[Sois-tcoa] Time Distribution - again

[gregory.menke at gsfc.nasa.gov]

Abhijit Sengupta writes:
 >     * Data structure:
 >         * The messaging system on requesting side puts a time stamp of its 
 > own clock on the message and also the messaging system on recipient side 
 > puts a time stamp of its own clock when message is received.
 >         * Similar time stamp while sending responses
 >         * These provide some estimate of propagation delay and delay 
 > because of processing overhead on each side. Such estimate might be useful 
 > for time critical cases.
 > 
 
I think this is a fantastic idea.  When we were developing the radhard
ethernet stuff for GPM, there were a number of occasions where it
would have been really handy to accumulate timestamps up and down the
protocol stacks on the sending and receiving ends.  

We kludged our way towards it by setting up a udp packet into which
the application layer would insert a timestamp.  The driver would
detect the packet at the datalink layer, obtain its own timestamp, and
with the difference we could evaluate the timing overhead of the
entire stack in both directions.  It was quite useful, even in that
clumsy form.

Greg