Activity log for bug #1671422

Date Who What changed Old value New value Message
2017-03-09 10:29:57 James Page bug added bug
2017-03-09 10:30:10 James Page bug task added charm-cinder-ceph
2017-03-09 10:30:16 James Page charm-cinder-ceph: importance Undecided Critical
2017-03-09 10:30:17 James Page charm-nova-compute: importance Undecided Critical
2017-03-09 10:30:20 James Page charm-cinder-ceph: status New Triaged
2017-03-09 10:30:22 James Page charm-nova-compute: status New Triaged
2017-03-09 10:30:30 James Page summary nova/cinder/ceph rbd integration broken on Ocata charms: nova/cinder/ceph rbd integration broken on Ocata
2017-03-09 10:35:22 James Page description https://github.com/openstack/nova/commit/b89efa3ef611a1932df0c2d6e6f30315b5111a57 introduced a change in Ocata where any data provided by cinder for rbd block devices is preferred over any local libvirt sectional configuration for rbd (which was used in preference in the past). As a result, its not possible to attach ceph block devices in instances in Ocata; the secret_uuid configuration is not populated in the cinder configuration file, and in any case the username on the compute units won't match the username for ceph being used on the cinder units (as compute and cinder units get different keys created) so I don't think the key created on the compute units will actually work with the username provided from cinder. I'm not 100% convinced this is a great change in behaviour; the cinder and nova keys have much the same permissions for correct operation (rwx on images, volumes and vms groups) however it does mean that the nova-compute units have to have the same keys as the cinder units. A key disclosure/compromise on a cinder unit would require revoke and re-issue across a large number of units (as compute units at likely to be 100-1000's whereas the number of cinder units will be minimal. https://github.com/openstack/nova/commit/b89efa3ef611a1932df0c2d6e6f30315b5111a57 introduced a change in Ocata where any data provided by cinder for rbd block devices is preferred over any local libvirt sectional configuration for rbd (which was used in preference in the past). As a result, its not possible to attach ceph block devices in instances in Ocata; the secret_uuid configuration is not populated in the cinder configuration file, and in any case the username on the compute units won't match the username for ceph being used on the cinder units (as compute and cinder units get different keys created) so I don't think the key created on the compute units will actually work with the username provided from cinder. I'm not 100% convinced this is a great change in behaviour; the cinder and nova keys have much the same permissions for correct operation (rwx on images, volumes and vms groups) however it does mean that the nova-compute units have to have the same keys as the cinder units. A key disclosure/compromise on a cinder unit would require revoke and re-issue across a large number of units (as compute units are likely to be 100-1000's whereas the number of cinder units will be minimal.
2017-03-09 10:48:18 James Page bug task added nova
2017-03-09 11:21:22 James Page description https://github.com/openstack/nova/commit/b89efa3ef611a1932df0c2d6e6f30315b5111a57 introduced a change in Ocata where any data provided by cinder for rbd block devices is preferred over any local libvirt sectional configuration for rbd (which was used in preference in the past). As a result, its not possible to attach ceph block devices in instances in Ocata; the secret_uuid configuration is not populated in the cinder configuration file, and in any case the username on the compute units won't match the username for ceph being used on the cinder units (as compute and cinder units get different keys created) so I don't think the key created on the compute units will actually work with the username provided from cinder. I'm not 100% convinced this is a great change in behaviour; the cinder and nova keys have much the same permissions for correct operation (rwx on images, volumes and vms groups) however it does mean that the nova-compute units have to have the same keys as the cinder units. A key disclosure/compromise on a cinder unit would require revoke and re-issue across a large number of units (as compute units are likely to be 100-1000's whereas the number of cinder units will be minimal. https://github.com/openstack/nova/commit/b89efa3ef611a1932df0c2d6e6f30315b5111a57 introduced a change in Ocata where any data provided by cinder for rbd block devices is preferred over any local libvirt sectional configuration for rbd (which was used in preference in the past). As a result, its not possible to attach ceph block devices in instances in a charm deployed Ocata; the secret_uuid configuration is not populated in the cinder configuration file, and in any case the username on the compute units won't match the username for ceph being used on the cinder units (as compute and cinder units get different keys created) so I don't think the key created on the compute units will actually work with the username provided from cinder. I'm not 100% convinced this is a great change in behaviour; the cinder and nova keys have much the same permissions for correct operation (rwx on images, volumes and vms groups) however it does mean that the nova-compute units have to have the same keys as the cinder units. A key disclosure/compromise on a cinder unit would require revoke and re-issue across a large number of units (as compute units are likely to be 100-1000's whereas the number of cinder units will be minimal.
2017-03-09 11:21:50 James Page charm-cinder-ceph: milestone 17.05
2017-03-09 11:21:53 James Page charm-nova-compute: milestone 17.05
2017-03-09 12:20:45 James Page charm-cinder-ceph: assignee James Page (james-page)
2017-03-09 12:20:47 James Page charm-nova-compute: assignee James Page (james-page)
2017-03-09 12:20:49 James Page charm-cinder-ceph: status Triaged In Progress
2017-03-09 12:20:55 James Page charm-nova-compute: status Triaged In Progress
2017-03-09 18:44:31 OpenStack Infra charm-nova-compute: status In Progress Fix Committed
2017-03-14 08:28:49 James Page bug task added charm-guide
2017-03-14 08:28:59 James Page charm-guide: status New In Progress
2017-03-14 08:29:02 James Page charm-guide: importance Undecided High
2017-03-14 08:29:05 James Page charm-guide: assignee James Page (james-page)
2017-03-14 08:29:07 James Page charm-guide: milestone 17.05
2017-03-14 09:20:11 OpenStack Infra charm-cinder-ceph: status In Progress Fix Committed
2017-03-15 09:42:06 OpenStack Infra charm-guide: status In Progress Fix Released
2017-03-15 09:51:34 James Page charm-cinder-ceph: status Fix Committed Fix Released
2017-03-15 09:51:37 James Page charm-nova-compute: status Fix Committed Fix Released
2017-03-22 18:54:33 Ryan Beisner bug added subscriber Ryan Beisner
2017-04-17 13:05:03 Sean Dague bug task deleted nova
2017-07-05 18:45:09 Ryan Beisner bug task added charm-cinder
2017-07-05 21:22:39 Darin Arrick bug added subscriber Darin Arrick
2017-08-17 13:32:30 Nobuto Murata bug added subscriber Nobuto Murata
2017-08-18 06:51:13 Dominique Poulain bug added subscriber Dominique Poulain
2017-09-21 15:44:24 James Page charm-cinder: status New Won't Fix
2018-12-20 17:10:04 Chris Sanders bug added subscriber Canonical Field Critical
2018-12-20 21:46:25 David Ames charm-cinder: status Won't Fix Confirmed
2018-12-20 21:46:28 David Ames charm-cinder: importance Undecided Critical
2018-12-20 21:46:30 David Ames charm-cinder: assignee David Ames (thedac)
2018-12-20 21:46:34 David Ames charm-cinder: milestone 19.04
2018-12-22 00:37:22 David Ames charm-cinder: assignee David Ames (thedac)
2018-12-22 21:33:07 Xav Paice bug added subscriber Canonical IS BootStack
2018-12-22 21:33:12 Xav Paice tags canonical-bootstack
2019-01-21 10:53:24 Liam Young charm-cinder: assignee Liam Young (gnuoy)
2019-01-24 19:28:06 Liam Young charm-cinder: status Confirmed Invalid
2019-01-25 07:42:47 Liam Young removed subscriber Canonical Field Critical