Table scan access implies that all records in a file (table or view) are read. The records are processed in the order in which they reside in the file. This means that no ordering is guaranteed unless ordering is specified in the query (seeSort Access). As each record is read, selection criteria (if any) are applied by the operating system, and only the records that match the criteria are returned to the application.
Table scan access provides good performance when approximately 20% or more of the records in the file match the selection criteria.
If the file contains a large percentage of deleted records, consider reorganizing the file to remove the deleted records to further improve performance.
Table scan access may not be the most efficient access method if only a small percentage (< 20%) of the records of a relatively large file (>2,000 records) is selected. In this case, you should consider creating an index with key fields that match the record selection fields.