SVC dump processing stores data in dump data sets that the system allocates automatically, as needed, or that you pre-allocate manually. IBM® recommends the use of automatically allocated dump data sets whenever possible. Only the space required for the dump being written is allocated. The dump is written using a system-determined block size, so write time is reduced. SMS extended attributes, such as compression and striping, can be assigned to further reduce the amount of space required and the time to write.
IBM recommends using extended format sequential data sets as dump data sets for SVC Dumps. For the reasons why, see Choosing SVC dump data sets.
Use pre-allocated dump data sets only as a back up, in case the system is not able to automatically allocate a data set. Otherwise, the dump can become truncated, making error diagnosis difficult.