This section provides a general description of IDMS Presspack, a large-volume database compression and decompression tool for IDMS/DB.
IDMS Presspack is a large-volume database compression and decompression tool for IDMS/DB. This section provides a description of the product and guidelines for using it.
Advantages of Using IDMS Presspack
Compressing your database records:
  • Slows the demand for new DASD.
    By compressing data, you use your DASD more efficiently and as a result, you can defer the purchase of more hardware.
  • Alleviates physical space constraints in your data center.
    DASD requires physical space that may not be available.
  • Helps I/O bound jobs.
    Since more data is stored in the same amount of space, fewer I/Os are required to retrieve the same amount of data.
  • Reduces journaling activity.
    Since the records that are written to the journals are compressed, more records fit onto a journal. As a result, you do not have to offload the journals as often.
  • Speeds backup and restoration of your database.
    Since the same amount of data can be stored on fewer pages, the backup/restore utilities have to do fewer I/Os and, as a result, run faster.
What IDMS Presspack Compresses
IDMS/DB records or tables may contain control field data and may non-control field data. This component compresses the data beyond the last control field.
How IDMS Presspack Compresses
IDMS Presspack compresses database records by:
  1. Eliminating repeating character strings
  2. Using the Huffman Encoding Technique
  3. Applying the concept of successor tables
The result is efficient compression of your data.
Huffman Encoding Technique
IDMS Presspack uses an algorithm that is called the Huffman Encoding Technique. Bit strings of varying length are assigned to characters that are being compressed. Characters that occur most frequently are assigned the shortest bit strings. Characters that occur less frequently are assigned longer bit strings. Instead of every character using a fixed 8-bit code, some characters are represented by as few as 2 bits and a few characters require more than 8 bits.
Since the most frequently occurring characters get the shortest bit strings, the set of data requires less space to be stored.
Successor Tables
The successor table concept allows for the use of multiple Huffman conversion tables. IDMS Presspack decides which table to use on the next character (that is, the "succeeding" character) based on the type of character it expects to compress or decompress next. This technique is based on the premise that user data occurs in patterns and not in a random fashion.
By applying the concept of successor tables, IDMS Presspack can achieve further compression by assigning the same short bit string to more than one frequently occurring character. For example, the most frequently occurring alphabetic character and the most frequently occurring digit could both be represented by the shortest bit string. The number of successor tables that are used is related to the degree of compression you specify when you create a custom DCT.
How You Can Optimize Compression
You can optimize compression by passing selected records or tables through IDMSPASS, the compression optimization utility of IDMS Presspack. IDMSPASS allows you to achieve the highest degree of compression possible with IDMS Presspack. For further information on IDMSPASS, see Using IDMSPASS.