[c-nsp] MTBF for Cisco products

[Peter Rathlev]

jay at west.net skrev man, 03 dec 2007:
> Ted Mittelstaedt wrote:
>> (assuming a standard bell curve applies to device failures, which  
>> it generally does)
>
> Actually, hardware failures generally follow a "bathtub curve", almost
> an inverse bell curve.  There is a certain amount of infant mortality
> where substandard parts will fail immediately or shortly after
> installation, bad connections and other mechanical and assembly issues
> are discovered or fail in transit, etc.  Then relatively few failures
> for a long time, and then a gradual ramping up of failures as things
> wear out, capacitors dry up, etc.

I think you both agree with each other. The "Bell curve" would  
describe the age of the devices and the "bathtub curve" would describe  
when they fail, so it's natural that one is the others inverse. :-)

Regards,
Peter Rathlev