Migrating VMs from pure paravirt hypervisor to hvm-capable hypervisor

I’m also changing all of my disk layouts from being block devices per each guest partition to one block device hypervisor side per each /dev/xvd* guest side. The block devices all live on a single storage server and are being served via iSCSI. One of the guests is making use of AoE, but I’m thinking I’m going to move it to iSCSI, too, so there’s one less dæmon I need to maintain.

It’s currently xen on squeeze. I want to experiment with kvm. I will miss xen-create-image, though.

Okay. Zelda wants to watch a movie, so I’m going to stop blathering now.



Leave a Reply