Control interval access gives you access to the contents of a control interval; keyed access and addressed access give you access to individual data records.

Restriction: You cannot use control interval access to access a compressed data set. The data set can be opened for control interval access to permit VERIFY and VERIFY REFRESH processing only.

With control interval access, you have the option of letting VSAM manage I/O buffers or managing them yourself (user buffering). Unless you specify user buffering, VSAM buffers obtained at OPEN time are used for reading or writing the control intervals. With keyed and addressed access, VSAM always manages I/O buffers. If you select user buffering, you have the further option of using improved control interval access, which provides faster processing than normal control interval access. With user buffering, only control interval processing is permitted. See Improved Control Interval Access.

Control interval access permits greater flexibility in processing entry-sequenced data sets. With control interval access, change the RBAs of records in a control interval and delete records by modifying the RDFs and the CIDF.

When using control interval processing, you are responsible for maintaining alternate indexes. If you have specified keyed or addressed access (ACB MACRF={KEY|ADR},...) and control interval access, then those requests for keyed or addressed access (RPL OPTCD={ KEY|ADR},...) cause VSAM to upgrade the alternate indexes. Those requests specifying control interval access will not upgrade the alternate indexes. You are responsible for upgrading them. Upgrading an alternate index is described in Maintaining Alternate Indexes.

Restriction: You should not update key-sequenced data sets or variable-length RRDSs with control interval access. You cannot use control interval access with compressed format data sets. When you process control intervals, you are responsible for how your processing affects indexes, RDFs, and CIDFs. Bypassing use of the control information in the CIDF and RDFs can make the control interval unusable for record level processing. For instance, key-sequenced data sets depend on the accuracy of their indexes and the RDFs and the CIDF in each control interval.