As an example, consider a queuing network which is a high-level
model for a 3-CPU multiprocessing computer on a single bus and two
ports to memory. The queue represents the serializing of memory data
entering the bus.
S1 and S2 are identical token sources each supplying
three tokens at a time with inter-arrival times uniformly distributed
between 1 and 100 time units. J1 receives all generated tokens
and provides a single stream of tokens to the single queue. Tokens are
removed from the queue whenever any of the three servers, S3,
S4, and S5, is free. J2 is a join which then
receives tokens from the servers and destroys them, thus freeing up
the small amount of memory used by a token.
