Records added to the end of a data set (that is, records with successively higher keys), can be added by the basic access method using WRITE KN, or by the queued access method using the PUT macro (resume load). In either case, records can be added to the prime data area.

When you use the WRITE KN macro, the record being added is placed in the prime data area only if there is room for it on the prime data track containing the record with the highest key currently in the data set. If there is not sufficient room on that track, the record is placed in the overflow area and linked to that prime track, even though additional prime data tracks originally allocated have not been filled.

When you use the PUT macro, records are added to the prime data area until the space originally allocated is filled. After this allocated prime area is filled, you can add records to the data set using WRITE KN, in which case they will be placed in the overflow area. Resume load is discussed in more detail under Creating an ISAM Data Set.

You can continue to add fixed-length records in this manner until the original space allocated for prime data is exhausted.

When you add records to an indexed sequential data set using the PUT macro, new entries are also made in the indexes. During resume load on a data set with a partially filled track or a partially filled cylinder, the track index entry or the cylinder index entry is overlaid when the track or cylinder is filled. If resume load abnormally ends after these index entries have been overlaid, a subsequent resume load will get a sequence check when adding a key that is higher than the highest key at the last successful CLOSE but lower than the key in the overlaid index entry. When the SYNAD exit is taken for a sequence check, register 0 contains the address of the highest key of the data set. Figure 1 graphically represents how records are added to an indexed sequential data set.

Figure 1. Adding Records to an Indexed Sequential Data Set