One of the interesting discussions going on right now across some of the teams around Linux and Power systems is around the ability to trace what's happening in the kernel and across the operating system.
The reasons are many... what's going on in the network stack? where are unexpected delays (latencies) happening around high speed transactions? where can we optimize the operating system better? is the behavior of the OS really doing what we think it's designed to do? etc etc etc
In that vein we plan to begin discussing some of those considerations here on the community, and encourage you to add your ideas, thoughts, and considerations. Our first topic can be discussed in this initial question posed on the message board. Check it out!