Essentially, adding a record amounts to updating by record identification. You can refer to the record using either a relative block address or a relative track address.

If you want to add a record passing a relative block address, the system converts the address to an actual track address. That track is searched for a dummy record. If a dummy record is found, the new record is written in place of it. If there is no dummy record on the track, you are informed that the write operation did not take place. If you request the extended search option, the new record will be written in place of the first dummy record found within the search limits you specify. If none is found, you are notified that the write operation could not take place.

In the same way, a reference by relative track address causes the record to be written in place of a dummy record on the referenced track or the first within the search limits, if requested. If extended search is used, the search begins with the first record on the track. Without extended search, the search can start at any record on the track. Therefore, records that were added to a track are not necessarily located on the track in the same sequence they were written in.