Re: [Nbd] NBD_CMD_RESIZE

[Alex Bligh <alex@...872...>]

On 18 May 2013, at 14:27, Goswin von Brederlow wrote:

> I think the inital idea was to send a resize ioctl to the block device
> on the client. The kernel would then send NBD_CMD_RESIZE to the server
> and the server would reply with the new size. I don't think that will
> ever work right with multiple clients for the same export. You would
> have to resize each individually.

The initial idea was that the client (kernel or otherwise) would send
NBD_CMD_RESIZE to the server (if the server supported it and
permitted it) and then the server would resize the volume to the
size requested by the client. That's simple enough when the client
is file backed. In things like qemu-nbd it's also useful as backing
stores such as ceph and qcow2 can be resized online.

I think multiple clients for the same export only work sanely with
readonly mounts (unless you are doing some crazy stuff like running
ocfs2 on nbd). I would presume resize would not work on a readonly
mount.

-- 
Alex Bligh