Thank Gallati, which means we donot need have to have shared rbd for live migration ?
On Tue, Dec 23, 2014 at 11:47 AM, René Gallati <ceph@xxxxxxxxxxx> wrote:
Hello,
On 23.12.2014 09:12, Zeeshan Ali Shah wrote:
Has any one tried running instances over ceph i.e using ceph as backend
for vm storage . How would you get instant migration in that case since
every compute host will have it's own RBD . other option is to have a
big rbd pool on head node and share it with NFS to have shared file system
When you use shared block devices, then the compute nodes don't need to have a shared file system in openstack. All their (runtime)information comes from either config files or the controller node/apis.
They mount RBDs and they contact each other in the case of a live migration so there is a sort of handover protocol, at least when you use libvirt+qemu as hypervisor. How this is set up is described in
http://ceph.com/docs/next/rbd/rbd-openstack/
Kind regards
René
Zeeshan Ali Shah
System Administrator - PDC HPC
PhD researcher (IT security)
Kungliga Tekniska Hogskolan
+46 8 790 9115
_______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
