:33 | |
| *** markmcclain has quit IRC | 02:34 | |
| *** shang_ has joined #openstack-dev | 02:35 | |
| *** Guest87128 has left #openstack-dev | 02:36 | |
| *** shang has quit IRC | 02:37 | |
| *** ajia_afk is now known as ajia | 02:40 | |
| *** baba has quit IRC | 02:41 | |
| *** bdpayne has joined #openstack-dev | 02:42 | |
| openstackgerrit | A change was merged to openstack/nova: Add and check data functions for test_migrations 141 https://review.openstack.org/21548 | 02:43 |
| *** Mandell has quit IRC | 02:49 | |
| *** imsplitbit has joined #openstack-dev | 02:50 | |
| *** johnpur has quit IRC | 02:51 | |
| *** espen__ has quit IRC | 02:52 | |
| *** espen__ has joined #openstack-dev | 02:52 | |
| *** vipul is now known as vipul|away | 02:52 | |
| *** hemna has joined #openstack-dev | 02:55 | |
| ayoung | topol, hold on, fighting other fires | 02:55 |
| topol | ayoung, no worries. I can wait. I will put something on tomorrow's keystone mtg agenda. I will go back and do some more reviews until you get some of the fires out | 02:58 |
| *** shang_ has quit IRC | 02:59 | |
| *** soody has joined #openstack-dev | 02:59 | |
| *** psedlak has joined #openstack-dev | 03:00 | |
| *** novas0x2a|laptop has quit IRC | 03:01 | |
| *** bdpayne has quit IRC | 03:02 | |
| *** ladquin has quit IRC | 03:06 | |
| *** thickskin has left #openstack-dev | 03:06 | |
| *** jcmartin has joined #openstack-dev | 03:07 | |
| *** vipul|away is now known as vipul | 03:08 | |
| *** nati_ueno has joined #openstack-dev | 03:08 | |
| *** Tross has quit IRC | 03:11 | |
| *** pabelanger has joined #openstack-dev | 03:12 | |
| *** darjeeling has quit IRC | 03:15 | |
| *** olaph_ has quit IRC | 03:18 | |
| *** olaph has joined #openstack-dev | 03:19 | |
| *** markmcclain has joined #openstack-dev | 03:22 | |
| *** nati_ueno_2 has joined #openstack-dev | 03:24 | |
| *** martine has joined #openstack-dev | 03:25 | |
| *** nati_ueno has quit IRC | 03:25 | |
| *** mohits has quit IRC | 03:32 | |
| openstackgerrit | A change was merged to openstack/python-keystoneclient: Added missing unit tests for shell.py https://review.openstack.org/21227 | 03:37 |
| *** amotoki has joined #openstack-dev | 03:37 | |
| openstackgerrit | A change was merged to openstack/keystone: Add an update option to run_tests.sh https://review.openstack.org/22057 | 03:38 |
| openstackgerrit | A change was merged to openstack/nova: Fix hacking tests on osx https://review.openstack.org/22098 | 03:38 |
| *** READ10 has quit IRC | 03:39 | |
| *** cloudchimp has quit IRC | 03:43 | |
| *** Tross has joined #openstack-dev | 03:44 | |
| *** asalkeld has quit IRC | 03:47 | |
| *** yamahata_ has joined #openstack-dev | 03:48 | |
| *** imsplitbit has quit IRC | 03:49 | |
| *** thingee_zzz is now known as thingee | 03:50 | |
| *** soody has quit IRC | 03:53 | |
| *** lbragstad has joined #openstack-dev | 03:53 | |
| *** fifieldt has quit IRC | 03:58 | |
| *** Tross has quit IRC | 03:58 | |
| *** hemna_ has joined #openstack-dev | 04:00 | |
| *** espen__ has quit IRC | 04:04 | |
| *** hemna has quit IRC | 04:04 | |
| *** espen__ has joined #openstack-dev | 04:04 | |
| *** topol has quit IRC | 04:05 | |
| *** topol has joined #openstack-dev | 04:06 | |
| *** networkstatic_ has joined #openstack-dev | 04:06 | |
| *** portante` has joined #openstack-dev | 04:07 | |
| *** dabo has quit IRC | 04:07 | |
| *** sgran has quit IRC | 04:07 | |
| *** networkstatic has quit IRC | 04:07 | |
| *** networkstatic_ is now known as networkstatic | 04:07 | |
| *** dabo has joined #openstack-dev | 04:07 | |
| *** jbr_ has quit IRC | 04:07 | |
| *** jbr_ has joined #openstack-dev | 04:08 | |
| *** sgran has joined #openstack-dev | 04:08 | |
| *** portante has quit IRC | 04:11 | |
| *** soody has joined #openstack-dev | 04:11 | |
| *** jcmartin has quit IRC | 04:11 | |
| *** nati_ueno has joined #openstack-dev | 04:12 | |
| *** Mathnerd314 has quit IRC | 04:13 | |
| *** Tross has joined #openstack-dev | 04:13 | |
| *** nati_ueno has quit IRC | 04:14 | |
| *** bdpayne has joined #openstack-dev | 04:14 | |
| *** nati_ueno has joined #openstack-dev | 04:15 | |
| *** nati_ueno_2 has quit IRC | 04:15 | |
| *** networkstatic has quit IRC | 04:16 | |
| *** pfreund has joined #openstack-dev | 04:21 | |
| *** adjohn has quit IRC | 04:21 | |
| *** kmartin has quit IRC | 04:23 | |
| *** Mandell has joined #openstack-dev | 04:25 | |
| *** pabelanger has quit IRC | 04:26 | |
| *** danwent has quit IRC | 04:28 | |
| *** sandywalsh has quit IRC | 04:31 | |
| *** tomoe_ has joined #openstack-dev | 04:34 | |
| *** soody has quit IRC | 04:36 | |
| *** soody has joined #openstack-dev | 04:37 | |
| *** bdpayne has quit IRC | 04:46 | |
| *** maurosr has quit IRC | 04:46 | |
| *** utlemming has joined #openstack-dev | 04:48 | |
| *** martine has quit IRC | 04:50 | |
| *** soody has quit IRC | 04:53 | |
| *** saschpe has quit IRC | 04:54 | |
| *** saschpe has joined #openstack-dev | 04:54 | |
| *** soody has joined #openstack-dev | 04:55 | |
| *** bencherian has quit IRC | 04:58 | |
| *** nati_ueno_2 has joined #openstack-dev | 05:04 | |
| *** dims has quit IRC | 05:05 | |
| openstackgerrit | A change was merged to openstack/nova: Fixes a race condition on updating security group rules https://review.openstack.org/21529 | 05:05 |
| *** nati_ueno has quit IRC | 05:06 | |
| *** jab416171 has quit IRC | 05:09 | |
| *** nunosantos has quit IRC | 05:10 | |
| *** rha has quit IRC | 05:11 | |
| *** Tross has quit IRC | 05:12 | |
| *** mlavalle has left #openstack-dev | 05:12 | |
| *** rcloran has quit IRC | 05:12 | |
| *** maoy has quit IRC | 05:13 | |
| *** maoy has joined #openstack-dev | 05:14 | |
| *** navid_ has joined #openstack-dev | 05:14 | |
| *** maoy has quit IRC | 05:15 | |
| *** rha has joined #openstack-dev | 05:16 | |
| *** Tross has joined #openstack-dev | 05:17 | |
| *** ivoks has quit IRC | 05:18 | |
| *** tomoe_ has quit IRC | 05:18 | |
| *** tomoe_ has joined #openstack-dev | 05:19 | |
| *** aeperezt has quit IRC | 05:20 | |
| *** ivoks has joined #openstack-dev | 05:22 | |
| *** tomoe_ has quit IRC | 05:23 | |
| *** rcloran has joined #openstack-dev | 05:24 | |
| *** kmartin has joined #openstack-dev | 05:25 | |
| *** jab416171 has joined #openstack-dev | 05:30 | |
| *** bdpayne has joined #openstack-dev | 05:30 | |
| *** adjohn has joined #openstack-dev | 05:32 | |
| *** avishay has joined #openstack-dev | 05:32 | |
| *** adjohn has quit IRC | 05:36 | |
| *** soody has quit IRC | 05:38 | |
| *** rushiagr has joined #openstack-dev | 05:43 | |
| *** markmcclain has quit IRC | 05:44 | |
| *** shang has joined #openstack-dev | 05:46 | |
| *** sacharya1 has quit IRC | 05:46 | |
| *** garyk has quit IRC | 05:47 | |
| *** adjohn has joined #openstack-dev | 05:51 | |
| *** nati_ueno has joined #openstack-dev | 05:52 | |
| *** almaisan-away is now known as al-maisan | 05:54 | |
| *** nati_ueno_2 has quit IRC | 05:55 | |
| *** darjeeling has joined #openstack-dev | 05:58 | |
| *** morganfainberg has quit IRC | 06:04 | |
| *** bdpayne has quit IRC | 06:08 | |
| *** tkammer has joined #openstack-dev | 06:09 | |
| *** hattwick has quit IRC | 06:17 | |
| *** tomoe_ has joined #openstack-dev | 06:19 | |
| *** zodiak has joined #openstack-dev | 06:23 | |
| *** bdpayne has joined #openstack-dev | 06:26 | |
| zyluo | amotoki, ping | 06:27 |
| amotoki | zyluo: pong | 06:27 |
| amotoki | zyluo: WADL one? | 06:28 |
| zyluo | amotoki, yup | 06:28 |
| amotoki | zyluo: sure. | 06:29 |
| zyluo | amotoki, thanks | 06:29 |
| *** al-maisan is now known as almaisan-away | 06:31 | |
| *** almaisan-away is now known as al-maisan | 06:34 | |
| *** navid_ has quit IRC | 06:36 | |
| *** brucer_afk has quit IRC | 06:38 | |
| *** navid_ has joined #openstack-dev | 06:38 | |
| *** topol has quit IRC | 06:41 | |
| *** kbrierly has quit IRC | 06:42 | |
| *** garyk has joined #openstack-dev | 06:51 | |
| *** kmartin has quit IRC | 06:51 | |
| *** hemna_ has quit IRC | 06:52 | |
| *** jdurgin has left #openstack-dev | 06:56 | |
| *** jdurgin has joined #openstack-dev | 06:57 | |
| *** armaan has joined #openstack-dev | 06:58 | |
| *** morganfainberg has joined #openstack-dev | 06:59 | |
| *** nmagnezi_ has joined #openstack-dev | 06:59 | |
| *** morganfainberg has quit IRC | 07:02 | |
| *** nmagnezi_ has quit IRC | 07:05 | |
| *** hemna_ has joined #openstack-dev | 07:05 | |
| *** tkammer has quit IRC | 07:05 | |
| openstackgerrit | A change was merged to openstack/nova: Small spelling fix in sqlalchemy utils. https://review.openstack.org/22247 | 07:08 |
| *** b1rkh0ff has quit IRC | 07:08 | |
| openstackgerrit | A change was merged to openstack/nova: Do not use abbreviated config group names (zookeeper) https://review.openstack.org/22229 | 07:08 |
| openstackgerrit | A change was merged to openstack/glance: Use install_venv_common.py from oslo. https://review.openstack.org/20914 | 07:09 |
| *** dachary has joined #openstack-dev | 07:10 | |
| openstackgerrit | A change was merged to openstack/glance: Add an update option to run_tests.sh https://review.openstack.org/22070 | 07:11 |
| *** doude has joined #openstack-dev | 07:12 | |
| *** obondarev has joined #openstack-dev | 07:13 | |
| openstackgerrit | A change was merged to openstack/nova: Add select_hosts to scheduler manager rpc https://review.openstack.org/21977 | 07:13 |
| openstackgerrit | A change was merged to openstack/nova: Live migration with an auto selection of dest https://review.openstack.org/22193 | 07:14 |
| *** nati_ueno has quit IRC | 07:14 | |
| openstackgerrit | A change was merged to openstack/nova: Add support for instance vif traffic control. https://review.openstack.org/22104 | 07:14 |
| *** pfreund has quit IRC | 07:14 | |
| *** nati_ueno has joined #openstack-dev | 07:14 | |
| *** mrunge has joined #openstack-dev | 07:16 | |
| *** jpich has joined #openstack-dev | 07:16 | |
| *** dachary has quit IRC | 07:17 | |
| *** bdpayne has quit IRC | 07:18 | |
| *** nati_ueno has quit IRC | 07:19 | |
| *** doude has quit IRC | 07:19 | |
| *** b1rkh0ff has joined #openstack-dev | 07:21 | |
| *** doude has joined #openstack-dev | 07:21 | |
| *** shang has quit IRC | 07:24 | |
| *** yolanda has joined #openstack-dev | 07:29 | |
| *** shang has joined #openstack-dev | 07:31 | |
| *** romcheg has quit IRC | 07:32 | |
| *** amerine has quit IRC | 07:36 | |
| *** xgauvrit has joined #openstack-dev | 07:40 | |
| *** xga_ has joined #openstack-dev | 07:40 | |
| *** nati_ueno has joined #openstack-dev | 07:41 | |
| *** mindpixel has joined #openstack-dev | 07:42 | |
| *** asalkeld has joined #openstack-dev | 07:47 | |
| *** salv-orlando has joined #openstack-dev | 07:47 | |
| *** koolhead17 has joined #openstack-dev | 07:51 | |
| *** eglynn has joined #openstack-dev | 07:53 | |
| *** reidrac has joined #openstack-dev | 07:55 | |
| *** rafaduran has joined #openstack-dev | 08:02 | |
| *** nati_ueno has quit IRC | 08:06 | |
| *** nati_ueno has joined #openstack-dev | 08:06 | |
| *** flaper87 has joined #openstack-dev | 08:07 | |
| *** mohits has joined #openstack-dev | 08:09 | |
| *** nati_ueno has quit IRC | 08:11 | |
| *** dachary has joined #openstack-dev | 08:11 | |
| *** thouveng has joined #openstack-dev | 08:13 | |
| *** jprovazn has joined #openstack-dev | 08:15 | |
| *** dachary has quit IRC | 08:15 | |
| openstackgerrit | A change was merged to openstack/nova: create new cidr type for data storage https://review.openstack.org/22238 | 08:17 |
| *** ondergetekende has joined #openstack-dev | 08:22 | |
| *** bing_bu has quit IRC | 08:22 | |
| ondergetekende | https://review.openstack.org/21958 | 08:22 |
| ondergetekende | oops, wrong channel | 08:22 |
| *** egallen has joined #openstack-dev | 08:24 | |
| *** giulivo has joined #openstack-dev | 08:25 | |
| *** gael__ has joined #openstack-dev | 08:25 | |
| *** jgallard has joined #openstack-dev | 08:27 | |
| *** zaneb has quit IRC | 08:29 | |
| garyk | amotoki: ping | 08:31 |
| *** dachary has joined #openstack-dev | 08:33 | |
| *** thingee is now known as thingee_zzz | 08:36 | |
| *** zoresvit has joined #openstack-dev | 08:37 | |
| *** kbrierly has joined #openstack-dev | 08:41 | |
| *** tomoe_ has quit IRC | 08:42 | |
| *** tomoe_ has joined #openstack-dev | 08:42 | |
| *** egallen has quit IRC | 08:42 | |
| openstackgerrit | A change was merged to openstack/nova: Fix a typo in two comments. networksa -> networks https://review.openstack.org/22270 | 08:43 |
| openstackgerrit | A change was merged to openstack/nova: Remove unused db calls https://review.openstack.org/22255 | 08:44 |
| *** aloga has quit IRC | 08:44 | |
| *** aloga has joined #openstack-dev | 08:44 | |
| ttx | ayoung: yeah, end of the day hawaii time | 08:46 |
| *** zoresvit has quit IRC | 08:46 | |
| *** tomoe_ has quit IRC | 08:47 | |
| *** shardy_afk is now known as shardy | 08:48 | |
| amotoki | garyk: hi | 08:51 |
| *** mmagr has joined #openstack-dev | 08:55 | |
| garyk | amotoki: hi. i think i made a mistake - when i reviewed https://review.openstack.org/#/c/20574/i approved by mistake. can you please take a look if possible | 08:55 |
| *** winston-d_ has quit IRC | 08:56 | |
| amotoki | sure. This patch looks almost ok last time I looked | 08:56 |
| amotoki | garyk: I checked and approved. Just the commit message was corrected :-) | 08:57 |
| garyk | amotoki: thanks! | 08:57 |
| *** zing has joined #openstack-dev | 08:58 | |
| *** dosaboy has joined #openstack-dev | 09:00 | |
| *** zoresvit has joined #openstack-dev | 09:00 | |
| *** dave_mcn has joined #openstack-dev | 09:00 | |
| *** egallen has joined #openstack-dev | 09:01 | |
| *** zing_ has joined #openstack-dev | 09:02 | |
| *** zing has quit IRC | 09:02 | |
| *** zing_ is now known as zing | 09:02 | |
| *** zoresvit has quit IRC | 09:05 | |
| *** pixelbeat has joined #openstack-dev | 09:08 | |
| *** Ryan_Lane has quit IRC | 09:09 | |
| *** egallen_ has joined #openstack-dev | 09:09 | |
| *** egallen has quit IRC | 09:09 | |
| *** egallen_ is now known as egallen | 09:09 | |
| *** zoresvit has joined #openstack-dev | 09:10 | |
| *** alobbs has joined #openstack-dev | 09:12 | |
| *** Mandell has quit IRC | 09:13 | |
| *** egallen has quit IRC | 09:13 | |
| *** darjeeling has quit IRC | 09:16 | |
| *** zoresvit has quit IRC | 09:16 | |
| *** tkammer has joined #openstack-dev | 09:16 | |
| *** nmagnezi_ has joined #openstack-dev | 09:17 | |
| *** egallen has joined #openstack-dev | 09:20 | |
| *** derekh has joined #openstack-dev | 09:22 | |
| *** egallen has quit IRC | 09:23 | |
| *** corXi has joined #openstack-dev | 09:23 | |
| *** iartarisi has joined #openstack-dev | 09:23 | |
| *** xga_ has quit IRC | 09:24 | |
| *** xgauvrit has quit IRC | 09:24 | |
| *** giulivo has quit IRC | 09:25 | |
| *** zoresvit has joined #openstack-dev | 09:27 | |
| *** mkollaro has joined #openstack-dev | 09:28 | |
| *** NobodyCam has quit IRC | 09:31 | |
| *** FlorianOtel has joined #openstack-dev | 09:38 | |
| *** gongysh has quit IRC | 09:42 | |
| *** darjeeling has joined #openstack-dev | 09:43 | |
| *** jruzicka has joined #openstack-dev | 09:43 | |
| *** afazekas has joined #openstack-dev | 09:43 | |
| *** egallen has joined #openstack-dev | 09:44 | |
| *** gongysh has joined #openstack-dev | 09:45 | |
| *** salv-orlando_ has joined #openstack-dev | 09:45 | |
| *** salv-orlando has quit IRC | 09:45 | |
| *** salv-orlando_ is now known as salv-orlando | 09:45 | |
| *** xchu has joined #openstack-dev | 09:45 | |
| *** egallen has quit IRC | 09:45 | |
| *** tkammer has quit IRC | 09:46 | |
| *** tkammer has joined #openstack-dev | 09:46 | |
| *** alexxu has quit IRC | 09:48 | |
| *** nmagnezi_ has quit IRC | 09:54 | |
| *** boris-42 has quit IRC | 09:59 | |
| *** xchu has quit IRC | 09:59 | |
| *** salv-orlando has quit IRC | 10:00 | |
| *** nmagnezi has joined #openstack-dev | 10:01 | |
| *** adjohn has quit IRC | 10:08 | |
| *** BobBall has quit IRC | 10:09 | |
| *** Yada has joined #openstack-dev | 10:10 | |
| *** tomoe_ has joined #openstack-dev | 10:12 | |
| openstackgerrit | A change was merged to openstack/nova: Remove unused db calls from nova.db.sqlalchemy.api https://review.openstack.org/22262 | 10:12 |
| *** mkollaro has quit IRC | 10:12 | |
| openstackgerrit | A change was merged to openstack/nova: Fix _get_instance_volume_block_device_info call parameter https://review.openstack.org/22235 | 10:13 |
| openstackgerrit | A change was merged to openstack/nova: Remove unused nova.db.api:network_get_by_instance https://review.openstack.org/22268 | 10:13 |
| openstackgerrit | A change was merged to openstack/nova: Delete baremetal interfaces when their parent node is deleted https://review.openstack.org/22013 | 10:14 |
| openstackgerrit | A change was merged to openstack/quantum: Imported Translations from Transifex https://review.openstack.org/22261 | 10:14 |
| *** tkammer has quit IRC | 10:17 | |
| *** markmc has joined #openstack-dev | 10:18 | |
| *** danpb has joined #openstack-dev | 10:18 | |
| *** zoresvit has quit IRC | 10:20 | |
| *** salv-orlando has joined #openstack-dev | 10:25 | |
| *** megha has joined #openstack-dev | 10:28 | |
| *** dave_mcn has quit IRC | 10:32 | |
| *** hattwick has joined #openstack-dev | 10:36 | |
| *** zoresvit has joined #openstack-dev | 10:36 | |
| *** milez has quit IRC | 10:38 | |
| *** vkmc has joined #openstack-dev | 10:41 | |
| *** darjeeling has quit IRC | 10:41 | |
| *** adjohn has joined #openstack-dev | 10:42 | |
| *** psedlak has quit IRC | 10:46 | |
| *** darjeeling has joined #openstack-dev | 10:46 | |
| *** egallen has joined #openstack-dev | 10:51 | |
| *** trapni has joined #openstack-dev | 10:52 | |
| *** trapni has joined #openstack-dev | 10:52 | |
| *** adjohn has quit IRC | 10:52 | |
| *** lucasagomes has joined #openstack-dev | 11:02 | |
| *** milez has joined #openstack-dev | 11:03 | |
| *** egallen has quit IRC | 11:09 | |
| al-maisan | is there a glance specific irc channel? | 11:09 |
| al-maisan | I am trying to figure out why this is failing: | 11:10 |
| al-maisan | $ glance image-list | 11:10 |
| al-maisan | Request returned failure status. | 11:10 |
| al-maisan | None | 11:10 |
| al-maisan | HTTPInternalServerError (HTTP 500) | 11:10 |
| vkmc | #openstack-glance | 11:10 |
| al-maisan | ok .. thanks vkmc! | 11:10 |
| vkmc | :) Make sure to paste the error trace in http://paste.openstack.org | 11:10 |
| vkmc | al-maisan, And... if you are receiving a server error please check Apache's logs and screen logs to have a much better idea of what's going on | 11:11 |
| openstackgerrit | A change was merged to openstack/quantum: Metadata support for NVP plugin https://review.openstack.org/21934 | 11:12 |
| *** xgauvrit has joined #openstack-dev | 11:13 | |
| *** xga has joined #openstack-dev | 11:13 | |
| al-maisan | vkmc: checked all except for apache .. will look at the latter now | 11:13 |
| vkmc | al-maisan, Ok, good luck with it :) | 11:14 |
| jpich | al-maisan: I think vkmc meant the glance logs, Apache logs are only relevant to Horizon really :) | 11:14 |
| danpb | russellb: pixelbeat if either of you have time to review these it'd be appreciated given grizzly freeze rsn https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/compute-driver-events,n,z | 11:14 |
| jpich | al-maisan: Adding the --debug switch can help when using the CLI tools | 11:14 |
| al-maisan | jpich: checked the glance logs, here's what I see: http://pastebin.ubuntu.com/1681029/ | 11:14 |
| al-maisan | will try that as well | 11:15 |
| pixelbeat | danpb, OK swapsies :) https://review.openstack.org/#/q/topic:bp/preallocated-images,n,z | 11:15 |
| vkmc | jpich, Thanks for fixing it! | 11:15 |
| vkmc | al-maisan, Sorry for the mistake | 11:15 |
| al-maisan | jpich: http://paste.ubuntu.com/1681410/ | 11:16 |
| al-maisan | Is there any way to turn off glance caching> | 11:16 |
| al-maisan | ? | 11:16 |
| jpich | al-maisan: The Glance folks on the channel vkmc directed you to will probably better be able to help you than I am, sorry, I'm not very familiar with glance | 11:17 |
| danpb | pixelbeat: sure | 11:18 |
| *** adjohn has joined #openstack-dev | 11:18 | |
| *** adjohn has quit IRC | 11:23 | |
| *** anniec has joined #openstack-dev | 11:28 | |
| *** anniec has quit IRC | 11:28 | |
| *** anniec has joined #openstack-dev | 11:28 | |
| openstackgerrit | A change was merged to openstack/python-cinderclient: adding v2 support to cinderclient https://review.openstack.org/21189 | 11:29 |
| zyluo | salv-orlando, ping | 11:29 |
| salv-orlando | zyluo: I am looking at the wadl code now | 11:29 |
| salv-orlando | ah you pushed another patch... | 11:30 |
| *** adalbas has joined #openstack-dev | 11:30 | |
| *** tkammer has joined #openstack-dev | 11:30 | |
| zyluo | salv-orlando, I just made some changes. Basically changed all singular attrs to xsd:int or xsd:string | 11:30 |
| salv-orlando | ok, thanks for telling me about the differences | 11:30 |
| *** pcm_ has joined #openstack-dev | 11:30 | |
| *** pcm_ has quit IRC | 11:31 | |
| *** pcm_ has joined #openstack-dev | 11:32 | |
| zyluo | salv-orlando, I'll be around for another two hours. | 11:34 |
| salv-orlando | k | 11:34 |
| *** dhellmann has joined #openstack-dev | 11:34 | |
| *** mohits has quit IRC | 11:40 | |
| *** jbr_1 has joined #openstack-dev | 11:44 | |
| *** armaan has quit IRC | 11:47 | |
| openstackgerrit | A change was merged to openstack/cinder: Install rtslib when installing cinder https://review.openstack.org/22194 | 11:47 |
| *** adjohn has joined #openstack-dev | 11:49 | |
| *** adjohn has quit IRC | 11:54 | |
| *** psedlak has joined #openstack-dev | 11:57 | |
| *** maurosr has joined #openstack-dev | 12:00 | |
| *** maurosr has quit IRC | 12:00 | |
| *** maurosr has joined #openstack-dev | 12:01 | |
| *** baseggio has left #openstack-dev | 12:02 | |
| *** rkukura has quit IRC | 12:03 | |
| *** timello_ has quit IRC | 12:04 | |
| *** radez is now known as radez_g0n3 | 12:05 | |
| *** maurosr has quit IRC | 12:07 | |
| *** mkollaro has joined #openstack-dev | 12:13 | |
| *** timello_ has joined #openstack-dev | 12:14 | |
| *** giulivo has joined #openstack-dev | 12:15 | |
| *** mkollaro has quit IRC | 12:18 | |
| *** maurosr has joined #openstack-dev | 12:18 | |
| *** adjohn has joined #openstack-dev | 12:19 | |
| *** yamahata has quit IRC | 12:20 | |
| *** yamahata has joined #openstack-dev | 12:23 | |
| *** adjohn has quit IRC | 12:24 | |
| *** johnthetubaguy has joined #openstack-dev | 12:25 | |
| *** sandywalsh has joined #openstack-dev | 12:30 | |
| *** martine has joined #openstack-dev | 12:31 | |
| *** al-maisan is now known as almaisan-away | 12:34 | |
| *** megha is now known as unix | 12:34 | |
| *** cabral has joined #openstack-dev | 12:36 | |
| *** digitalsanctum has joined #openstack-dev | 12:38 | |
| *** darjeeling has quit IRC | 12:39 | |
| *** digitalsanctum has quit IRC | 12:40 | |
| *** martine has quit IRC | 12:41 | |
| *** martine has joined #openstack-dev | 12:42 | |
| *** lucasagomes has quit IRC | 12:43 | |
| *** digitalsanctum has joined #openstack-dev | 12:44 | |
| *** sthaha has quit IRC | 12:47 | |
| *** timello_ has quit IRC | 12:48 | |
| *** timello_ has joined #openstack-dev | 12:50 | |
| *** adjohn has joined #openstack-dev | 12:50 | |
| *** krtaylor has quit IRC | 12:52 | |
| *** timello_ has quit IRC | 12:54 | |
| *** adjohn has quit IRC | 12:55 | |
| *** armaan has joined #openstack-dev | 12:55 | |
| *** egallen has joined #openstack-dev | 12:56 | |
| *** sthaha has joined #openstack-dev | 12:58 | |
| garyk | markmc: ping | 12:59 |
| markmc | garyk, yep? | 12:59 |
| garyk | markmc: hi, question. can you please look at http://paste.openstack.org/show/31929/ | 12:59 |
| garyk | markmc: any idea what i am missing? | 12:59 |
| markmc | garyk, pip-python install http://tarballs.openstack.org/oslo-config/oslo-config-2013.1b3.tar.gz | 13:00 |
| garyk | markmc: great, thanks. i'll try | 13:01 |
| *** timello_ has joined #openstack-dev | 13:07 | |
| *** lbragstad has left #openstack-dev | 13:08 | |
| *** markvoelker has joined #openstack-dev | 13:08 | |
| *** tkammer has quit IRC | 13:08 | |
| garyk | markmc: for some reason i still have the same problem after running the above command. | 13:09 |
| *** mrunge has quit IRC | 13:10 | |
| markmc | garyk, do 'pip-python uninstall oslo-config' and try again | 13:10 |
| markmc | garyk, also, there's an rpm here: http://koji.fedoraproject.org/koji/buildinfo?buildID=395798 | 13:10 |
| garyk | markmc: ok, will do | 13:10 |
| garyk | markmc: ok, thanks. i'll try | 13:10 |
| *** egallen has quit IRC | 13:11 | |
| *** dave_mcn has joined #openstack-dev | 13:12 | |
| *** unix has quit IRC | 13:13 | |
| amotoki | gongysh: ping | 13:13 |
| openstackgerrit | A change was merged to openstack/keystone: make fakeldap._match_query work for an arbitrary number of groups https://review.openstack.org/22196 | 13:15 |
| openstackgerrit | A change was merged to openstack/quantum: Add unit test for ryu-agent https://review.openstack.org/21419 | 13:15 |
| openstackgerrit | A change was merged to openstack/python-quantumclient: Match other python-*client prettytable dependency https://review.openstack.org/22151 | 13:15 |
| *** kbringard has joined #openstack-dev | 13:16 | |
| *** rods has quit IRC | 13:16 | |
| *** rods has joined #openstack-dev | 13:16 | |
| *** unix has joined #openstack-dev | 13:18 | |
| *** zyluo has quit IRC | 13:20 | |
| *** dolphm has joined #openstack-dev | 13:20 | |
| *** adjohn has joined #openstack-dev | 13:21 | |
| *** rods has quit IRC | 13:22 | |
| *** ladquin has joined #openstack-dev | 13:22 | |
| *** yaguang has left #openstack-dev | 13:22 | |
| *** rods has joined #openstack-dev | 13:22 | |
| *** vishy has quit IRC | 13:22 | |
| *** saschpe has quit IRC | 13:23 | |
| *** saschpe has joined #openstack-dev | 13:23 | |
| *** jerdfelt has quit IRC | 13:24 | |
| *** jerdfelt has joined #openstack-dev | 13:25 | |
| *** adjohn has quit IRC | 13:25 | |
| *** timello_ has quit IRC | 13:27 | |
| *** zyluo has joined #openstack-dev | 13:28 | |
| *** krtaylor has joined #openstack-dev | 13:28 | |
| zyluo | salv-orlando gongysh amotoki: I'm on the road. Response could be slow | 13:29 |
| salv-orlando | ok | 13:29 |
| amotoki | garyk: ping | 13:29 |
| garyk | amotoki: hi | 13:29 |
| *** vishy has joined #openstack-dev | 13:30 | |
| amotoki | garyk: when i tested with yong's patch 2, i see several times OVS agent is down in agent-list. | 13:30 |
| garyk | amotoki: by doing the agent-list? | 13:31 |
| amotoki | garyk: yes | 13:31 |
| amotoki | watch -n 1 quantum agent-list | 13:31 |
| amotoki | do you see the similar situation? | 13:31 |
| *** rods has quit IRC | 13:31 | |
| garyk | no. i always see the alive as :) | 13:32 |
| garyk | :-) | 13:32 |
| *** rods has joined #openstack-dev | 13:32 | |
| amotoki | garyk: my env is KVM on KVM. It may be the reason of it. | 13:32 |
| garyk | amotoki: i am running on a baremetal host (when i was running with nested virtualization i ran into problems a few months ago) | 13:33 |
| garyk | i can try on a nested setup. | 13:33 |
| amotoki | garyk: thanks. anyway yong's patch looks great. I tested dhcp-agent with/without notification and it worked well. | 13:34 |
| salv-orlando | garyk, amotoki: I have a nested env (KVM on KVM), not seeing ovs-agent going down | 13:34 |
| garyk | salv-orlando amotoki : thanks for the input. the patch looks good to me and i like the new generic notification treatment | 13:35 |
| amotoki | salv-orlando: : What I saw is ovs-agent is down just for a second or two and back to active. | 13:35 |
| salv-orlando | amotoki: will do further checks. | 13:35 |
| amotoki | we needs to check such things after G-3 release.. | 13:36 |
| salv-orlando | amotoki: indeed. It won't be a blocker for me | 13:37 |
| *** timello_ has joined #openstack-dev | 13:37 | |
| *** alexxu has joined #openstack-dev | 13:37 | |
| amotoki | salv-orlando: agree. | 13:38 |
| *** spzala has joined #openstack-dev | 13:38 | |
| salv-orlando | amotoki: if you are happy with it, we can start approving patch #2 - https://review.openstack.org/#/c/21069/ | 13:39 |
| *** aeperezt has joined #openstack-dev | 13:39 | |
| *** rods is now known as Guest86187 | 13:40 | |
| amotoki | salv-orlando: thanks, I already give +2. from now approve it. | 13:40 |
| salv-orlando | amotoki: thanks | 13:41 |
| *** darraghb has joined #openstack-dev | 13:45 | |
| *** rods has joined #openstack-dev | 13:46 | |
| *** cabral has quit IRC | 13:47 | |
| *** adjohn has joined #openstack-dev | 13:52 | |
| *** zeriouz has joined #openstack-dev | 13:53 | |
| *** adjohn has quit IRC | 13:55 | |
| *** dhellmann has quit IRC | 13:55 | |
| *** eharney has joined #openstack-dev | 13:56 | |
| *** dprince has joined #openstack-dev | 13:56 | |
| *** gongysh has quit IRC | 13:58 | |
| *** Clabbe has joined #openstack-dev | 14:02 | |
| *** jergerber has joined #openstack-dev | 14:03 | |
| *** rkukura has joined #openstack-dev | 14:04 | |
| *** topol has joined #openstack-dev | 14:04 | |
| *** anteaya has joined #openstack-dev | 14:06 | |
| *** fesp has joined #openstack-dev | 14:07 | |
| *** romcheg has joined #openstack-dev | 14:07 | |
| *** flaper87 has quit IRC | 14:08 | |
| *** troytoman-away is now known as troytoman | 14:09 | |
| *** romcheg has quit IRC | 14:09 | |
| *** romcheg has joined #openstack-dev | 14:10 | |
| *** romcheg has left #openstack-dev | 14:12 | |
| *** henrynash has joined #openstack-dev | 14:12 | |
| anteaya | koolhead17, ping | 14:15 |
| *** mtreinish has joined #openstack-dev | 14:15 | |
| *** armaan has quit IRC | 14:17 | |
| koolhead17 | anteaya: hi | 14:18 |
| anteaya | hello, I am taking a look at this bug you filed: https://bugs.launchpad.net/openstack-ci/+bug/1093812 | 14:18 |
| uvirtbot | Launchpad bug 1093812 in openstack-ci "nova API version doc should point to Nova developer doc" [High,Confirmed] | 14:18 |
| koolhead17 | anteaya: yes | 14:18 |
| anteaya | I am unable to discern what links should lead to the Nova developer doc | 14:18 |
| anteaya | I have gone through the source code of the API version doc and can find the link you reference | 14:19 |
| anteaya | is is a redirect do you know? | 14:19 |
| anteaya | sorry, I can't find the link you reference | 14:20 |
| anteaya | could you give me an example of /versionnumber sites | 14:21 |
| maroh | arosen : hi | 14:21 |
| koolhead17 | anteaya: seems like things got fixed in over month and half time :) | 14:21 |
| *** zoresvit has quit IRC | 14:21 | |
| *** amotoki has quit IRC | 14:22 | |
| *** zoresvit has joined #openstack-dev | 14:23 | |
| anteaya | koolhead17, so this bug can be closed? or marked invalid? | 14:23 |
| koolhead17 | yes please close it | 14:23 |
| *** tomoe_ has quit IRC | 14:23 | |
| anteaya | koolhead17, thank you | 14:23 |
| koolhead17 | np | 14:24 |
| *** tomoe_ has joined #openstack-dev | 14:24 | |
| *** esheffield has joined #openstack-dev | 14:24 | |
| *** woodspa has joined #openstack-dev | 14:25 | |
| *** dims has joined #openstack-dev | 14:25 | |
| *** tomoe_ has quit IRC | 14:27 | |
| *** tomoe_ has joined #openstack-dev | 14:27 | |
| *** Clabbe has quit IRC | 14:28 | |
| *** READ10 has joined #openstack-dev | 14:30 | |
| dolphm | ayoung: +2 on this from me https://review.openstack.org/#/c/22109/ | 14:30 |
| *** kmartin has joined #openstack-dev | 14:30 | |
| henrynash | ayoung: can you post the link of your trusts implementation, can only seem to find the API change…. | 14:31 |
| *** bnemec has joined #openstack-dev | 14:32 | |
| *** mindpixel has quit IRC | 14:32 | |
| *** bknudson has joined #openstack-dev | 14:33 | |
| *** henrynash has quit IRC | 14:34 | |
| *** gongysh has joined #openstack-dev | 14:34 | |
| *** zoresvit has quit IRC | 14:35 | |
| *** imsplitbit has joined #openstack-dev | 14:35 | |
| *** darjeeling has joined #openstack-dev | 14:42 | |
| *** zoresvit has joined #openstack-dev | 14:42 | |
| dolphm | henrynash: https://review.openstack.org/#/c/20289/ | 14:43 |
| *** b1rkh0ff has quit IRC | 14:44 | |
| *** b1rkh0ff has joined #openstack-dev | 14:45 | |
| openstackgerrit | A change was merged to openstack/quantum: Modify dhcp agent for agent management extension https://review.openstack.org/21069 | 14:47 |
| ayoung | dolphm, so, on Guangs, my only concern is /auth | 14:52 |
| *** dontalton has joined #openstack-dev | 14:52 | |
| ayoung | I'm really thinking about proposing X509s as an alternative to Tokens in Havana, and I would want them to be under /auth as well | 14:53 |
| *** adjohn has joined #openstack-dev | 14:53 | |
| dolphm | ayoung: i'm not sure you ever answered my questions about Accept on your identity-api review? | 14:53 |
| ayoung | dolphm, I don't think accepts headers provides a fine enough granularity | 14:54 |
| ayoung | for example | 14:54 |
| *** almaisan-away is now known as al-maisan | 14:54 | |
| ayoung | if a browser hits Keystone to get a cert, it is going to have HTTP as an accept header | 14:54 |
| ayoung | but we also want to be able to display information about tokens | 14:54 |
| ayoung | and which also might be relevant in an HTTP view | 14:55 |
| ayoung | so I think tokens and X509s are going to be fundamentally different entities, and not just two different views of the same entities | 14:55 |
| ayoung | dolphm, the thing is, the X509 support is in some ways antique and in some ways a moving target | 14:57 |
| *** adalbas has quit IRC | 14:57 | |
| ayoung | there are old, old apis which will allow a browser to import a certificate directly, and then the mozilla team is also working on a new scheme for short term certs, I'll see if I can find the link | 14:57 |
| *** adjohn has quit IRC | 14:57 | |
| *** radez_g0n3 is now known as radez | 14:58 | |
| *** gargya has joined #openstack-dev | 14:58 | |
| *** jimfehlig has joined #openstack-dev | 14:59 | |
| *** navid_ has left #openstack-dev | 14:59 | |
| ayoung | dolphm, it is also likely that X509 certs will have different information that we will have to track from tokens, which means that they can't be considered as different views of the same data. | 15:00 |
| *** maoy has joined #openstack-dev | 15:00 | |
| *** AlanClark has joined #openstack-dev | 15:00 | |
| ayoung | dolphm, also, when it comes to revocation lists, we will want to keep tokens and certs separate. No need to put tokens in a CRL and vice versa if only for size | 15:00 |
| *** koolhead17 has quit IRC | 15:02 | |
| *** henrynash has joined #openstack-dev | 15:02 | |
| *** markmcclain has joined #openstack-dev | 15:03 | |
| *** dhellmann has joined #openstack-dev | 15:04 | |
| ayoung | henrynash, I added you as a reviewer to the trusts review | 15:04 |
| henrynash | ayoung: thx | 15:05 |
| ayoung | https://review.openstack.org/#/c/20289/ | 15:05 |
| *** cloudchimp has joined #openstack-dev | 15:05 | |
| ayoung | henrynash, BTW, I've started using this view for open reviews: https://review.openstack.org/#/q/status:open+project:openstack/keystone,n,z | 15:05 |
| *** ek6 has joined #openstack-dev | 15:05 | |
| *** edmund has joined #openstack-dev | 15:05 | |
| henrynash | ayoung: ahh, nice... | 15:06 |
| *** bencherian has joined #openstack-dev | 15:06 | |
| *** armaan has joined #openstack-dev | 15:07 | |
| *** ctracey has quit IRC | 15:08 | |
| ayoung | henrynash, and, not to distract you, the view for keystone client. https://review.openstack.org/#/q/status:open+project:openstack/python-keystoneclient,n,z | 15:08 |
| henrynash | ayoung: np….thanks | 15:09 |
| *** dontalton has quit IRC | 15:09 | |
| *** ctracey has joined #openstack-dev | 15:09 | |
| *** ndipanov has quit IRC | 15:10 | |
| *** JonnyNomad has joined #openstack-dev | 15:12 | |
| *** fesp is now known as flaper87 | 15:15 | |
| *** beagles has quit IRC | 15:18 | |
| *** beagles has joined #openstack-dev | 15:19 | |
| *** pabelanger has joined #openstack-dev | 15:19 | |
| henrynash | anyone know anything about openstack client failing in gate vm quantum (version mismatch on pretty table)? | 15:21 |
| *** sacharya has joined #openstack-dev | 15:23 | |
| *** zykes- has quit IRC | 15:23 | |
| sdague | danpb: threw some comments on the libvirt events patch series, how keen are you to land the work in grizzly? | 15:24 |
| *** adjohn has joined #openstack-dev | 15:24 | |
| sdague | lifeless: you still awake? | 15:24 |
| *** nati_ueno has joined #openstack-dev | 15:24 | |
| *** ndipanov has joined #openstack-dev | 15:25 | |
| *** kmartin has quit IRC | 15:26 | |
| *** armaan has quit IRC | 15:26 | |
| *** darjeeling has quit IRC | 15:27 | |
| *** darjeeling has joined #openstack-dev | 15:28 | |
| *** troytoman is now known as troytoman-away | 15:28 | |
| *** adjohn has quit IRC | 15:28 | |
| danpb | sdague: well guess it is really a question of how much people care about the bug it is addressing | 15:28 |
| danpb | sdague: ie if a guest admin does 'shutdown -h now' in their guest, Nova won't change to RUNNING -> SHUTOFF for upto 10 minutes | 15:29 |
| danpb | this behaviour has existed for folsom/essex and presumably earlier without huge complaints | 15:29 |
| danpb | so while it is important to fix, I can't claim it is a release blocker | 15:29 |
| *** zykes- has joined #openstack-dev | 15:32 | |
| *** cp16net is now known as cp16net|away | 15:32 | |
| *** cp16net|away is now known as cp16net | 15:32 | |
| *** TerryH has joined #openstack-dev | 15:33 | |
| *** sacharya has quit IRC | 15:33 | |
| sdague | yeh, my only concern is whether we expose new issues with the native thread changes. it's good stuff to go in, wondering how risky you think that is | 15:33 |
| *** rnirmal has joined #openstack-dev | 15:33 | |
| *** otherwiseguy has joined #openstack-dev | 15:34 | |
| *** john5223 has joined #openstack-dev | 15:34 | |
| *** alexxu has quit IRC | 15:34 | |
| russellb | sdague: i'd like to see it go in, i've heard many complaints around the behavior this addresses | 15:34 |
| russellb | sdague: i think the benefit is enough for the risk | 15:34 |
| ayoung | henrynash, link? | 15:36 |
| ayoung | henrynash, I think that might be non-voting. | 15:36 |
| danpb | sdague: the main risk is whether the compute/manager.py will correctly handle events wrt other operations going on in parallel | 15:36 |
| danpb | sdague: there is already a parallelization race in there with the periodic task which can be trying to update a VM state at the same time as some other API call is trying todo the same | 15:37 |
| henrynash | ayoung: doesn't seem to be: the name space one just failed on it: https://review.openstack.org/#/c/21951/ | 15:37 |
| danpb | sdague: this change makes the race more likely - so if there was an existing flaw, we'd be more likely to hit it | 15:37 |
| ayoung | henrynash, reverify. Probably a bad node. | 15:38 |
| *** garyTh has joined #openstack-dev | 15:38 | |
| ayoung | henrynash, hm..lets see if a change went in elsewhere that changes which version of pretty table? | 15:39 |
| *** portante` is now known as portante | 15:41 | |
| ayoung | henrynash, https://etherpad.openstack.org/trusts_api | 15:42 |
| *** troytoman-away is now known as troytoman | 15:42 | |
| ayoung | trying to word smith this | 15:42 |
| *** AlanClark__ has joined #openstack-dev | 15:42 | |
| ayoung | https://review.openstack.org/#/c/22063/8/openstack-identity-api/src/markdown/identity-api-v3.md | 15:42 |
| *** winston-d_ has joined #openstack-dev | 15:44 | |
| *** jprovazn has quit IRC | 15:44 | |
| *** cp16net is now known as cp16net|away | 15:45 | |
| *** AlanClark has quit IRC | 15:46 | |
| sdague | danpb: ok, I think I can live with that | 15:46 |
| *** lglenden has joined #openstack-dev | 15:47 | |
| *** digitalsanctum has quit IRC | 15:47 | |
| *** sthaha has quit IRC | 15:48 | |
| *** zaneb has joined #openstack-dev | 15:49 | |
| *** esp has joined #openstack-dev | 15:49 | |
| *** esp has quit IRC | 15:50 | |
| *** digitalsanctum has joined #openstack-dev | 15:50 | |
| dave_mcn | Hey all, I keep seeing failures from Jenkins but they don't seem to be related to my change. I've rechecked it a few times now and the failures have happened each time but not always from the same jobs. Is anyone else experiencing problems? | 15:51 |
| *** radez is now known as radez_g0n3 | 15:51 | |
| *** hugokuo has joined #openstack-dev | 15:51 | |
| *** al-maisan is now known as almaisan-away | 15:52 | |
| henrynash | atyoung: will be back with you wordsmithing in a sec | 15:52 |
| henrynash | ayoung: will be back with you wordsmithing in a sec | 15:52 |
| pcm_ | anyone: I updated my dev branch with latest from master, and I'm getting an import error for oslo.config when doing "run_tests.sh -f" for quantum. Any ideas? | 15:52 |
| *** devoid has joined #openstack-dev | 15:52 | |
| *** yaguang has joined #openstack-dev | 15:52 | |
| *** adjohn has joined #openstack-dev | 15:54 | |
| *** nmagnezi has quit IRC | 15:57 | |
| *** zoresvit has quit IRC | 15:57 | |
| *** zoresvit has joined #openstack-dev | 15:57 | |
| *** danwent has joined #openstack-dev | 15:58 | |
| *** mkollaro has joined #openstack-dev | 15:59 | |
| garyk | bcwaldon: ping | 15:59 |
| *** dave_mcn has quit IRC | 15:59 | |
| *** olaph has quit IRC | 15:59 | |
| *** adjohn has quit IRC | 15:59 | |
| *** olaph has joined #openstack-dev | 16:00 | |
| *** datsun180b has joined #openstack-dev | 16:01 | |
| *** dachary has quit IRC | 16:02 | |
| *** reidrac has quit IRC | 16:02 | |
| *** anniec has quit IRC | 16:03 | |
| *** gargya has quit IRC | 16:03 | |
| *** nati_ueno_2 has joined #openstack-dev | 16:04 | |
| *** nati_ueno has quit IRC | 16:07 | |
| *** amotoki has joined #openstack-dev | 16:08 | |
| *** ndipanov has quit IRC | 16:08 | |
| *** egallen has joined #openstack-dev | 16:10 | |
| *** pabelanger has quit IRC | 16:10 | |
| *** dave_mcn has joined #openstack-dev | 16:12 | |
| *** cp16net|away is now known as cp16net | 16:12 | |
| *** bdpayne has joined #openstack-dev | 16:12 | |
| *** zyluo has quit IRC | 16:12 | |
| *** zyluo has joined #openstack-dev | 16:12 | |
| *** kmartin has joined #openstack-dev | 16:13 | |
| *** zyluo has quit IRC | 16:14 | |
| *** zyluo has joined #openstack-dev | 16:15 | |
| *** armaan has joined #openstack-dev | 16:17 | |
| *** sacharya has joined #openstack-dev | 16:18 | |
| *** gyee has joined #openstack-dev | 16:19 | |
| *** zaneb has quit IRC | 16:21 | |
| *** digitals_ has joined #openstack-dev | 16:21 | |
| *** ndipanov has joined #openstack-dev | 16:22 | |
| *** pabelanger has joined #openstack-dev | 16:22 | |
| *** gongysh has quit IRC | 16:23 | |
| *** mjfork has joined #openstack-dev | 16:24 | |
| *** mmagr has quit IRC | 16:24 | |
| *** digitalsanctum has quit IRC | 16:25 | |
| *** koolhead17 has joined #openstack-dev | 16:25 | |
| *** adjohn has joined #openstack-dev | 16:25 | |
| *** rpedde_away is now known as rpedde | 16:25 | |
| *** adalbas has joined #openstack-dev | 16:26 | |
| *** digitals_ is now known as digitalsanctum | 16:28 | |
| *** egallen has quit IRC | 16:28 | |
| *** rkukura has quit IRC | 16:28 | |
| *** njoy has joined #openstack-dev | 16:28 | |
| *** njoy1__ has quit IRC | 16:29 | |
| *** njoy_ has joined #openstack-dev | 16:30 | |
| *** adjohn has quit IRC | 16:30 | |
| *** jcmartin has joined #openstack-dev | 16:32 | |
| *** adjohn has joined #openstack-dev | 16:32 | |
| *** mrodden1 has joined #openstack-dev | 16:33 | |
| *** njoy has quit IRC | 16:33 | |
| *** Gordonz has joined #openstack-dev | 16:34 | |
| *** garyk has quit IRC | 16:35 | |
| *** Gordonz has quit IRC | 16:35 | |
| *** mrodden has quit IRC | 16:35 | |
| *** Gordonz has joined #openstack-dev | 16:36 | |
| *** alunduil has joined #openstack-dev | 16:36 | |
| *** mrodden1 has quit IRC | 16:37 | |
| *** nati_ueno_2 has quit IRC | 16:38 | |
| *** nati_ueno has joined #openstack-dev | 16:38 | |
| *** egallen has joined #openstack-dev | 16:38 | |
| *** yidclare has joined #openstack-dev | 16:40 | |
| *** dave_mcn has quit IRC | 16:41 | |
| *** rkukura has joined #openstack-dev | 16:41 | |
| *** anniec has joined #openstack-dev | 16:41 | |
| *** torandu has quit IRC | 16:41 | |
| *** anniec has quit IRC | 16:42 | |
| *** adjohn has quit IRC | 16:43 | |
| *** salgado is now known as salgado-lunch | 16:44 | |
| *** mrodden has joined #openstack-dev | 16:45 | |
| *** andrewbogott_afk is now known as andrewbogott | 16:45 | |
| *** mrodden has quit IRC | 16:46 | |
| *** mrodden has joined #openstack-dev | 16:46 | |
| *** markmc has quit IRC | 16:46 | |
| *** xgauvrit has quit IRC | 16:47 | |
| *** xga has quit IRC | 16:47 | |
| *** torandu has joined #openstack-dev | 16:47 | |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Add ConfigFilter wrapper class https://review.openstack.org/21243 | 16:48 |
| *** AlanClark__ has quit IRC | 16:49 | |
| *** AlanClark__ has joined #openstack-dev | 16:50 | |
| *** garyk has joined #openstack-dev | 16:50 | |
| *** ondergetekende has quit IRC | 16:50 | |
| *** roampune has joined #openstack-dev | 16:51 | |
| *** diogogmt has joined #openstack-dev | 16:51 | |
| *** xga has joined #openstack-dev | 16:52 | |
| bcwaldon | garyk: hey | 16:53 |
| garyk | bcwaldon: hi. seems like we have a problem with the quantum client after the dependency change today. you aware of it? | 16:54 |
| *** torandu has quit IRC | 16:54 | |
| *** torandu has joined #openstack-dev | 16:55 | |
| garyk | bcwaldon: maybe i am wrong but i see the following: error: Installed | 16:56 |
| garyk | distribution prettytable 0.7 conflicts with requirement | 16:56 |
| garyk | prettytable>=0.6,<0.7 | 16:56 |
| *** xga has quit IRC | 16:56 | |
| *** zoresvit has quit IRC | 16:57 | |
| garyk | bcwaldon: i guess that it could be related to the pypi => https change. | 16:57 |
| bcwaldon | garyk: hrm - I'd bet it was my fault | 16:58 |
| *** reed has joined #openstack-dev | 16:58 | |
| bcwaldon | garyk: so I was seeing similar issues when installing all of the clients at once, coming out at the end with the wrong version of pretty table | 16:58 |
| *** Ryan_Lane has joined #openstack-dev | 16:58 | |
| bcwaldon | garyk: I ended up fixing the problem temporarily a different way, by installing prettytable<0.7 explicitly before any of the clients | 16:58 |
| garyk | bcwaldon: nah, i do not think it was your fault. guess we have a ci hiccup at the moment. | 16:59 |
| bcwaldon | garyk: I am seeing general pypi issues at the moment, if that helps | 16:59 |
| danpb | yep, the pypi website seems to be fubar | 16:59 |
| danpb | it is randomly failing requests | 16:59 |
| bcwaldon | yep, thats what I see trying to install anything in pip | 16:59 |
| *** esp has joined #openstack-dev | 16:59 | |
| bcwaldon | the primary python mirror has an avg response time a magnitude higher than normal | 17:00 |
| *** radez_g0n3 is now known as radez | 17:00 | |
| *** pmyers has quit IRC | 17:00 | |
| *** egallen has quit IRC | 17:00 | |
| *** jcmartin has quit IRC | 17:01 | |
| *** colinmcnamara has joined #openstack-dev | 17:03 | |
| *** ctracey has quit IRC | 17:04 | |
| *** ctracey has joined #openstack-dev | 17:05 | |
| *** jcmartin has joined #openstack-dev | 17:05 | |
| *** zoresvit has joined #openstack-dev | 17:05 | |
| *** jcmartin has quit IRC | 17:06 | |
| *** yaguang has quit IRC | 17:06 | |
| *** egallen has joined #openstack-dev | 17:07 | |
| *** yidclare has quit IRC | 17:08 | |
| *** heckj has joined #openstack-dev | 17:15 | |
| *** jcmartin has joined #openstack-dev | 17:16 | |
| ayoung | dolphm, does the API doc have to be approved before trusts can go in? I don't mean semantics, just formatting and the like? | 17:16 |
| *** jruzicka has quit IRC | 17:16 | |
| ayoung | I want to prioritize. I assume docs are tweakable up until the release | 17:17 |
| openstackgerrit | A change was merged to openstack/python-quantumclient: Client for agent extension https://review.openstack.org/21064 | 17:19 |
| *** david-lyle has joined #openstack-dev | 17:19 | |
| *** zoresvit has quit IRC | 17:20 | |
| openstackgerrit | A change was merged to openstack/python-quantumclient: Allow known options after unknown ones in list and update command https://review.openstack.org/20574 | 17:20 |
| *** bencherian has quit IRC | 17:21 | |
| dolphm | ayoung: anything that merges after feature freeze i'd like to start tagging with something like *New in version 3.1*, and every other API-level feature change we've blocked on the api review, so i'd say definitely yes | 17:21 |
| *** Ryan_Lane has quit IRC | 17:21 | |
| *** jcmartin has quit IRC | 17:22 | |
| ayoung | dolphm, OK | 17:23 |
| *** jcmartin has joined #openstack-dev | 17:24 | |
| ayoung | dolphm, we are wordsmithing here | 17:24 |
| ayoung | https://etherpad.openstack.org/trusts_api | 17:24 |
| *** pmyers has joined #openstack-dev | 17:24 | |
| dolphm | ayoung: in part | 17:25 |
| heckj | ayoung, dolphm: should be tweakable in wording, but not in sematics | 17:25 |
| *** egallen has quit IRC | 17:25 | |
| dolphm | ayoung: we're also asking for hard changes to the API | 17:25 |
| dolphm | ayoung: i really like henrynash's suggestion of filtering GET /trusts instead of returning two separate collections in the same response to GET /users/{user_id}/trusts | 17:26 |
| *** nati_ueno_2 has joined #openstack-dev | 17:26 | |
| *** troytoman is now known as troytoman-away | 17:26 | |
| *** annegentle has joined #openstack-dev | 17:28 | |
| ayoung | dolphm, I'm find with that | 17:28 |
| *** rafaduran has quit IRC | 17:30 | |
| *** nati_ueno has quit IRC | 17:30 | |
| *** hugokuo has quit IRC | 17:30 | |
| *** nunosantos has joined #openstack-dev | 17:31 | |
| *** bswartz has quit IRC | 17:32 | |
| *** jcmartin has quit IRC | 17:32 | |
| *** jcmartin has joined #openstack-dev | 17:32 | |
| *** alszar has joined #openstack-dev | 17:33 | |
| *** dachary has joined #openstack-dev | 17:33 | |
| *** jcmartin has left #openstack-dev | 17:33 | |
| *** ctracey has quit IRC | 17:34 | |
| *** ctracey has joined #openstack-dev | 17:34 | |
| *** gyee has quit IRC | 17:35 | |
| *** alszar_ has joined #openstack-dev | 17:35 | |
| ayoung | dolphm, how about I leave get /user/trusts as is if no param, and then accept as_trustee and as_trustor as additional filters? | 17:35 |
| *** alszar has quit IRC | 17:37 | |
| *** alszar_ is now known as alszar | 17:37 | |
| *** avishay has quit IRC | 17:37 | |
| *** zoresvit has joined #openstack-dev | 17:39 | |
| *** nati_ueno_2 has quit IRC | 17:39 | |
| *** avishay has joined #openstack-dev | 17:39 | |
| *** hemnafk is now known as hemna | 17:39 | |
| *** nati_ueno has joined #openstack-dev | 17:39 | |
| *** pabelanger has quit IRC | 17:39 | |
| *** gael__ has quit IRC | 17:40 | |
| *** lloydde has joined #openstack-dev | 17:40 | |
| *** dachary has quit IRC | 17:40 | |
| *** winston-d_ has quit IRC | 17:41 | |
| *** doude has quit IRC | 17:41 | |
| dolphm | ayoung: you get the same functionality out of GET /trusts with filters | 17:41 |
| ayoung | dolphm, I'm also going to yank the LIST Trustors and list trustees APIs as I won';t have time to impl, and they are nice to have, not need to have | 17:41 |
| ayoung | dolphm, yes, just what do I return if no filter? | 17:42 |
| *** salgado-lunch is now known as salgado | 17:42 | |
| dolphm | ayoung: all trusts, regardless of enabled/disabled | 17:42 |
| ayoung | dolphm, ah, yeah, I am yanking that | 17:43 |
| dolphm | ayoung: you can achieve the old functionality with GET /trusts?enabled&trustor_user_id={my_user_id} | 17:43 |
| ayoung | filters can go in later | 17:43 |
| ayoung | not needed for first round | 17:43 |
| dolphm | ayoung: it's a one line in the docs for GET /trusts (see the examples on GET /projects and GET /domains) | 17:43 |
| ayoung | dolphm, I meant for user/id/trusts | 17:43 |
| ayoung | dolphm, lets cut scope for now | 17:44 |
| *** thingee_zzz is now known as thingee | 17:44 | |
| *** dhellmann is now known as dhellmann-afk | 17:46 | |
| *** dachary has joined #openstack-dev | 17:46 | |
| *** gyee has joined #openstack-dev | 17:46 | |
| *** diogogmt has quit IRC | 17:48 | |
| *** pmyers has quit IRC | 17:49 | |
| *** trapni has quit IRC | 17:49 | |
| *** pmyers has joined #openstack-dev | 17:49 | |
| *** diogogmt has joined #openstack-dev | 17:50 | |
| *** doude has joined #openstack-dev | 17:52 | |
| *** ondergetekende has joined #openstack-dev | 17:52 | |
| henrynash | I have a git rebasing question when someone has a moment | 17:52 |
| ayoung | henrynash, fire away | 17:53 |
| *** annegentle has quit IRC | 17:53 | |
| henrynash | ayoung: thx. So i have a patch that is dependent on the v3auth patch from guang | 17:53 |
| *** iartarisi has quit IRC | 17:54 | |
| henrynash | I made it dependant by: doing the fetch of gang's patch, checkouk FETCH_Head…. | 17:54 |
| gyee | s/gang/guang/ | 17:55 |
| gyee | :) | 17:55 |
| *** nachi has joined #openstack-dev | 17:55 | |
| henrynash | …and putting that into a branch on my git…and then rebasing my change of it….seemed fine, have submit that for review, shows dependancy on "Guang" (!) change etc. | 17:55 |
| henrynash | ..now I want to update to rebase on his latest version | 17:56 |
| henrynash | i tried the simple web rebase button - says can't do it due to conflicts | 17:56 |
| henrynash | how do I do this via git on the cmd line….I tried a few things…that didn't work :-) | 17:57 |
| henrynash | I think what I need to do is refresh the branch that Ihave with Guang's change in it…and then I rebase of it again | 17:57 |
| gyee | henrynash, I am pushing another patch shortly | 17:57 |
| *** reed has quit IRC | 17:57 | |
| henrynash | gyee: (once you've done that :-) ) | 17:58 |
| *** markwash has joined #openstack-dev | 17:59 | |
| gyee | henrynash, one of us will be busy rebasing all today today, with all these merges coming in :) | 17:59 |
| henrynash | gyee….err, both us maybe! | 18:00 |
| gyee | yeah | 18:00 |
| *** garyTh has quit IRC | 18:00 | |
| *** giulivo has quit IRC | 18:00 | |
| * heckj nods | 18:01 | |
| *** bencherian has joined #openstack-dev | 18:01 | |
| henrynash | meeting time! | 18:01 |
| *** rushiagr has quit IRC | 18:01 | |
| *** garyTh has joined #openstack-dev | 18:03 | |
| *** Mandell has joined #openstack-dev | 18:03 | |
| *** dachary has quit IRC | 18:03 | |
| *** jog0 has joined #openstack-dev | 18:05 | |
| *** jog0 has quit IRC | 18:05 | |
| *** njoy_ has quit IRC | 18:05 | |
| *** njoy_ has joined #openstack-dev | 18:05 | |
| *** derekh has quit IRC | 18:05 | |
| *** jog0 has joined #openstack-dev | 18:05 | |
| *** yidclare has joined #openstack-dev | 18:05 | |
| *** nati_ueno has quit IRC | 18:06 | |
| *** roampune has left #openstack-dev | 18:06 | |
| *** nati_ueno has joined #openstack-dev | 18:06 | |
| *** roampune has joined #openstack-dev | 18:07 | |
| *** Ryan_Lane has joined #openstack-dev | 18:08 | |
| *** amerine has joined #openstack-dev | 18:10 | |
| *** ondergetekende has quit IRC | 18:10 | |
| *** nati_ueno has quit IRC | 18:11 | |
| *** darraghb has quit IRC | 18:12 | |
| openstackgerrit | A change was merged to openstack/quantum: Exit if DHCP agent interface_driver is not defined https://review.openstack.org/22290 | 18:13 |
| *** zing has quit IRC | 18:16 | |
| *** mohits has joined #openstack-dev | 18:17 | |
| *** adjohn has joined #openstack-dev | 18:17 | |
| *** xga has joined #openstack-dev | 18:17 | |
| *** jgallard has quit IRC | 18:18 | |
| *** buzztroll_ has quit IRC | 18:18 | |
| *** bswartz has joined #openstack-dev | 18:24 | |
| *** ndipanov has quit IRC | 18:25 | |
| heckj | ayoung: you going to appear today? keystone meeting | 18:26 |
| *** nati_ueno has joined #openstack-dev | 18:28 | |
| cloudchimp | Hi All, still building multi-node grizzly cluster for early system integration work. But just noticed that glance had a failure because warlock 0.7 or greater is required. Is warlock 0.1 still supported? | 18:29 |
| *** xga has quit IRC | 18:31 | |
| *** egallen has joined #openstack-dev | 18:31 | |
| openstackgerrit | A change was merged to openstack/keystone: Update sample_data.sh to match docs https://review.openstack.org/22237 | 18:33 |
| openstackgerrit | A change was merged to openstack/horizon: Adding network-topology view for quantum https://review.openstack.org/21916 | 18:34 |
| openstackgerrit | A change was merged to openstack/python-keystoneclient: Implements token expiration handling https://review.openstack.org/20817 | 18:34 |
| openstackgerrit | A change was merged to openstack/keystone: Use oslo-config-2013.1b3 https://review.openstack.org/22080 | 18:34 |
| cloudchimp | Even better question. What is python-warlock 0.7? I thought it was at version 0.3.1 | 18:35 |
| *** avishay has quit IRC | 18:36 | |
| openstackgerrit | A change was merged to openstack/keystone: Remove old, outdated keystone devref docs https://review.openstack.org/22301 | 18:36 |
| *** yolanda has quit IRC | 18:37 | |
| openstackgerrit | A change was merged to openstack/python-keystoneclient: Use oslo-config-2013.1b3 https://review.openstack.org/22126 | 18:38 |
| *** rushiagr has joined #openstack-dev | 18:39 | |
| openstackgerrit | A change was merged to openstack/nova: Add a safe_minidom_parse_string function. https://review.openstack.org/22309 | 18:39 |
| *** rushiagr1 has joined #openstack-dev | 18:40 | |
| *** rushiagr2 has joined #openstack-dev | 18:41 | |
| *** egallen has quit IRC | 18:43 | |
| *** rushiagr has quit IRC | 18:44 | |
| *** rushiagr1 has quit IRC | 18:45 | |
| *** lcheng has joined #openstack-dev | 18:45 | |
| *** psedlak has quit IRC | 18:46 | |
| *** corXi has quit IRC | 18:46 | |
| *** johnthetubaguy has left #openstack-dev | 18:50 | |
| *** AnilV4 has quit IRC | 18:50 | |
| lifeless | sdague: that was 0430, so no :) | 18:52 |
| *** mrodden has quit IRC | 18:55 | |
| *** Yada has quit IRC | 18:55 | |
| *** cp16net is now known as cp16net|away | 18:56 | |
| *** cp16net|away is now known as cp16net | 18:56 | |
| *** cdub_ has joined #openstack-dev | 18:57 | |
| *** YorikSar has joined #openstack-dev | 18:57 | |
| *** zoresvit has quit IRC | 18:57 | |
| sdague | lifeless: still digging on it :) | 18:58 |
| *** morganfainberg has joined #openstack-dev | 18:58 | |
| sdague | lifeless: you see the bug? | 18:58 |
| *** roampune1 has joined #openstack-dev | 18:58 | |
| *** roampune has quit IRC | 18:58 | |
| sdague | basically we had to revert a patch because it managed to fail testr in such a spectacular way that it wiped out hundreds of other tests from running, and yet reported no fails | 18:59 |
| *** radez is now known as radez_g0n3 | 18:59 | |
| *** lloydde has quit IRC | 19:00 | |
| *** rushiagr has joined #openstack-dev | 19:00 | |
| dolphm | ayoung: instead of exposing enabled/disabled at the API level and overloading the term (it's use is different here), how about an attribute in the backend called 'deleted' which defaults to false | 19:01 |
| *** pabelanger has joined #openstack-dev | 19:02 | |
| dolphm | ayoung: trusts that have been DELETE'd through the API have deleted=True, and are not returned through GET /trusts | 19:02 |
| dolphm | ayoung: but their still available for auditing | 19:02 |
| *** spzala has quit IRC | 19:02 | |
| lifeless | sdague: thats pretty special; is it what clarkb pinged me on when he was going to sleep last night? | 19:02 |
| ayoung | dolphm, OK. And audit will have to go direct to backend to start | 19:02 |
| *** thingee is now known as thingee_zzz | 19:03 | |
| *** alunduil has quit IRC | 19:03 | |
| sdague | lifeless: this is something new I think | 19:03 |
| ayoung | we can provide audit and admin apis for cleanup later? | 19:03 |
| *** thingee_zzz is now known as thingee | 19:03 | |
| lifeless | what bug 3 ? | 19:03 |
| uvirtbot | Launchpad bug 3 in launchpad "Custom information for each translation team" [Low,Fix released] https://launchpad.net/bugs/3 | 19:03 |
| lifeless | s/3/#/ | 19:03 |
| sdague | lifeless: so I've tracked it into the XMLmatcher | 19:03 |
| *** rushiagr2 has quit IRC | 19:03 | |
| sdague | lifeless: https://bugs.launchpad.net/nova/+bug/1130146 | 19:03 |
| uvirtbot | Launchpad bug 1130146 in testrepository "Unit tests broke with recent commit" [Critical,Confirmed] | 19:03 |
| dolphm | ayoung: i'd want to include listing expired tokens in that conversation as well | 19:03 |
| *** AnilV4 has joined #openstack-dev | 19:04 | |
| ayoung | dolphm, so expired trusts are considered deleted? | 19:04 |
| dolphm | ayoung: is there a difference in utility between an expired trust and a DELETE'd trust? | 19:05 |
| dolphm | ayoung: i'm thinking it's the same lack of difference as between an expired token and a revoked token | 19:05 |
| ayoung | dolphm, notreally | 19:05 |
| ayoung | agreed | 19:05 |
| *** AlanClark__ has quit IRC | 19:05 | |
| *** zing has joined #openstack-dev | 19:06 | |
| *** AlanClark__ has joined #openstack-dev | 19:06 | |
| *** imsplitbit has quit IRC | 19:06 | |
| sdague | lifeless: so nova.tests.matchers.XMLMatcher() is the culprit, something about how it interacts with assertThat is causing the bork | 19:06 |
| dolphm | ayoung: are you supporting expiration time defined by the api-user? or is that a deployment configuration | 19:06 |
| ayoung | dolphm, for trusts, expiry is optional | 19:07 |
| ayoung | if it is not specified, they don't time out | 19:07 |
| ayoung | In the original use case, we didn't need them | 19:07 |
| *** danwent has quit IRC | 19:07 | |
| lifeless | sdague: lets see | 19:07 |
| clarkb | lifeless: it is the same thing. I submitted a bug against subunit | 19:07 |
| gyee | ayoung, dolphm, https://review.openstack.org/#/c/21487 | 19:08 |
| *** sacharya has quit IRC | 19:08 | |
| ayoung | but yes, it will be supported. I have an update that supports it | 19:08 |
| sdague | clarkb: I'm not sure it's a subunit bug | 19:08 |
| lifeless | sdague: that looks fine by eyeball | 19:08 |
| gyee | I changed /auth to /auth/tokens | 19:08 |
| lifeless | sdague: can I ask how you pinpointed it ? | 19:08 |
| ayoung | gyee, w00t | 19:08 |
| clarkb | sdague: not directly but subunit should fail appropriately | 19:08 |
| clarkb | lifeless: see the bug | 19:08 |
| ayoung | gyee, I assume you had to change tests too, right? | 19:08 |
| lifeless | clarkb: kk | 19:08 |
| sdague | lifeless: yeh, the bug documents this pretty well | 19:08 |
| lifeless | clarkb: whats the # ? | 19:08 |
| clarkb | https://bugs.launchpad.net/nova/+bug/1130146 | 19:09 |
| uvirtbot | Launchpad bug 1130146 in testrepository "Unit tests broke with recent commit" [Critical,Confirmed] | 19:09 |
| gyee | ayoung, yes, tests are updated | 19:09 |
| lifeless | oh right same one | 19:09 |
| lifeless | lets see | 19:09 |
| clarkb | sdague: in my tracing the matcher in nova/tests/matchers.py is returning a non boolean | 19:09 |
| ayoung | gyee, I'll +1 | 19:09 |
| sdague | clarkb: yep, I see that as well | 19:09 |
| clarkb | sdague: I am willing to bet that asserThat expects more specific True False type values | 19:09 |
| sdague | clarkb: yes, it's true | 19:10 |
| sdague | there is a lot of odd inverted logic as well | 19:10 |
| dolphm | gyee: from keystone.auth.core import AuthMethodHandler ? | 19:10 |
| ayoung | gyee, this is the one time diff to previous patch worked really well for me | 19:11 |
| *** doude has left #openstack-dev | 19:11 | |
| ayoung | the only other change I saw was the common config change | 19:11 |
| ayoung | AH! | 19:11 |
| *** pabelanger has quit IRC | 19:12 | |
| gyee | dolphm, yes, heckj wanted the import to be specific | 19:12 |
| heckj | preference towards it, but I can be argued out of it | 19:12 |
| dolphm | heckj: ^ we don't do that for other packages | 19:12 |
| heckj | I get a rash when I see from … import * | 19:13 |
| gyee | heckj, yeah, same here | 19:13 |
| dolphm | heckj: we do from ...core import * in catalog, identity, token and policy -- but *only* from core | 19:13 |
| lifeless | the bug is on line 387 | 19:13 |
| lifeless | sdague: ^ | 19:13 |
| heckj | gyee: but don't block this, its the pattern we're already using, even if I don't care for it :-) | 19:13 |
| lifeless | of mactchers.py | 19:13 |
| *** jaypipes has quit IRC | 19:14 | |
| dolphm | gyee: i'm fine with an explicit import, but everything from core should be explicitly imported then | 19:14 |
| *** sacharya has joined #openstack-dev | 19:14 | |
| gyee | dolphm, there's only one thing in core | 19:14 |
| gyee | so we should be fine | 19:14 |
| dolphm | gyee: then i'm cool with it | 19:14 |
| sdague | lifeless: what should that be? | 19:15 |
| lifeless | sdague: the match() method must return either a mismatch or None | 19:15 |
| lifeless | sdague: its returning a string on that branch | 19:15 |
| dolphm | gyee: GET /auth/tokens/revoked isn't in the spec, i don't think | 19:16 |
| gyee | dolphm, no, that's for PKI tokens only | 19:17 |
| sdague | lifeless: ok, sorry, not seeing that in the density of the code, the only returns I see seem to be objects, or True, which filters to the top | 19:18 |
| gyee | I suspect that API will go away once we have short-live PKI tokens | 19:18 |
| *** jimfehlig has quit IRC | 19:18 | |
| sdague | lifeless: also, why is assertThat exploding | 19:18 |
| sdague | in a weird way | 19:18 |
| dolphm | gyee: i'm confused | 19:18 |
| gyee | dolphm, /auth/tokens/revoked is basically returning a CRL in CMS | 19:19 |
| ayoung | A TRL as it were | 19:19 |
| gyee | I don't think keystoneclient is even checking it | 19:19 |
| dolphm | gyee: then why expose it? | 19:19 |
| ayoung | gyee, it was | 19:19 |
| *** buzztroll_ has joined #openstack-dev | 19:19 | |
| ayoung | dolphm, no it is used, it has been there for a long while | 19:20 |
| gyee | dolphm, I was replicating what we have in /v2 | 19:20 |
| ayoung | if the client can;t get to /auth/tokens/revoked the token validation fails | 19:20 |
| dolphm | ayoung: gyee: show me? https://github.com/openstack/identity-api/ | 19:20 |
| *** adjohn has quit IRC | 19:20 | |
| gyee | dolphm, that API is not there right now | 19:20 |
| gyee | ayoung, can you amend it to your /auth/tokens changes? | 19:21 |
| *** rushiagr has left #openstack-dev | 19:21 | |
| dolphm | GET /v3/auth/tokens/revocation ? | 19:23 |
| gyee | yes | 19:24 |
| gyee | sorry GET /v3/auth/tokens/revoked | 19:24 |
| dolphm | or GET /v3/auth/tokens?deleted or GET /v3/auth/tokens?revoked | 19:24 |
| gyee | dolphm, its not returning tokens, just a TRL | 19:25 |
| *** annegentle has joined #openstack-dev | 19:26 | |
| *** aeperezt has quit IRC | 19:26 | |
| ayoung | gyee, yes, I can ammend | 19:26 |
| *** annegentle has quit IRC | 19:27 | |
| gyee | dolphm, we're good? | 19:27 |
| *** annegentle has joined #openstack-dev | 19:27 | |
| dolphm | gyee: ayoung: i'd rather see it as a seperate change rather than hijack a change that's only partially related | 19:27 |
| sdague | lifeless: so the testr bug here appears to be the fact that when _run_test runs, the finally block puts the test on the queue | 19:27 |
| *** FlorianOtel has quit IRC | 19:28 | |
| dolphm | gyee: i'm suggesting that r' | 19:28 |
| sdague | but because it didn't process, some time later it gets rejected and dropped silently | 19:28 |
| sdague | so it seems that in these cases it needs to add an explicit fail | 19:28 |
| eafonichev | hi guys! I need some help with horizon(stable/folsom). Trying to customize logo I got error like this 'OfflineGenerationError: You have offline compression enabled but key "1056718f92f8d4204721bac759b3871a" is missing from offline manifest. You may need to run "python manage.py compress"'. Invoking collectstatic and compress doesn't work. | 19:28 |
| *** jaypipes has joined #openstack-dev | 19:29 | |
| gyee | dolphm, without that, it'll be hard to check for revoked tokens without calling back to Keystone | 19:29 |
| *** pabelanger has joined #openstack-dev | 19:29 | |
| lifeless | sdague: yes, something like that will be the fix. | 19:29 |
| *** melwitt has joined #openstack-dev | 19:29 | |
| lifeless | sdague: I'll make a SCCE and fix later | 19:29 |
| *** romcheg has joined #openstack-dev | 19:29 | |
| dolphm | gyee: so, what's the structure of the response? | 19:29 |
| lifeless | sdague: assertThat is exploding because | 19:29 |
| lifeless | sdague: the match object is returning "something" | 19:30 |
| dolphm | gyee: i thought the token revocation list was a cache maintained by middleware in a file | 19:30 |
| gyee | dolphm bas64(CMS) | 19:30 |
| lifeless | sdague: rather than a Mismatch object | 19:30 |
| *** aeperezt has joined #openstack-dev | 19:30 | |
| jpich | eafonichev: Unfortunately the compress errors tend to hide real issues. If this is a dev or test environment (not production) you could try setting DEBUG to True in the local_settings to see what's the real problem | 19:30 |
| gyee | dolphm, token revocation list is fetched via /v3/auth/tokens/revoked | 19:31 |
| eafonichev | jpick, ok, thanks, will try | 19:31 |
| dolphm | gyee: is that call applicable to uuid tokens? | 19:31 |
| sdague | lifeless: ok, well I just keep dumping my findings into the bug. I think I'm at about the extent of things here | 19:31 |
| sdague | so I'll let you run with it now that it's morning there :) | 19:31 |
| dolphm | gyee: if it's not, then the call should certainly appear as an API extension /v3/auth/tokens/OS:PKI/revoked | 19:32 |
| lifeless | sdague: a mismatch can return multiple attached errors for one failure | 19:32 |
| gyee | dolphm, https://github.com/openstack/python-keystoneclient/blob/master/keystoneclient/middleware/auth_token.py#L939 | 19:32 |
| gyee | no it is only used for PKI tokens I think | 19:32 |
| dolphm | *facepalm* | 19:32 |
| lifeless | sdague: so yeah its just that line 'return result' is the bug. | 19:32 |
| gyee | ha | 19:32 |
| eafonichev | jpich: you were so right! indeed, there is some another problem, thanks a lot | 19:33 |
| *** cp16net is now known as cp16net|away | 19:33 | |
| dolphm | this needs to be independently documented as an extension then, not part of identity-api-v3.md | 19:33 |
| jpich | eafonichev: Cool! Good luck with it :) | 19:33 |
| dolphm | identity-api-v3-os-pki.md or something? | 19:34 |
| gyee | dolphm, fine with me | 19:34 |
| gyee | ayoung, we're good? | 19:34 |
| gyee | so GET /v3/auth/tokens/OS:PKI/revoked then? | 19:35 |
| dolphm | gyee: sure | 19:35 |
| dolphm | or GET /v3/auth/tokens/OS:PKI/trl if that's more broadly appropriate? (is it even json?) | 19:36 |
| gyee | not json | 19:37 |
| gyee | CMS is ASN.1 | 19:37 |
| ayoung | ASN.1 which is then encoded... | 19:37 |
| gyee | dolphm, I think we should stick with revoked as it is useful for x.509 as well | 19:38 |
| dolphm | i'm pretty sure we require the client to specify Accept: application/json somewhere in middleware (json_body?) | 19:38 |
| sdague | lifeless: ok, I'm not seeing how it would be anything other than an XMLMismatch or True gets returned to that point | 19:38 |
| gyee | trl is too implementation-specific | 19:38 |
| dolphm | gyee: sure, whatevers more appropriate | 19:38 |
| *** romcheg has left #openstack-dev | 19:38 | |
| lifeless | sdague: ' | 19:39 |
| *** n0ano has quit IRC | 19:39 | |
| lifeless | elif result is not True: | 19:39 |
| lifeless | return result' | 19:39 |
| lifeless | sdague: doesn't that return result when it is *not* True ? | 19:39 |
| lifeless | sdague: if you put a pdb statement in there and run it, I would be interested in the value of result :) | 19:40 |
| clarkb | <nova.tests.matchers.XMLTagMismatch object at 0x4387710> | 19:40 |
| lifeless | sdague: if result is True at that point, I'll be amazed. | 19:40 |
| *** nachi has quit IRC | 19:40 | |
| lifeless | clarkb: oh, its recursive? | 19:40 |
| clarkb | _compare* is recursive | 19:40 |
| gyee | dolphm, ayoung, https://review.openstack.org/#/c/21487 | 19:41 |
| ayoung | gyee, BTW, foreign keys are out as Id might be in a different backend. I started out with FK but had to remove them | 19:41 |
| clarkb | I did it the lazy way and just put a print statement before return resutl :) | 19:41 |
| gyee | ayoung, I was worrying about unusable data | 19:41 |
| gyee | say a role definition is deleted, how does it triggered a data cleanup in trust? | 19:42 |
| ayoung | gyee, I agree, but the thing is, if the trut is bad, no token.... | 19:42 |
| ayoung | we can do all of the the test work we want....make it nice. right now I just need it to be capable of working | 19:42 |
| sdague | lifeless: yeh, it has to be recursive, xml is nested :) | 19:42 |
| *** cp16net|away is now known as cp16net | 19:42 | |
| gyee | ayoung, I am fine with a FIXME right now | 19:43 |
| ayoung | gyee, +2 from me. | 19:43 |
| ayoung | gyee, OK | 19:43 |
| gyee | ayoung, with so many backend, I would even argue that we generally have a data cleanup problem | 19:44 |
| *** yidclare has quit IRC | 19:46 | |
| *** novas0x2a|laptop has joined #openstack-dev | 19:46 | |
| ayoung | gyee, can you git commit amend with a friendly title? Say token v3? | 19:48 |
| ayoung | blueprint pluggable-identity-authentication-handlers blueprint stop-ids-in-uris blueprint multi-factor-authn (just the plumbing) v3 authentication and token APIs | 19:48 |
| ayoung | kind of rolls off the tongue, I know | 19:48 |
| *** adjohn has joined #openstack-dev | 19:48 | |
| gyee | ayoung, sure, give a sec | 19:49 |
| dolphm | gyee: you're passing token='ADMIN' in an auth request (get_scoped_token) | 19:50 |
| ewindisch | russellb: I'd love to get your feedback on my kombu-fast-serialization patch. | 19:50 |
| dolphm | gyee: i assume you don't actually need an X-Auth-Token to authenticate.. | 19:50 |
| gyee | dolphm, you do | 19:50 |
| * dolphm *brainsplode* | 19:50 | |
| dolphm | gyee: what? | 19:51 |
| gyee | authentication is a privilege operation :( | 19:51 |
| dolphm | gyee: wtf | 19:51 |
| gyee | yeah, same here :) | 19:51 |
| dolphm | gyee: so, keystone doesn't support authentication, period... | 19:51 |
| gyee | authentication need to either accompany by a ADMIN token in the header or SSL | 19:52 |
| dims | do we need this quick for the pypi issue? - https://bugs.launchpad.net/tempest/+bug/1130273 | 19:53 |
| dolphm | gyee: that's completely unusable by anyone -- why? | 19:53 |
| uvirtbot | Launchpad bug 1130273 in tempest "install_venv_common.py is not robust to pypi instability" [Undecided,New] | 19:53 |
| ayoung | gyee, merging your changes to get ahead of the curve and I get | 19:53 |
| ayoung | from oslo.config import cfg | 19:53 |
| ayoung | ImportError: No module named oslo.config | 19:53 |
| gyee | wth? I didn't touch oslo | 19:53 |
| dolphm | gyee: authenticate isn't wrapped by @protected... where's the requirement coming from? i'm not clear if you did this intentionally or not | 19:53 |
| ayoung | dolphm, only an admin user can verify a token | 19:53 |
| ayoung | wait... | 19:54 |
| sdague | lifeless: the XMLTagMismatch is what's returned when things go funny | 19:54 |
| dolphm | ayoung: this is authentication | 19:54 |
| ayoung | you don't need an X-Auth-Token to authenticate | 19:54 |
| ayoung | you can use it token-to-token | 19:54 |
| dolphm | ayoung: like "hi my name is ayoung, here's my password, and here's the admin_token; can i please have a token?" | 19:55 |
| *** gabrielhurley has joined #openstack-dev | 19:55 | |
| ayoung | dolphm, heh | 19:55 |
| ayoung | get_scoped_token needs an unscoped | 19:55 |
| ayoung | unscoped doesn't need to be admin | 19:55 |
| gyee | ayoung, now I am getting the oslo error | 19:56 |
| gyee | what the hell's going on here | 19:56 |
| gyee | it was working 30 mins ago | 19:56 |
| ayoung | dolphm, gyee I think it comes from | 19:57 |
| dolphm | gyee: if i remove token from that call test_v3_identity passes just fine | 19:57 |
| dolphm | ayoung: gyee: ... install oslo? | 19:57 |
| gyee | dolphm, yeah, I have @controller.protected in my v3 authenticate call | 19:57 |
| gyee | removing it now | 19:57 |
| dolphm | gyee: ayoung: http://paste.openstack.org/raw/31968/ | 19:58 |
| ayoung | https://review.openstack.org/#/c/21487/11..12/keystone/cli.py | 19:58 |
| *** yidclare has joined #openstack-dev | 19:59 | |
| dolphm | gyee: which method is that? keystone.auth.controllers ... authenticate() ? | 19:59 |
| gyee | authenticate_for_token | 19:59 |
| dolphm | gyee: ah, i had misread that as "authenticate with token" | 20:00 |
| dolphm | given that it was protected | 20:00 |
| *** mrodden has joined #openstack-dev | 20:00 | |
| dolphm | gyee: while your at is, is all the is_admin stuff going away in the tests? | 20:01 |
| dolphm | gyee: i assume that's all fallout from protecting authentication lol | 20:01 |
| gyee | dolphm, no, is_admin is still needed | 20:01 |
| gyee | user_foo is not an admin user | 20:01 |
| dolphm | gyee: okay? | 20:02 |
| *** danwent has joined #openstack-dev | 20:02 | |
| lifeless | sdague: FYI I have a protocol revision I'm working on for subunit that will offer much better recovery in this sort of situation | 20:05 |
| *** koolhead17 has quit IRC | 20:05 | |
| lifeless | sdague: its weekends and evenings stuff though, unless mordred decides to offer HP time on it :) | 20:05 |
| mordred | lifeless: you're a high-level tech guy - do you think it's important for you to work on? :) | 20:06 |
| lifeless | sdague: oh - so self.expected and self.actual are the next candidates for the issue | 20:06 |
| ayoung | dolphm, OK, so merging in the oslo config broke things | 20:06 |
| ayoung | No virtual environment found...create one? (Y/n) y | 20:06 |
| ayoung | ImportError: No module named oslo.config | 20:06 |
| ayoung | it doesn't get fetched automagically | 20:07 |
| lifeless | mordred: I think protocol design is hard and has far reaching implications; so yeah :> But there are many other things I need to work on too. | 20:07 |
| dolphm | ayoung: http://paste.openstack.org/raw/31968/ | 20:07 |
| *** sandywalsh has quit IRC | 20:07 | |
| ayoung | chicken egg thing | 20:07 |
| gyee | ayoung, I am aving the same problem here | 20:07 |
| ayoung | gyee, I understand | 20:07 |
| ayoung | gyee, I htink we need to install so the venv set up runs | 20:07 |
| mordred | lifeless: I know - but I trust you to prioritize your time appropriately | 20:07 |
| openstackgerrit | A change was merged to openstack/keystone: Implement name space for domains https://review.openstack.org/21951 | 20:07 |
| ayoung | gyee, I think it needs to be | 20:08 |
| ayoung | sudo easy_install http://tarballs.openstack.org/oslo-config/oslo-config-2013.1b3.tar.gz#egg=oslo-config | 20:08 |
| ayoung | but that didn't do it either | 20:08 |
| *** AnilV4 has quit IRC | 20:08 | |
| gyee | ayoung, we need it inside venv | 20:08 |
| lifeless | mordred: yeah; the tough thing for me in this context is that its close to being a conflict of interest; subunit&testr are personal projects, *and* projects HP/Openstack depend on. | 20:08 |
| * dolphm wonders if i'm the only one that manages my own venv | 20:09 | |
| ayoung | /usr/lib/python2.7/site-packages/oslo_config-2013.1b3-py2.7.egg/ | 20:09 |
| lifeless | sdague: state = XMLMatchState(self.expected_xml, actual_xml) | 20:09 |
| ayoung | dolphm, somehow mine just got wiped and I needed to rebuild it | 20:09 |
| lifeless | sdague: I think thats the bug | 20:09 |
| lifeless | sdague: the details object should contain Content objects, not strings | 20:09 |
| lifeless | sdague: but self.expected_xml is a string (possibly unicode?) | 20:09 |
| lifeless | sdague: that will flow into the Mismatch | 20:10 |
| lifeless | sdague: and get_details() will then find a string where it expects a Content object, which will trigger the 'str object has no attribute content_type' error we see. | 20:10 |
| mordred | lifeless: fair | 20:12 |
| *** imsplitbit has joined #openstack-dev | 20:12 | |
| ayoung | dolphm, how do I manage my own venv | 20:12 |
| lifeless | mordred: in nearly any other context I would do first and tell you later :) | 20:13 |
| topol | ayoung, so I added in the use_dumb_member = True and I think everything is happy now | 20:13 |
| dolphm | ayoung: virtualenv --help :) | 20:13 |
| gyee | dolphm, launch a VM off nova? | 20:13 |
| *** adalbas has quit IRC | 20:13 | |
| dolphm | gyee: lol | 20:13 |
| *** AlanClark__ has quit IRC | 20:13 | |
| topol | dolphm, that is so wrong on code freeze day | 20:13 |
| dolphm | topol: lol | 20:14 |
| *** AlanClark__ has joined #openstack-dev | 20:14 | |
| topol | :-) | 20:14 |
| dolphm | ayoung: do you know where your venv is? | 20:14 |
| dolphm | if you do, source it's bin/activate, and then pip install the oslo tarball (no sudo) | 20:14 |
| ayoung | dolphm, it is gone | 20:15 |
| lifeless | mordred: you might find the three top posts on http://rbtcollins.wordpress.com/ interesting | 20:15 |
| ayoung | dolphm, this is show stopper | 20:15 |
| ayoung | at least for dev | 20:15 |
| gyee | ayoung, just manually extract the oslo tarball somewhere for now | 20:16 |
| *** tiamar has quit IRC | 20:17 | |
| ayoung | gyee, I have it in site-packages | 20:17 |
| gyee | then do env PYTHONPATH=oslo_path ./run_tests.sh to recreate your .venv | 20:17 |
| dolphm | ayoung: virtualenv ~/somedir && source ~/somedir/bin/activate && pip install -r tools/pip-requires && pip install -r tools/test-requires && sh run_tests.sh -N | 20:18 |
| gyee | dolphm, now I got this | 20:19 |
| gyee | Could not find any downloads that satisfy the requirement oslo-config from http://tarballs.openstack.org/oslo-config/oslo-config-2013.1b3.tar.gz | 20:19 |
| dolphm | gyee: damn, i thought by "i got this" you meant "... under control" | 20:19 |
| gyee | translates into "shit just hit the fan" in Chinese :) | 20:20 |
| *** sandywalsh has joined #openstack-dev | 20:20 | |
| openstackgerrit | A change was merged to openstack/oslo-config: Add missing files to the tarball. https://review.openstack.org/22344 | 20:20 |
| dolphm | gyee: pip install http://tarballs.openstack.org/oslo-config/oslo-config-2013.1b3.tar.gz#egg=oslo-config | 20:21 |
| dolphm | gyee: what's that do? | 20:21 |
| ayoung | dolphm, here's the deal. At 5 I have to get my kid. We're going to drive to my folks place, which is 2 hours away. After that, I will be back on line. So I have about 1.5 hours now, and then.... | 20:21 |
| *** Gomes has joined #openstack-dev | 20:21 | |
| lifeless | clarkb: do you have enough to address the issue ? | 20:22 |
| *** Yada has joined #openstack-dev | 20:22 | |
| dolphm | ayoung: i have about the same amount of time, and then food -> coffee -> and ready to stay up all night with you :P | 20:22 |
| dolphm | ayoung: does pip installing oslo-config directly work for you? ^^ | 20:23 |
| clarkb | lifeless: uh I wasn't really following along | 20:23 |
| *** voliveirajr has joined #openstack-dev | 20:23 | |
| ayoung | dolphm, so long as Freeze doesn't happen at COB tonight, I'm cool | 20:23 |
| ayoung | dolphm, venv is building | 20:23 |
| dolphm | ayoung: awesome | 20:23 |
| openstackgerrit | A change was merged to openstack/nova: Assign unique names with os-multiple-create. https://review.openstack.org/21386 | 20:23 |
| clarkb | lifeless: XMLMatchState should be passed a COntent object instead of a string? | 20:23 |
| ayoung | dolphm, I did it the way that gyee suggested | 20:23 |
| openstackgerrit | A change was merged to openstack/nova: Stub additional FloatingIP methods in FlatManager https://review.openstack.org/20642 | 20:23 |
| clarkb | lifeless: yeah I think I should be able to correct the problem wit hthat info | 20:23 |
| ayoung | PYTHONPATH= | 20:23 |
| *** dhellmann-afk is now known as dhellmann | 20:23 | |
| ayoung | I'm going to undo the osl install and try the one that dprince just merged | 20:23 |
| gyee | ayoung, install_venv still choked on me | 20:24 |
| ewindisch | ayoung: yeah, the oslo-config thing has been a bit of a pain to come this late. I had to rebase and fix merge conflicts on a changes that just needed a +2… | 20:24 |
| lifeless | clarkb: well, there are clearly several routes to fix it. | 20:24 |
| lifeless | clarkb: the Mismatch object's details dict needs to be str->Content though | 20:24 |
| clarkb | ok | 20:24 |
| lifeless | clarkb: its currently str->str AFAICT | 20:25 |
| lifeless | at least in this failure mode | 20:25 |
| danpb | lifeless: i've already posted exactly that fix | 20:25 |
| lifeless | danpb: oh cool | 20:25 |
| clarkb | woot | 20:25 |
| lifeless | danpb: I missed that while I took Cynthia to school | 20:25 |
| lifeless | danpb: linky link ? | 20:25 |
| danpb | https://review.openstack.org/#/c/22350/ | 20:26 |
| *** radez_g0n3 is now known as radez | 20:26 | |
| danpb | presume this is a new API requirement by latest testtools ? since i swear this xml matcher did the right thing in the past | 20:27 |
| *** yolanda has joined #openstack-dev | 20:30 | |
| *** dachary has joined #openstack-dev | 20:30 | |
| *** pabelanger has joined #openstack-dev | 20:31 | |
| lifeless | devananda: https://review.openstack.org/#/c/21481/ needs a rebase | 20:32 |
| lifeless | danpb: I don't believe so | 20:33 |
| lifeless | danpb: last changed may 2010 | 20:34 |
| lifeless | danpb: sorry, june 2010. | 20:34 |
| danpb | oh, wierd | 20:34 |
| danpb | maybe we were just lucky that it didn't previously trip up the test harness | 20:35 |
| lifeless | Yeah, I think you had no failures there until now | 20:35 |
| devananda | lifeless: yes, and needed a fix as well. about to do push it again | 20:35 |
| lifeless | devananda: ah cool | 20:36 |
| ayoung | dolphm, OK, after I did a pip instll and reran the tests, building the venv continued to the point that I got and an error, but the tests seem to be running. | 20:38 |
| dolphm | ayoung: what was the error? | 20:38 |
| ayoung | dolphm, could not apply redhat-greelet patch? | 20:38 |
| ayoung | greenlet | 20:38 |
| *** jimfehlig has joined #openstack-dev | 20:38 | |
| *** digitalsanctum has quit IRC | 20:39 | |
| ayoung | gyee, File "/opt/stack/keystone/tests/test_v3_auth.py", line 276, in test_revoke_token | 20:39 |
| ayoung | failed after rebase. | 20:39 |
| ayoung | I think that was the only failure | 20:40 |
| ayoung | {"error": {"message": "revocation_list() got an unexpected keyword argument 'PKI'", "code": 400, "title": "Bad Request"}} | 20:40 |
| ewindisch | dhellmann: re:redis-driver -- thoughts to agreeing to the tests and letting me add mock-outs later? | 20:40 |
| ayoung | I wonder if something is parsing the OS:PKI url? | 20:40 |
| gyee | ayoung, yeah, we need to change it to OS-PKI instead | 20:40 |
| *** lglenden_ has joined #openstack-dev | 20:41 | |
| dhellmann | ewindisch: yeah, that makes sense | 20:41 |
| dhellmann | ewindisch: I'll post on the changeset after this meeting ends | 20:41 |
| ayoung | gyee, coo | 20:42 |
| ewindisch | thanks | 20:42 |
| dolphm | gyee: my bad | 20:42 |
| *** pixelbeat has quit IRC | 20:42 | |
| *** armaan has quit IRC | 20:43 | |
| *** lglenden has quit IRC | 20:44 | |
| ayoung | gyee, when you post for review, clean up the commit message | 20:44 |
| gyee | ayoung, you want me to remove the bp/s in the commit? | 20:45 |
| ayoung | gyee, nah | 20:45 |
| ayoung | gyee, just put a top level description | 20:45 |
| ayoung | v3 auth api | 20:45 |
| gyee | I did | 20:45 |
| ayoung | https://review.openstack.org/#/c/21487 doesn't show it | 20:46 |
| *** sulrich has joined #openstack-dev | 20:46 | |
| devananda | lifeless: 21481 is up, with a vastly different test_migrations file | 20:46 |
| lifeless | devananda: want eyeballs? | 20:46 |
| devananda | lifeless: please | 20:47 |
| devananda | lifeless: this is holding up the whole stack :( | 20:47 |
| ayoung | gyee, do you have new one to post? | 20:47 |
| ayoung | If not, I can fix and repost | 20:47 |
| gyee | ayoung, yeah, in a min | 20:47 |
| *** annegentle has quit IRC | 20:47 | |
| *** agentle_ has joined #openstack-dev | 20:47 | |
| ayoung | gyee, you figure out your venv issues? | 20:49 |
| gyee | ayoung, yeah, I had to manually do a few things | 20:53 |
| *** yolanda has quit IRC | 20:53 | |
| gyee | the tests are running now | 20:53 |
| *** Gomes has quit IRC | 20:54 | |
| *** digitalsanctum has joined #openstack-dev | 20:55 | |
| *** stevebaker has quit IRC | 20:57 | |
| gyee | ayoung, dolphm, https://review.openstack.org/#/c/21487 | 20:58 |
| *** dprince has quit IRC | 20:59 | |
| *** digitalsanctum has quit IRC | 21:01 | |
| *** epim has joined #openstack-dev | 21:01 | |
| *** digitalsanctum has joined #openstack-dev | 21:02 | |
| hemna | gate-tempest-devstack-vm-full is failing now due to missing nova.networks table not existing. | 21:02 |
| *** timello_ has quit IRC | 21:02 | |
| dolphm | gyee: you modified pip-requries | 21:05 |
| *** olaph has quit IRC | 21:05 | |
| *** maurosr has quit IRC | 21:06 | |
| *** olaph has joined #openstack-dev | 21:06 | |
| devananda | lifeless: also, care to remove your -1 on 21605 since there is now a bug tracking the issue you raised? | 21:06 |
| gyee | dolphm, yeah, sorry, corrected it | 21:08 |
| ayoung | dolphm, how do filters show up in the controllers? | 21:08 |
| openstackgerrit | A change was merged to openstack/quantum: Improve error handling when nvp and quantum are out of sync https://review.openstack.org/22128 | 21:08 |
| dolphm | gyee: trying to use v3 auth and it's not going well | 21:08 |
| *** johnpur has joined #openstack-dev | 21:09 | |
| gyee | dolphm, what's the issue? | 21:09 |
| dolphm | gyee: keystone.common.controller.protected doesn't work with v3 tokens | 21:09 |
| lifeless | devananda: there is? | 21:10 |
| dolphm | gyee: i'd like to see the v3 tests actually using v3 auth instead of bypassing it with 'ADMIN' | 21:10 |
| devananda | lifeless: https://bugs.launchpad.net/nova/+bug/1129519 | 21:10 |
| uvirtbot | Launchpad bug 1129519 in nova "ComputeManager running_deleted_instances doesn't support instance uuids" [Undecided,New] | 21:10 |
| gyee | controller.protected works with that authenticate call | 21:11 |
| lifeless | devananda: ah. Is that the only method that needs us to cache the instance name ? | 21:11 |
| *** matiu_ has joined #openstack-dev | 21:11 | |
| *** matiu_ has quit IRC | 21:11 | |
| *** matiu_ has joined #openstack-dev | 21:11 | |
| lifeless | devananda: and did you look at just doing a try:except: within it ? | 21:11 |
| lifeless | in the same way as _get_instances_on_driver ? | 21:12 |
| devananda | lifeless: based on a chat with vishy, it's generally a good thing if baremetal acts like other hypervisors and supports the (legacy) list_instances() method, which is supposed to return names, not uuids | 21:12 |
| devananda | lifeless: naturally, baremetal also needs to support list_instance_uuids(). but the solution is not to change the list_instances() interface on all the other drivers.... | 21:12 |
| ayoung | gyee, OK, my review now official depends on yours | 21:13 |
| lifeless | devananda: I thought the general idea with legacy things was to move all the generic code to avoid it | 21:13 |
| *** johnpur has left #openstack-dev | 21:13 | |
| lifeless | devananda: so that it becomes ... legacy | 21:13 |
| gyee | ayoung, which one? | 21:13 |
| lifeless | devananda: this sounds like we're not doing that? | 21:13 |
| dolphm | gyee: i'm looking at changing the keystone_admin role in default_fixtures to just be 'admin', which matches policy.json | 21:13 |
| devananda | lifeless: instead of trying to wedge in a patch to nova at the last minute, i'd rather add support for a common (if old) interface to baremetal that it was sorely lacking | 21:13 |
| dolphm | gyee: then user_foo is admin for free | 21:14 |
| lifeless | devananda: I don't quite follow the last minute thing | 21:14 |
| lifeless | devananda: is there some special time pressure here? | 21:14 |
| dolphm | gyee: but of course the produced tokens won't work with @protected | 21:14 |
| devananda | lifeless: uh, yes... today is the cutoff ... | 21:14 |
| *** matiu_ has quit IRC | 21:14 | |
| lifeless | devananda: for what? | 21:14 |
| gyee | dolphm, that's still not going to work | 21:14 |
| lifeless | devananda: I mean, I know about g3. | 21:15 |
| gyee | user_foo don't not have admin role by default | 21:15 |
| *** timello_ has joined #openstack-dev | 21:15 | |
| gyee | does not | 21:15 |
| lifeless | devananda: but - baremetal is not going to be prod ready in G | 21:15 |
| dolphm | gyee: a lot of the setUp methods grant keystone_admin to user_foo | 21:15 |
| lifeless | devananda: so, I'm worried about caching the name only because I don't understand the semantics | 21:16 |
| lifeless | devananda: because other drivers are so varied in their implementations. | 21:16 |
| *** yidclare has quit IRC | 21:16 | |
| *** donaldh has joined #openstack-dev | 21:16 | |
| lifeless | devananda: at the moment we depend on the manager layer's mapping of uuid->name , which means we dont' need to care, with one exception | 21:16 |
| devananda | lifeless: fwiw, we already have folks trying to use it. no, it's not prod ready at scale, but it _should_ be usable in small environments, and this is one of the bugs which prevents that right now | 21:16 |
| lifeless | and that is dangling instances | 21:17 |
| gyee | dolphm, I don't we'll need to change all our tests to grant admin role | 21:17 |
| gyee | seems like a separate review to me | 21:17 |
| dolphm | gyee: no you don't -- 1 min | 21:18 |
| gyee | s/don't/think | 21:18 |
| lifeless | so it seems to me that the conservative - less risky thing - is to fix the one code path that remains that calls list_instances, rather than adding an implementation that might not really match. | 21:18 |
| *** kmartin has quit IRC | 21:18 | |
| dolphm | gyee: the v3 tokens coming out of hte token backend don't look anything like v2 tokens | 21:18 |
| dolphm | gyee: you created a container called token_data? | 21:19 |
| *** unix has quit IRC | 21:19 | |
| *** nerd has joined #openstack-dev | 21:19 | |
| *** Mr_T has joined #openstack-dev | 21:19 | |
| gyee | dolphm, should be the same as in the spec | 21:19 |
| dolphm | gyee: i'm talking about the backend, ignoring the API layer | 21:19 |
| cp16net | does there exist a way to migrate/resize an instance to a specific host with admin priv? | 21:19 |
| cp16net | unless i am missing something this doesnt seem possible | 21:20 |
| lifeless | devananda: I've removed my minus 1 | 21:20 |
| lifeless | devananda: but I still think its a mistake; just don't want to have a long conversation, other opinions are still valid :) | 21:20 |
| *** yidclare has joined #openstack-dev | 21:20 | |
| lifeless | devananda: so the decision is up to you. | 21:20 |
| cp16net | i see the create allows you to do scheduler_hints to force_hosts=['host1'] | 21:20 |
| gyee | dolphm, where do you see the token_data container? | 21:21 |
| dolphm | gyee: this is what i'm getting out of token_api.get_token() http://paste.openstack.org/raw/31975/ | 21:21 |
| lifeless | devananda: I don't see anything incorrect in the details of the change, its all conceptual. | 21:21 |
| dolphm | gyee: there was another expires in there which was actually a datetime object, and not json serializable, so i pulled that out so i could use json.dumps() | 21:22 |
| gyee | dolphm, yeah, that's for storing the token | 21:23 |
| dolphm | ayoung: #openstack-meeting | 21:23 |
| gyee | we don't need to store bits and pieces | 21:23 |
| dolphm | gyee: what you store in the backend for v2 should be identical for what you store in the backend for v3 | 21:23 |
| gyee | dolphm no need to | 21:24 |
| gyee | v2 and v3 token data are not interchangeable | 21:24 |
| dolphm | gyee: YES, you need to -- otherwise you run into the exact failure i'm seeing | 21:24 |
| dolphm | gyee: the data is NOT compatible | 21:24 |
| gyee | they are not supposed to | 21:24 |
| *** aeperezt has quit IRC | 21:24 | |
| devananda | lifeless: thanks :) from an idealistic POV, I agree that it would be great if everything used UUIDs, but looking at other drivers, I'd also like to have parity with them. Having a non-functional list_instances() just doesn't make sense, even if nothing calls it right now. | 21:24 |
| dolphm | gyee: so a token generated on v3 is unusable on v2 and vice versa? | 21:25 |
| gyee | no | 21:25 |
| gyee | can't validate a v2 token with v3 auth api | 21:26 |
| dolphm | gyee: ... why not | 21:26 |
| dolphm | heckj: ^^ | 21:26 |
| gyee | v3 token data format is competely different | 21:26 |
| gyee | domains, auth methods, etc | 21:26 |
| dolphm | gyee: it should still validate, they both have id's | 21:26 |
| dolphm | gyee: v2 tokens have a domain as well, it's just default' | 21:27 |
| gyee | if we have use cases that mixes v2 and v3 APIs we have a serious problem | 21:27 |
| *** aeperezt has joined #openstack-dev | 21:28 | |
| openstackgerrit | A change was merged to openstack/cinder: XenAPINFS: fix capacity reporting https://review.openstack.org/22187 | 21:28 |
| dolphm | gyee: we're supporting a mixed deployment for the next two releases | 21:28 |
| openstackgerrit | A change was merged to openstack/cinder: Skip tests if cinder is not installed https://review.openstack.org/22195 | 21:29 |
| *** kmartin has joined #openstack-dev | 21:29 | |
| gyee | dolphm, that's going to suck | 21:29 |
| heckj | gyee: yes, it does | 21:29 |
| gyee | can you imagine validating a v3 token with v2 api? | 21:29 |
| dolphm | gyee: uh huh, but that's sort of our promise of backwards compatibility | 21:29 |
| dolphm | gyee: yes i can | 21:29 |
| heckj | gyee: even if the V3 auth can't handle a V2 token, we need to ahve the V2 API avialable to do so | 21:29 |
| gyee | v2 clients are going to choke on the v3 token data | 21:29 |
| heckj | so the back-end data *must* support both at the same time | 21:29 |
| dolphm | gyee: v2 clients should be able to receive a v3 token **ID** and work with it | 21:30 |
| *** adjohn has quit IRC | 21:30 | |
| gyee | dolphm, how does v2 client handle v3 token data returned from token validation? | 21:30 |
| heckj | There's not going to be a hard cut/swtich to this - we need to handle it gracefully | 21:30 |
| dolphm | gyee: keystone needs to return v2 token data for a v3 token ID | 21:31 |
| gyee | the format is completely different | 21:31 |
| dolphm | gyee: the format is a superficial issue at the controller level, it's not something that should have leaked to the driver | 21:31 |
| *** pabelanger has quit IRC | 21:31 | |
| gyee | dolphm, what if domain is involved? | 21:31 |
| dolphm | gyee: that's where default domain comes in | 21:31 |
| gyee | catalog is completely different as well | 21:32 |
| dolphm | gyee: v2 controllers assume 'default' domain when calling drivers, so if the token is not applicable to the 'default' domain, it raises 401 | 21:32 |
| gyee | dolphm, then we are inconsistent in backward compatible | 21:32 |
| lifeless | devananda: I don't think we have to change other drivers though | 21:32 |
| dolphm | gyee: the v2 catalog controllers are currently transalating v2 requests to a v3-"aware driver | 21:32 |
| dolphm | "aware" | 21:32 |
| lifeless | devananda: as long as the core doesn't call list_instances without trying list_instance_uuids, why should we care? | 21:33 |
| dolphm | gyee: the whole goal of the 'default' domain was to make it an opt-in to start using v3 and multi-domain deployments | 21:33 |
| devananda | lifeless: cause an incomplete or broken API is just generally a bad thing? | 21:33 |
| lifeless | devananda: and list_instances is not broken ? | 21:34 |
| dolphm | gyee: you should be able to start using a v3 aware client intermixed with v2 aware clients as long as you don't go multi-domain (or you understand that v2 clients are limited to default domain resources) | 21:34 |
| lifeless | devananda: by comparison, look at the legacy/new nwinfo stuff | 21:34 |
| lifeless | devananda: the legacy is something to eliminate wherever possible; I see this the same way. | 21:35 |
| *** voliveirajr has quit IRC | 21:35 | |
| lifeless | devananda: its a defective API whose use is a mistake, and implementations are only needed to support us while we migrate the calling code. | 21:35 |
| devananda | sdague: got a moment to re+2 21855? I changed the comment in it yesterday... | 21:35 |
| gyee | dolphm, so we are saying v3 client will take to one instance of keystone to get a v3 token then turn around and use it to take to a service which is running a v2 instance of keystone? | 21:35 |
| devananda | lifeless: except that several other drivers don't implement list_instance_uuids at all, afaict | 21:36 |
| dolphm | gyee: all of this to say that the data in/out of the token driver should be identical regardless of the API/controller it's being utilized from | 21:36 |
| devananda | lifeless: so it's not really a legacy api | 21:36 |
| gyee | s/take/talk/ | 21:36 |
| dolphm | gyee: sure | 21:36 |
| devananda | s/it/list_instances/ | 21:36 |
| openstackgerrit | A change was merged to openstack/oslo-config: Add AUTHORS and ChangeLog to .gitignore. https://review.openstack.org/22346 | 21:36 |
| dolphm | gyee: although i assume that by 'v2 instance of keystone' you mean 'v2-aware version of auth_token' | 21:37 |
| *** pabelanger has joined #openstack-dev | 21:37 | |
| dolphm | gyee: (we're not mixing folsom and grizzly installs of keystone in the same deployment) | 21:37 |
| gyee | dolphm, only way we can get into this scenario is that we have v2 and v3 instances in the same env | 21:37 |
| *** rkukura has quit IRC | 21:38 | |
| dolphm | gyee: right now, today, at this moment, auth_token is hardcoded on v2 and openstackclient is being written against v3 ... so we will have tokens produced on v3 being validated against v2 for all of grizzly | 21:38 |
| sdague | lifeless: hey, so syncing up about where the matcher stuff stands | 21:38 |
| dolphm | gyee: (considering a few steps away from v3-aware auth_token at the moment, even if it lands soon, i doubt it'll make it into grizzly docs) | 21:39 |
| *** johnpur has joined #openstack-dev | 21:39 | |
| *** adjohn has joined #openstack-dev | 21:40 | |
| lifeless | sdague: danpb has a patch just needs +2's. | 21:40 |
| lifeless | sdague: see backscroll | 21:40 |
| openstackgerrit | A change was merged to openstack/oslo-config: Refactor _cparser_get_with_deprecated() https://review.openstack.org/22074 | 21:40 |
| gyee | dolphm, that fine, so I presume domain-scoped is not going to work as well | 21:40 |
| openstackgerrit | A change was merged to openstack/oslo-config: Allow MultiConfigParser get from mutliple sections https://review.openstack.org/22075 | 21:40 |
| lifeless | devananda: network stuff is the same | 21:40 |
| *** stevebaker has joined #openstack-dev | 21:40 | |
| dolphm | gyee: right, a domain scoped token has no authz on v2 | 21:40 |
| *** eglynn_ has joined #openstack-dev | 21:41 | |
| lifeless | devananda: some drivers do the old, some do the new, as the code base is mid-migration | 21:41 |
| lifeless | devananda: but that doesn't make the old code non-legacy. | 21:41 |
| dolphm | gyee: i'd be fine if it returned an unscoped token on v2? | 21:41 |
| lifeless | devananda: I don't see whats different here. | 21:41 |
| dolphm | validated* as an unscoped token | 21:41 |
| sdague | lifeless: ok, so that fixes that Matcher, but it still means it's trivially easy to sneak broken tests past testr | 21:41 |
| gyee | if we are going to store the same data as v2, we really don't have much of a choice | 21:42 |
| gyee | it will be treated as unscoped | 21:42 |
| *** eglynn has quit IRC | 21:42 | |
| *** cp16net is now known as cp16net|away | 21:45 | |
| *** rnirmal has quit IRC | 21:47 | |
| lifeless | sdague: otp, one minute | 21:47 |
| openstackgerrit | A change was merged to openstack/glance: Create connection for each qpid notification. https://review.openstack.org/22206 | 21:49 |
| openstackgerrit | A change was merged to openstack/cinder: Update Storwize/SVC driver for Grizzly. https://review.openstack.org/21457 | 21:50 |
| openstackgerrit | A change was merged to openstack/quantum: Support Port Binding Extension in BigSwitch plugin https://review.openstack.org/22239 | 21:50 |
| *** dolphm has quit IRC | 21:50 | |
| *** gray--_ has joined #openstack-dev | 21:51 | |
| *** martine has quit IRC | 21:53 | |
| *** esp has left #openstack-dev | 21:54 | |
| *** zyluo has quit IRC | 21:55 | |
| lifeless | sdague: hi | 21:57 |
| lifeless | sdague: yes, the existing subunit stream can be corrupted; thats why - http://rbtcollins.wordpress.com/ | 21:57 |
| lifeless | sdague: new subunit protocol with much better handling of test runners that go boom | 21:58 |
| lifeless | sdague: or tests that hang | 21:58 |
| *** pcm_ has quit IRC | 21:58 | |
| *** markvoelker has quit IRC | 21:58 | |
| *** esp has joined #openstack-dev | 21:58 | |
| lifeless | sdague: I'll get a fix for the ConcurrentTestSuite in place in testtools shortly as well, we clearly don't want to fail to record failures :) | 21:59 |
| *** cp16net|away is now known as cp16net | 22:00 | |
| *** kagan has joined #openstack-dev | 22:00 | |
| sdague | lifeless: ok cool :) | 22:00 |
| *** danpb has quit IRC | 22:01 | |
| lifeless | sdague: this is a particularly egregious foot-gun | 22:01 |
| lifeless | sdague: because we have a result object that is streaming potentially GB's of stuff out but its all one protocol unit | 22:02 |
| ttx | vishy: around ? | 22:02 |
| *** hugokuo has joined #openstack-dev | 22:02 | |
| gray--_ | hi chaps, i've run into this issue on ubuntu 12.10, https://bugs.launchpad.net/nova/+bug/993663, with this ver: 2012.2.1+stable-20121212-a99a802e-0ubuntu1.1 | 22:02 |
| ttx | clarkb, jeblair: around ? | 22:02 |
| uvirtbot | Launchpad bug 993663 in nova "[SRU] dns_domains table mysql charset is 'latin1'. Should be 'utf8'" [Medium,Fix released] | 22:02 |
| clarkb | ttx: yes | 22:02 |
| jeblair | ttx: o/ | 22:02 |
| lifeless | sdague: I'm inclined to paper over it and focus on the new codebase which is nearly at the point we could start consuming it [though the python API wouldn't be stable until at least after pycon when I hope to corner voidspace on it :)] | 22:03 |
| ttx | clarkb, jeblair: so it was suggested to turn off some tests in order to have the gate process all the stuff | 22:03 |
| *** gary_th has joined #openstack-dev | 22:03 | |
| *** zing has quit IRC | 22:03 | |
| sdague | lifeless: well as long as we don't run into this situation again | 22:03 |
| ttx | in time for a feature freeze pushed back on e day, ending tomorrow EOD | 22:03 |
| clarkb | ttx: jeblair in particular the tempest tests because they take an hour and a half often | 22:04 |
| ttx | I'm not a big fan of it, but I guess that risk is limited if we use the check results as a pre-approval test | 22:04 |
| jeblair | ttx: nobody does that | 22:04 |
| lifeless | sdague: paper over -- not ignore :). | 22:04 |
| gray--_ | was this expected to have been backported into ubuntu's nova-common | 22:04 |
| gray--_ | ? | 22:04 |
| *** mlavalle has joined #openstack-dev | 22:05 | |
| jeblair | ttx: i'm not a fan of it either. | 22:05 |
| lifeless | sdague: I'll put an appropriate error thingy into the except clause, and we'll say no more about it | 22:05 |
| sdague | lifeless: great, appreciate it | 22:05 |
| lifeless | sdague: vs fixing the generator to handle details blowing up... | 22:05 |
| gray--_ | seems a pretty fundamental issue on a vanilla ubuntu install | 22:05 |
| jeblair | ttx: are people ignoring flakey tests? | 22:05 |
| clarkb | ttx: jeblair neither am I. a couple problems with it are check tests can be ignored and check tests don't check in the order thigns are merged | 22:05 |
| sdague | once that's in, and we're past queue full, I'll bump all the projects | 22:05 |
| soren | I haven't followed this closely... Do we really have so many patches landing right now that 1½ to merge one is a big problem? | 22:05 |
| sdague | jeblair: the issue is raw volume | 22:05 |
| ttx | soren: yes | 22:06 |
| jeblair | sdague: i'm not sure i agree with that | 22:06 |
| clarkb | soren: http://status.openstack.org/zuul/ gives you the current queues | 22:06 |
| soren | 1½ hours, I mean. | 22:06 |
| *** garyTh has quit IRC | 22:06 | |
| jeblair | sdague: volume only matters if patches fail tests | 22:06 |
| sdague | jeblair: well pypi breaking all to hell blew everything out of the water this morning | 22:06 |
| jeblair | let's be clear what we're saying: "Testing OpenStack doesn't work very well and takes so long that we don't want to do it any more." | 22:06 |
| *** eglynn_ has quit IRC | 22:06 | |
| jeblair | sdague: indeed, not arguing there, and we were poised to take drastic action to fix that, but it is not necessary because pypi is working now | 22:07 |
| sdague | jeblair: I agree with you, I'm trying to bring balanced perspective | 22:07 |
| soren | Oh, there isnt a per-project queue, but rather a global one? | 22:07 |
| ttx | unfortunately the ones pushing for that are not around | 22:07 |
| ttx | soren: yes | 22:07 |
| sdague | soren: yes, it's a global queue | 22:07 |
| jeblair | soren: OpenStack is one project. | 22:07 |
| clarkb | soren: yes because the integration tests check that one project doesn't break another | 22:07 |
| ttx | vishy: ping | 22:07 |
| sdague | ttx: where's vishy ? | 22:07 |
| vishy | hi | 22:08 |
| ttx | hah, here he comes | 22:08 |
| ttx | prepare the rotten tomatoes | 22:08 |
| sdague | jeblair: is there an easy way to see what keeps failing in the gate? | 22:08 |
| * vishy puts on his helmet | 22:08 | |
| soren | clarkb: Yeah, makes sense, just hadn't really thought about that. | 22:08 |
| sdague | like a list of all the rejected jobs? maybe there is a pattern | 22:08 |
| *** stevebaker has quit IRC | 22:08 | |
| clarkb | sdague: we can get a list. | 22:09 |
| *** zaneb has joined #openstack-dev | 22:09 | |
| jeblair | sdague: i imagine the tempest jobs cause most of the failures, because they are the ones common to multiple projects | 22:09 |
| vishy | so looking at the current nova queue of things that have to go in for ffe | 22:09 |
| vishy | we may be able to squeak by | 22:09 |
| jeblair | sdague: though the actual _causes_ are more interesting, and ideally listed here: http://status.openstack.org/rechecks/ | 22:09 |
| ttx | vishy: we could also limit the number of OTHER reviews being approved | 22:09 |
| jeblair | sdague: (but that relies on people) | 22:09 |
| vishy | but time to merge is getting pretty frustrating :) | 22:09 |
| sdague | jeblair: right, and it's not time boxed | 22:09 |
| ttx | i.e. stop approving random patches, concentrate on prio stuff | 22:09 |
| sdague | what I really want to know is how many queue resets did we have today | 22:09 |
| sdague | and what caused them | 22:10 |
| ttx | there aren't that many changes tat are feature-related in queue | 22:10 |
| buzztroll_ | Are any glance folks around? I am wondering if the location of images which are stored outside of glance should be returned in the metadata | 22:10 |
| jeblair | sdague: what _job_ caused them or what _bug_ caused them? | 22:10 |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Support RPC envelopes in impl_zmq https://review.openstack.org/21790 | 22:10 |
| buzztroll_ | they currently are not, and i am wondering if this is a bug or a feature | 22:10 |
| ttx | vishy: I hate the idea of disabling tests just because everyone wants to push their change at the very end | 22:10 |
| sdague | ttx: yeh, well telling everyone to put down the +A would solve it from that side as well | 22:10 |
| *** dontalton has joined #openstack-dev | 22:11 | |
| ttx | we could ask all -core to stop approving non blueprint-stuff for a day or two | 22:11 |
| *** winston-d has quit IRC | 22:11 | |
| ttx | only approve targeted blueprints code | 22:11 |
| clarkb | off the top of my head there are a couple problems that have been slowing down the gate lately | 22:11 |
| ttx | so that the queue prioritizes correctly | 22:11 |
| ttx | vishy: would that work for you ? | 22:12 |
| ttx | vishy: + 24/7 support from Infra team watching the queue when it chokes ? | 22:12 |
| clarkb | https://bugs.launchpad.net/glance/+bug/1108985 happens fairly frequently. The quantum tests fail a bit due to instances not becoming active. volume tests fail for the same reason. and when Jenkins gets overloaded it loses connectivity to its slaves | 22:12 |
| uvirtbot | Launchpad bug 1108985 in glance "Spurious ECONNREFUSED errors during Glance unit tests" [Undecided,New] | 22:12 |
| vishy | we could try | 22:12 |
| russellb | fwiw it's not just this week, people have been annoyed for the last few weeks with how long stuff takes to get merged | 22:12 |
| dansmith | +1 | 22:12 |
| rainya | i know it was a point of annoyance for the last two status projects at least | 22:12 |
| ttx | russellb: the problem is only in the number of false negatives in tests | 22:12 |
| ttx | without false negatives, you can consider merging is instant | 22:13 |
| jeblair | russellb: is anyone annoyed enough to fix the bugs? | 22:13 |
| jeblair | https://bugs.launchpad.net/glance/+bug/1108985 | 22:13 |
| ttx | (i.e. merging n+1 is not longer than merging n) | 22:13 |
| *** winston-d has joined #openstack-dev | 22:13 | |
| jeblair | ^ that one clarkb pointed out is huge and according to lp, no one is working on it. | 22:13 |
| uvirtbot | jeblair: Error: "that" is not a valid command. | 22:13 |
| ttx | OK, so here is my proposal... | 22:14 |
| clarkb | it affects devstack and glance unittests. each time it fails in the gate everything behind it must restart | 22:14 |
| ttx | * Ask core to only approve targeted blueprints for a day | 22:14 |
| *** vuntz has quit IRC | 22:14 | |
| ttx | * Sped extra cycles fixing test false negatives causing rechecks | 22:14 |
| ttx | Spend | 22:14 |
| *** gabrielhurley has quit IRC | 22:14 | |
| jeblair | fwiw, clarkb and mordred are trying very hard to get hpcloud to increase our cloud account quota there | 22:15 |
| ttx | * Infra does not sleep and fix new issues if any come along | 22:15 |
| *** zyluo has joined #openstack-dev | 22:15 | |
| *** vuntz has joined #openstack-dev | 22:15 | |
| clarkb | also, merge conflicts stop everything | 22:15 |
| ttx | vishy: would that work for you ? Should we try ? | 22:15 |
| jeblair | hpcloud is running tempest in 40 minutes or less there, as opposed to 90 minutes or more on rackspace. :( | 22:15 |
| clarkb | if you want to be careful you can try and check that the code you are approving can actually merge | 22:15 |
| sdague | clarkb: that's non solvable with more resources | 22:16 |
| vishy | lets try it | 22:16 |
| vishy | if we need to reevaluate tomorrow we can | 22:16 |
| ttx | I can announce the FF extension and the temporary approval rule | 22:16 |
| clarkb | sdague: in theory you can click the rebase button in gerrit and don't approve if it cannot rebase | 22:16 |
| ttx | and reevaluate early tomorrow | 22:16 |
| lifeless | jeblair: do we need more hp instances ? | 22:16 |
| jeblair | ttx, vishy: ^ clarkb's 'rebase' suggestion is pretty good | 22:16 |
| sdague | clarkb: yeh, though the collisions that we most often get wouldn't be solved by that | 22:16 |
| jeblair | lifeless: we've needed them for a week but getting them hasn't been going well. | 22:16 |
| lifeless | jeblair: oh, I see you are on that in backscroll... | 22:16 |
| sdague | api samples is the fun one there | 22:17 |
| clarkb | lifeless: yeah I think I derped the request on friday | 22:17 |
| lifeless | clarkb: doh! | 22:17 |
| vishy | interesting idea on rebase | 22:17 |
| jeblair | we _are_ seeing a lot of merge conflicts | 22:17 |
| *** stevebaker has joined #openstack-dev | 22:17 | |
| sdague | jeblair: any chance the rebase check could happen first automatically? | 22:18 |
| *** lcheng has quit IRC | 22:18 | |
| *** Mr_T has left #openstack-dev | 22:18 | |
| sdague | I keep trying to think of ways we can throw things out early | 22:18 |
| jeblair | sdague: clarkb has been working on speeding that up, but it's...complicated | 22:19 |
| sdague | yeh, understood | 22:19 |
| jeblair | sdague: enough so that even if we had a patch in hand for it, i don't think we'd merge it this week due to potential disruption. | 22:19 |
| sdague | yep, fair | 22:19 |
| jeblair | sdague: (since we already reverted a first attempt at that because of unanticipated consequences) | 22:20 |
| jeblair | it's possible, though conterintuitive, that we might speed things up by reducing rackspace's contribution to the node pool | 22:20 |
| *** salv-orlando_ has joined #openstack-dev | 22:21 | |
| jeblair | we don't have enough hpcloud machines to meet demand, but since they turn over faster, it may be a net win | 22:21 |
| *** njoy1__ has joined #openstack-dev | 22:21 | |
| openstackgerrit | A change was merged to openstack/quantum: Update latest OSLO files https://review.openstack.org/22302 | 22:21 |
| sdague | yeh it's too bad you can't place them first in the queue | 22:21 |
| sdague | or what about dedicating all the hp nodes to the gate instead of the check? | 22:22 |
| *** anteaya has left #openstack-dev | 22:22 | |
| *** woodspa has quit IRC | 22:22 | |
| clarkb | thats a possibility. we would need to update a bunch of jenkins jobs but that mostly automagic | 22:22 |
| jeblair | ttx: i'm surveying the recheck bugs, and they mostly seem to be getting ignored | 22:22 |
| lifeless | jeblair: ttx: recheck bugs should be critical, no ? | 22:23 |
| clarkb | jeblair: what do you think about changing the label on hpcloud instances and devoting them to the gate? | 22:23 |
| jeblair | clarkb: we'd have to make new jobs | 22:23 |
| clarkb | hmm that would require a zuul restart too | 22:23 |
| clarkb | yeah | 22:23 |
| sdague | zuul will be empty in the morning | 22:23 |
| sdague | there were only 5 jobs in zuul at 7:30am EST today | 22:24 |
| *** brucer has joined #openstack-dev | 22:24 | |
| jeblair | clarkb: actually, not new jobs, i think... | 22:24 |
| *** salv-orlando has quit IRC | 22:24 | |
| *** salv-orlando_ is now known as salv-orlando | 22:24 | |
| jeblair | clarkb: i believe the node is already a parameter, so we could update the parameter function to add that info | 22:24 |
| *** njoy_ has quit IRC | 22:25 | |
| jeblair | on the plus side, we'd have finer control over which resources we give to which pipeline... | 22:25 |
| jeblair | on the other hand, we would be partitioning our resources | 22:25 |
| clarkb | it opens us to catastrophic failure like pypi this mornign (eg hpcloud goes down) | 22:26 |
| jeblair | clarkb: (but that's still a zuul config change, so a bit disruptive) | 22:26 |
| sdague | jeblair: I think right now that's ok, the real constraint is on the gate side | 22:26 |
| ttx | lifeless: yes, as well as security bugs. Use common sense :) | 22:26 |
| sdague | clarkb: sure, but if the gate queue crashes, no one will care if you have to restart zuul to flip it to the other cloud | 22:26 |
| jeblair | sdague: hopefully we'll get new test resources soon too, enough so that that matters less. | 22:26 |
| ttx | You have reached your quota for directly contacting other Launchpad users. You can try again in 23 hours. | 22:26 |
| ttx | pfff | 22:26 |
| clarkb | ttx: nice | 22:27 |
| lifeless | ttx: I don't have triage acls on most projects; I was curious about policy :) | 22:27 |
| sdague | ttx: man, even with your karma? | 22:27 |
| lifeless | its not karma based | 22:27 |
| lifeless | very simple code | 22:27 |
| *** kmansel has joined #openstack-dev | 22:27 | |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Add function for listing native threads to eventlet backdoor https://review.openstack.org/22197 | 22:27 |
| ttx | ok, so I need someone to warn quantum/cinder reviewerzs | 22:28 |
| *** rods has quit IRC | 22:28 | |
| *** stevebaker has quit IRC | 22:29 | |
| *** AlanClark__ has quit IRC | 22:29 | |
| ttx | danwent, jgriffith: please relay http://lists.openstack.org/pipermail/openstack-dev/2013-February/005841.html to your core reviewers | 22:30 |
| *** AlanClark__ has joined #openstack-dev | 22:30 | |
| jgriffith | ttx: check | 22:30 |
| *** nati_ueno has quit IRC | 22:30 | |
| ttx | or someone can run https://launchpad.net/~quantum-core/+contactuser :) | 22:30 |
| danwent | ttx: but you're ok with quantum not doing the extension, correct? | 22:31 |
| danwent | its just that we should also hold off on bug fixes? | 22:31 |
| ttx | danwent: not doing the extension ? | 22:32 |
| ttx | danwent: ideally bugfixes would get in Thursday, in time for G3 | 22:32 |
| *** Yada has left #openstack-dev | 22:33 | |
| danwent | ttx: yes, my plan was to still hold the requirement that quantum blueprints are done by tonight. | 22:33 |
| *** jbr_1 has quit IRC | 22:33 | |
| ttx | danwent: oh, you mean not extending Feature Freeze for you | 22:33 |
| danwent | ttx: yes | 22:33 |
| danwent | no | 22:33 |
| danwent | or maybe yes, depending on what you mean :) | 22:33 |
| danwent | as in, if something is not in by tonight for quantum, it needs an FFE | 22:34 |
| danwent | rather than allowing people to push any blueprint in past the deadline tonight | 22:34 |
| jeblair | sdague, ttx: i've changed the pool config so we shouldn't get any more precise nodes from rax. we'll see if hpcloud only will speed things up. | 22:34 |
| ttx | danwent: your call | 22:34 |
| danwent | ttx: k | 22:34 |
| ttx | danwent: if you want, I can cut milestone-proposed for quantum g3 tomorrow morning | 22:34 |
| *** nati_ueno has joined #openstack-dev | 22:34 | |
| ttx | effectively declaring feature freeze for quantum | 22:35 |
| danwent | ttx: no need to break the routine, i will just enforce it with reviews. | 22:35 |
| *** pabelanger has quit IRC | 22:35 | |
| sdague | jeblair: can you pull them out of the check queue? | 22:36 |
| sdague | so they will only be on the gate | 22:36 |
| sdague | and checks can't eat them up | 22:36 |
| *** epim has quit IRC | 22:37 | |
| *** nati_ueno_2 has joined #openstack-dev | 22:37 | |
| *** stevebaker has joined #openstack-dev | 22:38 | |
| SpamapS | lifeless: how hard would you estimate it might be to have the ec2 metadata service implement SSL? | 22:39 |
| gyee | dolphm, heckj, for PKI tokens, token data is baked into the token itself | 22:40 |
| jeblair | sdague: that's what clarkb and i were discussing; it would require a few changes and a bit of time to do, but it's possible. i'm thinking it is probably worth it, but realistically, i'm not sure it could be in place before tomorrow. | 22:40 |
| gyee | so we can't really expect data compatibility in that case | 22:41 |
| sdague | jeblair: ok | 22:41 |
| *** bswartz has quit IRC | 22:42 | |
| *** gray--_ is now known as gray-- | 22:42 | |
| henrynash | dolphm, gyee: in regards to the protected wrapper in the controllers not understanding the token format, I have that change in my query filter patch: https://review.openstack.org/#/c/22223/ | 22:42 |
| *** jpich has quit IRC | 22:43 | |
| lifeless | SpamapS: if the API server is running SSL it probably already does | 22:44 |
| lifeless | SpamapS: try having that setup, an iptables rule for port 443 on 169.254.169.254 as well as port 80 | 22:44 |
| *** shardy is now known as shardy_afk | 22:44 | |
| lifeless | SpamapS: and see what happens | 22:44 |
| gyee | henrynash, for v3, roles are no longer stashed under the user container | 22:45 |
| gyee | the are part of root container | 22:45 |
| SpamapS | lifeless: thats what I think too. So perhaps just making it possible to have multiple metadata ports would be enough. Hm | 22:45 |
| henrynash | gyee: err, you mean in the token? | 22:46 |
| gyee | henrynash, correct, it will be token['roles'] instead of token['user']['roles'] | 22:47 |
| henrynash | gyee: have we updated the spec on that? Doesn't seem to be what it says right now... | 22:47 |
| *** buzztroll_ has quit IRC | 22:47 | |
| *** afazekas has quit IRC | 22:48 | |
| *** jergerber has quit IRC | 22:48 | |
| gyee | henrynash, yes, https://github.com/openstack/identity-api/blob/master/openstack-identity-api/src/markdown/identity-api-v3.md | 22:49 |
| *** dolphm has joined #openstack-dev | 22:50 | |
| henrynash | gyee: damn….somehow I didn't sync that change…ok, I'll fix that in my patch | 22:50 |
| *** dolphm has quit IRC | 22:50 | |
| henrynash | gyee: thx | 22:50 |
| *** gray-- has quit IRC | 22:50 | |
| *** dolphm has joined #openstack-dev | 22:50 | |
| zykes- | ttx: ping; when does the summit talk registration close ? | 22:51 |
| SpamapS | lifeless: darn, don't think I can fake it with stunnel because it changes the source IP. | 22:51 |
| dolphm | gyee: ayoung: heckj: henrynash: did i miss anything in the past hour? | 22:53 |
| zykes- | or does anyone else know ? | 22:53 |
| henrynash | dolphm: only just back on myself | 22:53 |
| dolphm | zykes-: i heard it was supposed to have already closed, but was briefly extended -- not even sure if it's still open | 22:53 |
| *** Ryan_Lane has quit IRC | 22:53 | |
| gyee | dolphm, I just submitted another patch | 22:54 |
| *** Ryan_Lane has joined #openstack-dev | 22:54 | |
| gyee | and added a test for v2-v3 intermix | 22:54 |
| *** johnpur has quit IRC | 22:55 | |
| henrynash | gyee: do we need to clone oslo now for your patch? | 22:55 |
| gyee | henrynash, no | 22:55 |
| henrynash | gyee: where does olso.config come from? | 22:56 |
| *** Ryan_Lane has quit IRC | 22:56 | |
| gyee | henryhash, see tools/pip-requires | 22:56 |
| gyee | you need to install it over http | 22:56 |
| dolphm | gyee: i meant to send this to you earlier http://paste.openstack.org/raw/31993/ | 22:56 |
| zykes- | dolphm: seems it just opened :p | 22:56 |
| zykes- | 6 days ago that is | 22:56 |
| zykes- | http://summit.openstack.org/ | 22:56 |
| dolphm | gyee: i think this is based on your patch 12 | 22:57 |
| dolphm | zykes-: ooh, i was thinking of the other half of the conference i suppose | 22:57 |
| *** pixelbeat has joined #openstack-dev | 22:58 | |
| gyee | dolphm, thanks, lemme check | 22:58 |
| zykes- | dolphm: which one you mean then ? | 22:58 |
| russellb | heh, it appears an improperly dated post will be on top of planet.openstack.org for a while ... | 22:59 |
| dolphm | zykes-: the talks that developers don't care about | 22:59 |
| zykes- | dolphm: I was wanting to do dev stuff yes :p | 22:59 |
| zykes- | dolphm: as in Billing or network stuff maybe | 22:59 |
| dolphm | zykes-: cool, yeah design summit stuff definitely just opened :) | 23:00 |
| *** openstackgerrit has quit IRC | 23:00 | |
| Kiall | zykes-, you going to make it? | 23:00 |
| dolphm | zykes-: "The call for speakers for the OpenStack Summit Spring 2013 is now open. The deadline for speaker submissions is February 15, 2013. We will again be making the submissions public for community vote starting February 18, with the goal of locking the "Conference" portion of the agenda by March 11." | 23:00 |
| zykes- | Kiall: uncertain yet | 23:00 |
| dolphm | zykes-: from http://www.openstack.org/summit/portland-2013/ | 23:00 |
| *** esp has quit IRC | 23:00 | |
| *** soren has quit IRC | 23:00 | |
| *** esp has joined #openstack-dev | 23:00 | |
| *** soren has joined #openstack-dev | 23:01 | |
| *** Ryan_Lane has joined #openstack-dev | 23:01 | |
| dolphm | henrynash: i had no trouble running pip install -r tools/pip-requires | 23:01 |
| henrynash | dolphm: just about to do the smae | 23:01 |
| henrynash | same | 23:01 |
| *** john5223 has quit IRC | 23:03 | |
| *** lglenden_ has quit IRC | 23:04 | |
| *** tiamar has joined #openstack-dev | 23:06 | |
| *** metral_ has quit IRC | 23:06 | |
| *** dosaboy has quit IRC | 23:07 | |
| dolphm | gyee: can you explain why PKI tokens generated on v3 don't work on v2? | 23:07 |
| dolphm | gyee: you should also have a test in the opposite direction -- a token generated on v2 should work on v3 | 23:07 |
| *** Ryan_Lane has quit IRC | 23:09 | |
| *** digitalsanctum has quit IRC | 23:09 | |
| *** digitalsanctum has joined #openstack-dev | 23:10 | |
| *** mtreinish has quit IRC | 23:11 | |
| gyee | dolphm, is v2 to v3 a valid use case? | 23:11 |
| gyee | I thought we only need to worry about v3 to v2 | 23:11 |
| *** zaneb has quit IRC | 23:12 | |
| ayoung | dolphm, gyee BTW I just submitted http://summit.openstack.org/cfp/edit/22 | 23:12 |
| *** Ryan_Lane has joined #openstack-dev | 23:12 | |
| *** flaper87 has quit IRC | 23:12 | |
| dolphm | gyee: compatibility goes both ways | 23:12 |
| ayoung | gyee, do you need help with the is_admin thing? I had to deal with it in test_v3_trusts | 23:13 |
| dolphm | gyee: so, technically we didn't document PKI support in folsom, I'm wondering if we can get away with saying that PKI tokens *only* work on v3 -- but we need to update client support correspondingly... | 23:13 |
| gyee | ayoung, I need help translating a v3 PKI token into a v2 PKI token | 23:13 |
| dolphm | heckj: ^ thoughts? | 23:13 |
| gyee | dolphm, +2 :) | 23:13 |
| ayoung | gyee, ok, once you run the CMS verify call, you have the raw data. | 23:14 |
| dolphm | is keystoneclient the only codebase that can introspect a PKI token? | 23:14 |
| ayoung | from there you should be able to tell the format. Is there a version field or something? | 23:14 |
| ayoung | dolphm, conditional yes | 23:15 |
| dolphm | ayoung: conditional? | 23:15 |
| ayoung | dolphm, we used to, but not any more/ | 23:16 |
| ayoung | so, for Grizzly, yeah, only keystone client | 23:16 |
| *** kbringard has quit IRC | 23:16 | |
| *** bodepd has joined #openstack-dev | 23:17 | |
| *** dachary has quit IRC | 23:17 | |
| dolphm | ayoung: gyee: so, i'm not super familiar with the client-side validation logic... but if a v3 PKI token is given a v2 client that attempts to perform offline validation, if the token is in an unrecognizable format, will it fall back to online validation? | 23:17 |
| ayoung | yes it will | 23:18 |
| *** dachary has joined #openstack-dev | 23:18 | |
| dolphm | ayoung: gyee: or does it only fall back if it's obviously not encrypted | 23:18 |
| ayoung | dolphm, it would have failed by then.,... not sure | 23:18 |
| *** mjfork has quit IRC | 23:18 | |
| *** Ryan_Lane has quit IRC | 23:19 | |
| ayoung | verified = self.verify_signed_token(user_token) | 23:20 |
| ayoung | 23:20 | |
| ayoung | data = json.loads(verified) | 23:20 |
| *** buzztroll_ has joined #openstack-dev | 23:20 | |
| ayoung | so all the CMS is going to do is confirm the signature | 23:20 |
| *** Ryan_Lane has joined #openstack-dev | 23:20 | |
| ayoung | then the JSON data gets loaded. | 23:20 |
| *** imsplitbit has quit IRC | 23:20 | |
| ayoung | https://github.com/openstack/python-keystoneclient/blob/master/keystoneclient/middleware/auth_token.py#L882 | 23:21 |
| ayoung | then line https://github.com/openstack/python-keystoneclient/blob/master/keystoneclient/middleware/auth_token.py#L575 | 23:23 |
| gyee | right, v3 and v2 token data are incompatible | 23:23 |
| gyee | so you can have an environment with mixture of v2 and v3 middleware | 23:23 |
| *** maoy has quit IRC | 23:23 | |
| ayoung | gyee, but from the auth_token middleware, you get the token data as JSON | 23:23 |
| ayoung | can you test a token to seeif it is v2 or v3? | 23:24 |
| *** jergerber has joined #openstack-dev | 23:24 | |
| bodepd | I get a few tempest errors about quotas when I run --smoke | 23:24 |
| bodepd | is there some initial config required to set up quatas properly for the smoke tests to work? | 23:24 |
| ayoung | gyee, dolphm OK, I have to disappear here for a few hours. | 23:24 |
| ayoung | I'm headed to my folks place, and I'll get back on line to check in once I am there. | 23:24 |
| *** diogogmt has quit IRC | 23:25 | |
| henrynash | dolphm: minor question, given the other subjects being discussed, but installing the oslo.config should end up creating such in keystone/Keystone, yes? | 23:25 |
| *** agentle_ has quit IRC | 23:26 | |
| *** sacharya has quit IRC | 23:26 | |
| dolphm | ayoung: gyee: if PKI tokens can't be used in an intermixed environment, i'm not sure that PKI should be enabled by default | 23:27 |
| *** digitalsanctum has quit IRC | 23:27 | |
| dolphm | henrynash: no, it should install into your python site-packages | 23:28 |
| henrynash | dolphm: hmm, ok, it complete ok, but tests still can't find it... | 23:29 |
| *** ayoung has quit IRC | 23:29 | |
| heckj | back | 23:29 |
| dolphm | henrynash: when you use run_tests.sh, do you let it build a venv? | 23:29 |
| dolphm | heckj: o/ | 23:29 |
| heckj | yeah - reading back | 23:30 |
| *** buzztroll_ has quit IRC | 23:30 | |
| henrynash | dolphm: only when I force it to…. | 23:30 |
| henrynash | dolphm: I activate the .venv and did the pp install in there…. | 23:30 |
| heckj | dolphm: henrynash, gyee : markmc has a patch going into install_venv_… for that ick, should be up shortly | 23:30 |
| henrynash | hoping to avoid rebuilding the whole venv | 23:30 |
| dolphm | heckj: so, a PKI-aware v2 client attempting to validate a v3 PKI token will decrypt it and find a structure it doesn't understand | 23:30 |
| gyee | dolphm, for PKI tokens, services running v2 middleware won't be able to parse a v3 PKI token | 23:30 |
| heckj | (at least if you use the ./run-tests.sh) | 23:30 |
| dolphm | heckj: that same client might not be smart enough to fall back to online validation | 23:31 |
| *** buzztroll_ has joined #openstack-dev | 23:31 | |
| dolphm | henrynash: i don't really know how the install_venv stuff works, i manage my own virtualenv manually and always use ./run_tests.sh --no-virtual-env (-N for short) | 23:32 |
| henrynash | dolphm: ok, I'll fix it...thx | 23:33 |
| *** soody has joined #openstack-dev | 23:33 | |
| heckj | dolphm: henrynash: FYI: https://review.openstack.org/#/c/22368/ | 23:33 |
| SpamapS | lifeless: so, interesting .. the way nova wants to run, you can only have metadata be SSL, or not SSL | 23:33 |
| SpamapS | lifeless: so to run both, we'd have to run two copies of nova-api (or fix that... ;) | 23:33 |
| *** cloudchimp has quit IRC | 23:34 | |
| *** almaisan-away is now known as al-maisan | 23:34 | |
| heckj | dolphm: going to run all the way over to play devil's advocate here for a sec... | 23:34 |
| dolphm | heckj: lol k | 23:34 |
| heckj | the token format should be set by API and deployment, and while we need to support both at once, do we absolutely need to support V2 API interfaces accepting and understanding V3 tokens (and vice versa)? I think we need the client to be able to support both, but perhaps each on their own API endpoint | 23:35 |
| dolphm | henrynash: missed your message earlier about activating the venv and pip installing... that should have worked :-/ | 23:35 |
| *** al-maisan is now known as almaisan-away | 23:35 | |
| *** openstackgerrit has joined #openstack-dev | 23:36 | |
| henrynash | dolphm: indeed | 23:36 |
| heckj | I think the only issue with this stance would be in multiple keystone's coordinating together, but we don't have federation support at all in V2, and only the first parts of it in V3 | 23:36 |
| dolphm | heckj: so, with UUID's, there's no issue -- everyone understands them | 23:36 |
| dolphm | heckj: so it's not an API issue -- it's an issue with our implementation | 23:36 |
| heckj | so it's getting a PKI token at the client, and not knowing which version you asked from? I missing it... | 23:37 |
| *** yamahata has quit IRC | 23:37 | |
| dolphm | heckj: end user with v3 API client trying to use a service protected by v2 auth_token, for example | 23:37 |
| zykes- | problems with grizzly features ? | 23:37 |
| *** AlanClark__ has quit IRC | 23:38 | |
| dolphm | zykes-: yes, they aren't going | 23:38 |
| *** AlanClark__ has joined #openstack-dev | 23:38 | |
| *** kagan has quit IRC | 23:38 | |
| heckj | dolphm: how about I rephrase that | 23:38 |
| zykes- | dolphm: domain etc? | 23:38 |
| dolphm | zykes-: v3 auth, full stop | 23:38 |
| zykes- | dolphm: that is .... sad... | 23:39 |
| heckj | dolphm: end-user with a client that knows both V2 and V3 API, requests a V2 token - the client had the token, can now pass it in | 23:39 |
| dolphm | zykes-: we're all working on it :P | 23:39 |
| heckj | Isn't it safe to assume that the deployment will be understanding of V2 tokens since the keystone that was deployed with it generated such? | 23:40 |
| SpamapS | hrm, has anyone used/tested the enabled_ssl_apis option in nova.conf ? | 23:40 |
| dolphm | heckj: absolutely | 23:40 |
| dolphm | heckj: there's not an issue within the scope of v2 alone | 23:40 |
| SpamapS | I think I have it working, openssl s_client sets up SSL properly, but then the connection is immediately shut down | 23:40 |
| dolphm | heckj: but then you're saying that you can't use v3 for auth? | 23:40 |
| heckj | dolphm: I guess I'm asserting that when you deploy an environment, all the services have to effectively agree (at least at this point) on either using V2 or V3 . | 23:41 |
| heckj | God, I'm taking gyee's side now | 23:41 |
| dolphm | heckj: =) | 23:41 |
| heckj | alright, let me try and noodle my way out of this noose | 23:41 |
| dolphm | heckj: so then, it's sort of pointless to say that grizzly support v3 auth, because nothing else will consume it (as of this moment) | 23:42 |
| heckj | dolphm: yeah - that's pretty true | 23:42 |
| dolphm | heckj: there's another issue with the current review worth discussing, and the issue is limited in scope to keystone | 23:42 |
| heckj | Not surprising for when we have V3 auth available though - just at feature freeze would make me expect nothing would be supporting it yet | 23:42 |
| *** jkyle has joined #openstack-dev | 23:43 | |
| heckj | dolphm: I think the key for supporting both will be the auth_token middleware (once again) - knowing how to identify and pull apart and use both tokens | 23:43 |
| *** metral has joined #openstack-dev | 23:43 | |
| *** zyluo has quit IRC | 23:44 | |
| *** roampune1 has quit IRC | 23:44 | |
| *** zyluo has joined #openstack-dev | 23:44 | |
| heckj | I think the two API's and their respective formats can stand alone, as long as they can be easily identified from each other and the services accept either | 23:44 |
| jkyle | I ran across this in the keystone wiki | 23:44 |
| jkyle | "support proxying external services and AuthN/AuthZ mechanisms such as oAuth, SAML and openID in future versions." | 23:44 |
| *** mjfork has joined #openstack-dev | 23:44 | |
| heckj | dolphm: what's the other topic? | 23:44 |
| heckj | jkyle: what's the question? | 23:45 |
| jkyle | does that mean it does not currently support using an external authentication system for user accounts? like oauth | 23:45 |
| jkyle | that's exactly what we'd like to do. | 23:45 |
| heckj | jkyle: right now, that means creating a custom back-end for the identity component in keystone to proxy | 23:45 |
| dolphm | heckj: before we move on -- i sort of took that position in my client support of v3 auth (still WIP) | 23:45 |
| gyee | heckj, dolphm, so do we still need to worry about v2-v3 interchangeability for PKI tokens at this point? | 23:45 |
| dolphm | heckj: Access has a sort of factory method that takes a v2 or v3 token and returns an object that knows how to read values from it | 23:46 |
| jkyle | heckj: ok, thanks. I thought it was something like that. I'm digging through the documentation, does a "writing an identity backend" example project exist? | 23:46 |
| heckj | jkyle: check out keystone, and look at keystone/identity/core.py and the backends/ examples of them, and it should be pretty clear what the internal API is that you need to support | 23:46 |
| dolphm | heckj: https://review.openstack.org/#/c/21942/ | 23:46 |
| jkyle | excellent, thanks! | 23:46 |
| *** buzztroll_ has quit IRC | 23:46 | |
| heckj | dolphm: sounds like a generally good plan for attacking just that | 23:47 |
| heckj | reading through the PR now | 23:47 |
| heckj | er, review | 23:47 |
| heckj | (sorry, too used to github for my day-to-day) | 23:47 |
| *** fifieldt has joined #openstack-dev | 23:47 | |
| heckj | jkyle: an example of someone who's done a keystone backend entirely outside of keystone is https://github.com/icgood/keystone-redis, and there's another (although not public) that I know at - OSU's SSO system | 23:49 |
| *** edmund has quit IRC | 23:49 | |
| *** jimfehlig has quit IRC | 23:50 | |
| *** digitalsanctum has joined #openstack-dev | 23:50 | |
| heckj | dolphm: I think with https://review.openstack.org/#/c/21942/ (which I just +2'd), we can work to support dual token formats in auth_token middleware. | 23:51 |
| bodepd | figured out the differences that are causing that test to fail. I have these extra default quotas that are not expected by the test: {'gigabytes': 1000, 'volumes': 10} | 23:52 |
| heckj | It can't see us getting that fully out the door before feature freeze, but the clients aren't under the same timelines... | 23:52 |
| gyee | dolphm, the patch you give me doesn't work | 23:52 |
| gyee | dolphm, http://paste.openstack.org/raw/31993/ | 23:52 |
| dolphm | gyee: i know, it reveals lots of test failures that are hidden by bypassing v3 auth in the v3 tests | 23:53 |
| gyee | user_foo doesn't exist | 23:53 |
| openstackgerrit | A change was merged to openstack/keystone: Update the Keystone policy engine to the latest openstack common https://review.openstack.org/22109 | 23:54 |
| *** pabelanger has joined #openstack-dev | 23:55 | |
| dolphm | gyee: exactly, none of the tests are actually using get_scoped_token | 23:55 |
| dolphm | gyee: they're all bypassing it with is_admin=True | 23:55 |
| heckj | urg | 23:55 |
| heckj | *facepalm* | 23:55 |
| jkyle | heckj: looks straight forward enough, good stuff. | 23:55 |
| gyee | dolphm, still getting a 403, even with admin role | 23:56 |
| heckj | jkyle: if you get stuck, holler - although we're a touch distracted right now with the grizzly release feature freeze and the usual chaos leading into it | 23:56 |
| dolphm | gyee: policy.json needs to be updated to include the new controller-level methods you introduced | 23:56 |
| heckj | jkyle: we also have keystone-general meetings weekly - you're welcome to pop in and pester: https://wiki.openstack.org/wiki/Meetings/KeystoneMeeting | 23:57 |
| *** bswartz has joined #openstack-dev | 23:57 | |
| jkyle | yeah, I'm subscribed to the calendars. I'll keep an eye on it next time it dings me :) | 23:57 |
| *** rpedde is now known as rpedde_away | 23:58 | |
| gyee | dolphm, I am not testing my controller, just test_v3_identity | 23:58 |
| *** alszar has quit IRC | 23:58 | |
| gyee | those tests all failed with a 403, even with user_foo, which has the admin role | 23:58 |
| dolphm | test_v3_identity wants to authenticate with get_scoped_token() and run it's tests with a real user account, and it's failing | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!