Hi all,
I'm testing a Pike overcloud deployment with an external Ceph installation. This is a small setup (1 controller and 1 compute) with libvirt/KVM VMs.
The undercloud has been installed from the RDO pike repositories, overcloud images have been built using `openstack overcloud image build` on the undercloud.
The deployment is done with the following script:
*************
TD=/usr/share/openstack-tripleo-heat-templates/
TE=$TD/environments
openstack overcloud deploy --templates \
-e $TE/disable-telemetry.yaml \
-e $TE/network-isolation.yaml \
-e $TE/net-single-nic-with-vlans.yaml \
-e $TE/storage/external-ceph.yaml \
-e local-config.yaml
*************
The ceph related parameters in local-config.yaml are:
parameter_defaults:
NovaEnableRbdBackend: true
CinderEnableRbdBackend: true
CinderBackupBackend: ceph
GlanceBackend: rbd
GnocchiBackend: rbd
NovaRbdPoolName: vms
CinderRbdPoolName: volumes
CinderBackupRbdPoolName: backups
GlanceRbdPoolName: images
GnocchiRbdPoolName: metrics
CephClientUserName: openstack
CephClusterFSID: '6ad0a5f7-ee40-4c04-bcdb-6de3d049b82f'
CephClientKey: 'AQC9SVparYGINhAA5xTsOldCDA8p3JnNnIyHFw=='
CephExternalMonHost: '172.18.0.9'
I ran the deploy_steps_playbook.yaml ansible playbook to get more information, the error log is attached to this bug report (failure on both nodes).
If I create `/etc/ceph/` on the nodes and redeploy, the deployment succeeds where it failed before.
I also explicitly defined the following resource to have a functional overcloud at the end of the deployment (not sure if this is related):
OS::TripleO::Services::CephClient: /usr/share/openstack-tripleo-heat-templates/puppet/services/ceph-client.yaml
I'm not sure if this is a bug or a configuration problem. Let me know if you need more information.
There is no bug, my setup was not correct: my overcloud image didn't have the ceph packages installed.
Maybe it would be a good idea to add a note about this requirement to https:/ /docs.openstack .org/tripleo- docs/latest/ install/ advanced_ deployment/ ceph_external. html ?