Revision history [back]

click to hide/show revision 1
initial version

Update: When i use the --block_migrate option the live migration succeeds (which is nice but unexpected). This seems inconsistent with the operation that apparently occurred because there was nothing to migrate (with a the cinder volume residing on a NetApp NFS share being served to the hypervisor on two compute nodes). What actually got moved was the vm and it's memory, processes, etc. Is this the recommended way to perform this operation? It seems like the first command without the --block_migrate option should have performed a pre-check to see if shared cinder storage for the vm was there and if so perform a standard migration.

Update: When i use the --block_migrate option the live migration succeeds (which is nice but unexpected). This seems inconsistent with the operation that apparently occurred because there was nothing to migrate (with a the cinder volume residing on a NetApp NFS share being served to the hypervisor on two compute nodes). What actually got moved was the vm and it's memory, processes, etc. Is this the recommended way to perform this operation? It seems like the first command without the --block_migrate option should have performed a pre-check to see if shared cinder storage for the vm was there and if so perform a standard migration.