This session will walk the audience through some common scenarios where latencies hit an embedded linux system causing loss of real-time performance that the embedded system expects. I will share experiences using Linux based latency tracing and deep system tracing tools and show how to use them. I will also share some methodologies such as instrumentation techniques and dynamic probing to detect problems. Lastly we will go through scheduler statistics which may indicate processes in a system that are suffering from higher latencies and how to use them.