[vbox-dev] VirtualBox faster than actual hardware?

Robin Green greenrd at gmail.com
Thu Dec 10 13:54:55 GMT 2009


You don't need virtualisation to move the blocks around on the disk to get
better performance. This has already been implemented, e.g. there is some
Intel software for Windows that does this. However, such software wouldn't
work as expected for a virtual disk file, so it might still be useful to
reimplement the idea in VirtualBox.

-- 
Robin

2009/12/9 Joseph Miller <josephcmiller2 at gmail.com>

> Hello,
> I am new to this list and I am not sure if this idea has been addressed in
> the past, so please forgive me if it has.
>
> Most virtualizers end up with performance somewhere in the range of 98% to
> 99% of running on real hardware (AFAIK).  This is pretty good considering
> what they have to go through in order to run guest systems.  But it is well
> known that the perfomance bottleneck in most desktop systems is in the disk
> access area.  I think this is an area where VirtualBox could actually exceed
> the performance of actual hardware.
>
> Operating systems disk access are limited by several factors like the type
> of filesystem that is being used and the number of files (not just the size)
> that must be loaded at any given time.  However, it is common that groups of
> files are all loaded in a certain sequence or basically *at the same time*
> such as when the OS is booting or when a particular program is loading and
> all its data files and libraries must be loaded.  Since disk access is
> virtual with VirtualBox, the mapping of these disk sectors on the virtual HD
> vs the physical HD doesn't have to be the same.  Perhaps it would be
> worthwhile for VirtualBox to keep an index of which sectors are commonly
> loaded together and to either physically group these together on the
> physical HD (with mapping to their virtual location of course) or a simple
> disk caching mechanism.  For files that are changed constantly (data files)
> the former may be a better solution but may carry difficult technical
> hurdles.  The latter would be fine for most static data (such as libraries
> or OS booting files).  If this data can be loaded sequentially (if the host
> has a sufficiently non-fragmenting filesystem) performance can be
> *significantly* improved.  Maybe I could finally get that reiserfs
> performance on my XP guest :)
>
> Please let me know what you think.  I do not think that I have the
> technical ability to implement this on my own, I'm mostly a PHP hacker, but
> I would contribute in any way I can if this is reasonable.
>
> Thanks for your consideration.
>
> -Joseph
>
> (P.S. - I just *inherited* a dual-head display and am wondering if there is
> any development on using a physical graphics card directly in the guest
> system. I googled but couldn't find anything.)
>
> _______________________________________________
> vbox-dev mailing list
> vbox-dev at virtualbox.org
> http://vbox.innotek.de/mailman/listinfo/vbox-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.virtualbox.org/pipermail/vbox-dev/attachments/20091210/f1ed3a8f/attachment.html>


More information about the vbox-dev mailing list