The IMS Spool API as a data manager

The IMS Spool API controls the input, execution, and output of jobs on the z/OS® system. Because the IMS Spool API is at the JOB or time sharing (TSO) user boundary, JES creates special considerations for IMS environments where each print data set is a logical message and is not part of or associated with a job boundary.

A data manager associated with IMS provides the DL⁄I application programming interface, which provides temporary buffering of data with the sync point process. The sync point process commits the changes for a unit of work if the unit of work terminates normally, or backs out the changes if the unit of work terminates abnormally. The IMS Spool API does not use the sync point process, but it can back out the effect of changes made by an abending unit of work.

The IMS Spool API does not recover data as does the IMS Transaction Manager or the IMS Database Manager. Because the IMS Spool API does not log changes to the spool data set, data on the spool can be lost in a restart if the spool device finds data errors.