Networking Pearl Harbor in the Making

Mon Nov 7 16:04:02 UTC 2005

Seems everyone considering the options would be well advised to  
consider how availability/reliability is actually calculated and  
based on that exercise make a more educated decision as to whether  
this does yield improvements at a cost that can be absorbed.

Just because you have n different flavors doesn't mean availability  
goes up.  And you might find some surprises in how costs develop.   
This isn't just about equipment, it's the operational impact as well.

Unfortunately, short of a verifiable economic cost being associated  
with such a doomsday scenario, what a business case can carry is what  
will be deployed.  And regulation doesn't necessarily solve anything  
here either (as it isn't cost neutral).

You can always build more availability.  But can you afford to pay  
for it.  (IMHO, the DoD JSF effort is real world testament to what  
happens when the cost of an ideal becomes so high that a compromise  
must be reached to sustain the effort -- this very much has its  
analogy in networking as well).

Or those are my $.02 anyway,
Christian

On Nov 7, 2005, at 10:50 AM, Simon Waters wrote:

>
> On Monday 07 Nov 2005 3:42 pm, Hannigan, Martin wrote:
>>
>> It's an argument for vendor diversity.
>
> No it is an argument for code base diversity (or better software  
> engineering).
>
> Vendor diversity doesn't necessarily give you this, and you can get  
> this with
> one vendor.
>
> Vendor diversity might be a good idea, but for other reasons.