On Wednesday, 01/16/2008 at 01:48 EST, "McBride, Catherine" 
<[EMAIL PROTECTED]> wrote:
> For a SOX audit I'd almost agree with you, as you bring up some valid 
points.  
> This was a PCI audit.  The key difference that we've found between SOX 
and PCI 
> is that for SOX you create policy statements to meet SOX guidelines and 
are 
> tested on how well you adhere to your own policies. For PCI you are 
tested 
> against the external PCI standards (as issued by the Payment Card 
Industry 
> Council).  A hipersocket would have more than met the standard of a 
private, 
> dedicated connection, had anyone been willing to listen.   But instead 
the guy 
> formulated a strong opinion and would not alter his position.   

I doesn't really matter if it is SOX or PCI.  The only difference is who 
establishes the policy.  If you can establish an audit point that can be 
used to demonstrate that you have a "private dedicated" connection, then 
your auditor is wrong.  Of course, the second you attach a 3rd LPAR (or 
another guest) to the HiperSocket, you no longer meet the criteria since 
you cannot establish access controls on a HiperSocket that allow LPARs 2 
and 3 to talk only with LPAR 1, not with each other.  It might be 
"private," but it sure is hard to call it "dedicated".

Alan Altmark
z/VM Development
IBM Endicott

Reply via email to