On 11/26/2010 10:43 PM, vincent paul wrote:
Hi All,
Should user's zero-window time be considered as user's tier
processing or network delay?
Well, it depends on what you consider network delay. It's not a
transmission delay, but it can be due to inefficient TCP stack. So
is that considered OS tier processing or network processing time? I
would consider it a protocol processing time.
*But* you can also make an argument that the application is too slow
in sucking up the datagrams, in which case it's an application/tier
processing time.
So it depends on your definition. BTW, your use of "tier
processing" and "network delay" makes me think you're using Opnet's
product and questioning its output? Do I have that right?
hsb
|
