[Gllug] Forcing CPU and Mem limits on a process

[Daniel P. Berrange]

On Mon, Feb 14, 2005 at 05:00:15PM -0000, David Abbishaw wrote:
> The idea is really to be able to virtually partition a machine so that we
> can guarantee a customer an amount of CPU time and RAM, thus if someone else
> application starts hogging a CPU /RAM it doesn't affect other customers.  Ee
> can do this in Windows 2003 (with wsrm) and with VMWare ESX server although
> ESX has a huge overhead per VM (in disk, cpu, ram and support and licences)
> 
> Oh well sounds like a feature that's not quite there yet in Linux, I'm sure
> one day it will emerge.

There's a project 'Class-based Kernel Resource Management' on SF.net[1] that
is working on providing these kind of capabilities to the Linux kernel. The
project site has been too active recently, but I see renewed interest in it
on various mailing lists, it'll be a very important capability to have with
virtualization technologies such as Xen.

> Does this mean that UML suffers from one instance being able to use all
> resources on a machine?

Yeah, UML instances can definite suffers from CPU & I/O starvation if
one VM is particuarly busy. Memory isn't so much of aproblem, since (on
the UML version run by Bytemark hosting at least), they allocate a fixed
amount of real physical RAM to each UML instance & lock it into memory
in the host system. (UML kernel performs /terribly/ if what it considers
to be physical RAM is in fact swapped out by the host kernel!)

Regards,
Dan.

[1] http://ckrm.sourceforge.net
-- 
|=-            GPG key: http://www.berrange.com/~dan/gpgkey.txt       -=|
|=-       Perl modules: http://search.cpan.org/~danberr/              -=|
|=-           Projects: http://freshmeat.net/~danielpb/               -=|
|=-   berrange at redhat.com  -  Daniel Berrange  -  dan at berrange.com    -=|
-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug