Linear Data Set

From Wikipedia, the free encyclopedia

An Linear Data Set (LDS) is a type of data set used by the VSAM computer data storage system. A Linear Data Set contains data that can be accessed as byte-addressable strings in virtual storage. The LDS is composed of VSAM Control Intervals (CI's) that are a multiple of 4096 bytes. An LDS contains no imbedded control information in the CI. There is no Control Interval Information Field (CIDF) or Record Information Field (RDF) that is in other VSAM file types. The entire control interval is data. If logical records exist in LDS, they must be unblocked or blocked by the program. The data is stored as undifferentiated bytes addressed by byte address. LDS is unlike other VSAM types except that it is defined by the IDCAMS VSAM utility program. When the LDS is defined, there is only a data component, similar to EDSD and RRDS. LDS's are used by DB2 and other programs which require byte addressable storage.

In other IBM Access Method techniques, records are retrieved or stored using requests at the record level. Linear datasets may use a technique called Data In Virtual (DIV). DIV allows an application program to map an LDS or a portion of an LDS into an address space, a data space, or a hiperspace. An LDS cluster may be referred to as a DIV object. Once the data set is mapped, the access of the data is essentially a paging operation.

Data is read into central storage using the paging algorithms. Pages are only read in when the address referenced refers to a page (all or part of a CI) that is not presently in storage. During Real Storage Manger (RSM) page steal processing, only changed pages in virtual storage are written back to auxiliary storage. Unchanged pages are discarded. If the pages no longer in storage are referenced again, they will be retrieved from the disk data set.

DIV is designed to improve the performance of applications that process large files non-sequentially. This make it useful for DBMS's, large arrays or table files.