We have a wide mix of hardware all running the same SL images and
configurations.  Most of our hardware is supplied by Dell, with many
different models, but we also have a few IBM's and the odd home-built
box here & there too. For desktops (we have hundreds running SL... 400
maybe? more? a lot anyway) differences with graphics cards are usually
the only thing that may change our "normal" build.  We use a combination
of kickstart over PXE with cfengine for config changes. 
  
HTH,
Deb  


On Wed, 2006-09-27 at 18:36 +0100, John Rowe wrote:
> Does anybody have experience of running a single image of SL on hardware
> which was not all bought at the same time? Ie, is it possible to buy
> machines from people like Dell so that you know that you can buy a
> machine in a years time that will be sufficiently identical that a
> pxeboot and an rsync to the known image will bring the machine up? 
> 
> I'm familiar with doing this for the classic beowulf cluster, it's the
> procurement over a period of time that I'm interested in. Does anybody
> actually manage this? Obviously we can do it with Macs and OS X, I would
> like to be able to do it with Scientific Linux too. I'm relatively
> neutral with respect to a vendor like Dell or a home-built solution like
> asus.
> 
> Thanks
> 
> John
-- 
d e b b i e   m a z u r e k                 [log in to unmask]
system admin . department of computer science . university of calgary