Specifying Spool Data Sets as Members of Spool Partitions

For each spool partition, you can specify the data sets that are to be members of that partition. You do this by specifying the name of the spool partition on the FORMAT or TRACK statement associated with each spool data set. A spool data set, however, can be a member of only one spool partition at any time.

If you do not name some spool data sets as members of a spool partition, JES3 makes them members of the default spool partition. If you have defined no spool partitions, JES3 makes all spool data sets members of the spool partition named JES3PART. JES3 defines and names the JES3PART spool partition.

Except for the default spool partition, it is not necessary for every spool partition to have spool data sets. For the default spool partition, you must specify at least one spool data set as a member of that partition, or you must let JES3 make a spool data set a member by default. During JES3 initialization, if the default spool partition has no spool data sets, JES3 terminates with an abend code of DM012. Any spool partition without a spool data set is called a “dummy partition”.

You can also specify which spool partition you want JES3 to use when it writes the initialization data that it needs to perform hot starts and local starts. Isolating the initialization data in this way causes JES3 to allocate the most accessible track groups to that data. To define a spool partition for initialization data, specify INIT=YES on the SPART statement that defines the spool partition you want to use.

For better performance, keep spool data sets within a partition about the same size.