If a database file is cached in memory, the DBMS looks in the cache before reading a database page from disk. If the page resides in the cache, the retrieval I/O is eliminated. If the page must be read from the disk, it is saved in the cache to satisfy future retrieval operations. Database files with a high number of I/Os are good candidates for caching in memory.
There are two basic types of file caching: shared cache which uses coupling facility services to enable a single cache to be shared by multiple central versions and memory cache which is accessible only by a single central version. The remainder of this discussion focuses on memory cache. For information about shared cache, see the CA IDMS System Operations Guide.
Note: Memory caching is available only for non-native VSAM files.
To enable the use of memory cache, take the following steps:
If the operating system supports 64-bit storage, the cache is allocated in 64-bit storage if sufficient memory is available. If no or not enough 64-bit storage is available to hold the entire file, the file will not be cached in memory. For details, see 7.13, "DMCL Statements".
Note: For more information about operating-specific considerations in using memory cache and 64-bit storage, see the CA IDMS System Operations Guide.
|
Copyright © 2014 CA.
All rights reserved.
|
|