Well, I could just enrich the trace/event with sample data from CPU, RAM, Disk I/O, etc...