|
|
|
CA Compress uses variations of three fundamental data compression methods. Run Length data compression basically eliminates repeating characters. A string of repeating characters is replaced by the character and the number of times it is repeated.
The Huffman method of data compression is fundamentally the substitution of bit codes for data characters. The codes are constructed such that frequently occurring characters have very short codes while those occurring less frequently or rarely have codes that are longer than the character being replaced. The net overall storage required to store the data is typically less after code substitution.
Hardware Compression Assist uses the IBM Hardware Compression facility, which uses Ziv-Lempel compression dictionaries.
In addition to other considerations, the generic algorithms—Super Express, Standard Tables, and Hardware Compression—do not use a user-defined FDT. For this reason, they require no serialization of their compression and expansion logic, and this increases I/O throughput in busy CICS systems or other instances in which there are many concurrent I/O requests.
This section contains the following topics:
| Copyright © 2012 CA. All rights reserved. |
|