Re: [Xen-users] Why is xen allocating entire disk/file?

["Fajar A. Nugraha" <fajar@xxxxxxxxx>]

On Thu, Jan 29, 2009 at 12:20 PM, Charlie Reddington
<charlie.reddington@xxxxxxxxx> wrote:
> # df -h
> Filesystem            Size  Used Avail Use% Mounted on
> /dev/sda1             446G   43G  381G  10% /
> tmpfs                 3.9G     0  3.9G   0% /dev/shm
>
> Right here alone, has me kind of sketcy. Total - 446G - 43G should = 403G

No comment on that one :)

>
> But in var/lib/xen/images.....
>
> # du -sh *
> 9.8G    ns1.img
> 9.8G    ns2..img
> 119M    shared01.img
> 20G             www.img
>
> Doing the command as you say.....
>
> 9.8G -rwxrwxr-x 1 root root 9.8G Jan 28 15:18 ns1.img
> 9.8G -rwxrwxr-x 1 root root 9.8G Jan 28 15:18 ns2.img
> 134M -rwxrwxr-x 1 root root  98G Jan 28 15:18 shared01.img
>  20G -rwxrwxr-x 1 root root  20G Jan 28 15:18 www.img
>

So it doesn't use 100G right away, right?
Both "du -sh" and "ls -lash" (leftmost column) shows space actually
used. It can be different from the right column if it's created as
sparse file (in your case, "not to allocate the entire disk yet").

> I'm just overall kind of confused how this is working.  I should in theory
> be able to over allocate a number of vm's, so that when I get close to
> running out, I can just add more disk space as cash allows.

Yes, it should work like that. Note however that using sparse file has
the downside of :
- lower I/O performance initially (i.e. only when it has to allocate
disk space previously used)
- there might be a condition when you run out of space (e.g. when all
domU unexpectedly needs more space) which can lead to unpredictable
result.

Regards,

Fajar

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users