LinuxCon+ContainerCon Europe 2016

The Linux kernel offers a lot of great debugging tools. The most powerful one is the tracing infrastructure. It's not just one single debugging method, it offers a lot of different methods for collecting and analyzing data within the Operating System. This presentation will give an introduction to the usage of the tracing infrastructure and the different methods for collecting data. This includes: Event tracing, using the tracers, collecting events during the boot process, dynamically adding events using kprobes and injecting events into your application with uprobes. It will also cover the tools which can be used for tracing, like trace-cmd and kernelshark. Furthermore the usage of the perf CTF converting function (which can be used to analyze traces with Tracecompass and Eclipse) will be explained.