We use our MPI-aware extension of OTFX described in
Section 4 to demonstrate the resulting trace size reduction
and to examine the introduced overhead for the event tracing
library. In addition, we evaluate the usefulness of the
resulting traces for performance analysis.