If the Event Management daemon detects an error in the shared memory segment used by the daemon, or a resource monitor instance, it creates a dump file in the /var/ha/run/haem.syspar directory. This dump contains the first 4KB of the shared memory segment. The dump is called rzdump.RMrmname.rminst.time, where rmname.rminst is the resource monitor name and instance, and time is a timestamp.
The Event Management daemon (haemd) generates a full core dump in most of the error conditions listed in Event Management daemon errors.