Tell me how to increase the virtual disk with ZFS?

[Paul Kraus]

On May 11, 2013, at 8:59 AM, "Vladislav Prodan" <universite at ukr.net> wrote:

> Add another virtual disk and do a RAID0 - not an option. It is not clear how to distribute the data from the old virtual disk to the new virtual disk.

When you add an additional "disk" to a zpool (to create a STRIPE), the ZFS code automatically stripes new writes across all top level vdevs (drinks in this case). You will see a performance penalty until the data distribution evens out. One way to force that (if you do NOT have snapshots) is to just copy everything. The new copy will be striped across all top level vdevs.

The other option would be to add an additional disk that is as large as you want to the VM, attach it to the zpool as a mirror. The mirror vdev will only be as large as the original device, but once the mirror completes resilvering, you can remove the old device and grow the remaining device to full size (it may do that anyway based on the setting of the auto expand property of the zpool. The default under 9.1 is NOT to autoexpand:

root at FreeBSD2:/root # zpool get autoexpand rootpool
NAME      PROPERTY    VALUE   SOURCE
rootpool  autoexpand  off     default
root at FreeBSD2:/root # 

--
Paul Kraus
Deputy Technical Director, LoneStarCon 3
Sound Coordinator, Schenectady Light Opera Company