Parallel Libtrace HOWTO: Scenario

For this HOWTO, we are going to concentrate on a particular packet analysis scenario so that everything that we demonstrate is going to have some relevant context.

The initial scenario is reasonably simple: we want to calculate the average payload size for UDP packets in a packet trace and we are going to use the libtrace parallel API to write a program to do it. Once we've done that, we are going to modify our program to report our UDP packet statistics every 10 seconds so that we can run the program against a live source in real-time.

First up, let's think about our packet processing threads.