Overview of database reorganization
As information is added, updated, and deleted, a database becomes physically disorganized, decreasing operating efficiency. More I/O operations are needed to retrieve a segment and its dependents when they are physically disorganized than when they are physically adjacent. When this occurs, response time slows noticeably, with a corresponding decrease in end-user productivity.
A database must be reorganized periodically to improve operating efficiency and to implement changes. Physical reorganization involves placing each root segment together with its dependent segments into one block (or into adjacent blocks if they do not fit into a single block) so that fewer I/O operations are needed to retrieve the root and its dependents. Any secondary indexes for the reorganized database must be rebuilt, and logical relationship pointers between databases must be resolved and updated. To enable recovery of the reorganized database, an image copy is required after the reorganization.
As the size and complexity of a database increases, reorganization processing time increases. Traditional reorganization utilities force a database to be offline during reorganization, and a sufficiently large window of down time must be found for the reorganization. You might not be able to find this window of down time, or you might not be able to take advantage of secondary indexes and logical relationships because of the additional reorganization requirements.
The need for an online reorganization utility has become more urgent as more and more installations strive to be available 24 hours a day, seven days a week.
This section contains the following topics: