Various methods of repairing a corrupted database

There are three basic ways:

Use standard IMS recovery methods

This is the best way to repair a database. It requires much less expertise, and usually can be accomplished in less time.

Modify the database manually with a zapping utility

This method requires a lot of expertise to accomplish successfully. It is highly error-prone, even when done by an experienced programmer. The person who decides where and how to zap the database must have a good knowledge of the internal formats of IMS databases.

Reorganize the database

This method can be used only for certain kinds of errors involving logical pointers or counter fields.

There are times when it is impossible to repair a database with standard IMS recovery methods. For example, when HALDB databases contain corrupted HALDB partition reorganization numbers, duplicate ILKs, or potentially duplicate ILKs. In this case, the HD Pointer Checker and the IMS Database Repair Facility are very valuable tools for the systems programmer or database administrator.

It might also be desirable to use zapping methods when your recovery job is known to be long. You can determine that you can analyze the database errors and repair them with zaps in a shorter time than is required to recover the databases. If you have expertise in the use of IMS, this can be a cost-effective way to repair a large database.