Index Build function—features and benefits

The Index Build function performs the following functions:

• Creates a new secondary index for an existing database
• Maintains or rebuilds an existing secondary index
• Rebuilds a primary index for a HIDAM or PHIDAM database

The Index Build function generates several comprehensive reports that provide information about index availability and performance.

The Index Build function offers the following features and benefits:

• Support for HALDBs
  Index Build function supports PHIDAM databases. You can use the stand-alone Index Build function to build primary and secondary indexes for a HALDB. You can specify selected partitions.
• Index Build function does not require the healing process thereby reducing overhead.
  If a database has been opened for update, the healing process heals the pointers and any subsequent reference will not have to heal the pointer. If the database is opened without update (read-only), the healing occurs but the database is not updated; therefore, all subsequent references must heal the pointer again.
• Ability to build the primary index for HIDAM and PHIDAM databases and the secondary indexes in the same job step
  The Index Build function enables you to simplify your JCL by building multiple secondary indexes in one job step rather than in several separate steps. Only one command is required.
• Ability to build the primary index for a HIDAM or PHIDAM database and the secondary indexes in the same job step
  The Index Build function can rebuild the primary index for a HIDAM or PHIDAM database that has been destroyed. It takes less time for the Index Build function to rebuild the primary index than it does to reorganize the entire database or to recover the primary index from an image copy and apply the logs.
  You can build all indexes for a database in the same job step, with one pass of the data.
• Ability to scan a database sequentially rather than hierarchically
  The Index Build function determines the most efficient scanning method for a database. If possible, the Index Build function reads blocks sequentially, identifies segments, and extracts data needed to build an index. This sequential scan can run significantly faster than a traditional hierarchical scan.
• Ability to create empty secondary indexes for use with empty databases
  If you need to create a secondary index for an empty database, you can simplify the process by using the Index Build function.
• Simplified command and keyword structure
  The Index Build function has only one command: BUILD. You can scan the database and build the index in the same job step with the same BUILD command; you no longer need to specify XSCN and XLOD commands. The Index Build function recognizes most SECONDARY INDEX UTILITY commands and keywords; it converts the commands and keywords to the appropriate Index Build function equivalents and writes a report that explains which commands and keywords were converted.
• Elimination of image copies of index databases
  Most installations perform frequent, scheduled image copies of index databases, even if they do not need to recover those databases. Because the Index Build function can rebuild lost or damaged indexes by using only the primary database as input, image copies of indexes are no longer needed for recovery. You can avoid the overhead and expense of these image copies without compromising data recoverability.
• Ability to sort and load indexes in multiple address spaces
  You can save elapsed time by sorting and loading indexes in separate address spaces. When you specify SIUSORT(E), the Index Build function creates a separate address space for each index. While the Index Build function job is running, you will see a separate MVS job for each index. When all jobs are complete, the output is merged into the original job. For more information, see SIUSORT.