Previous Topic: Monitoring and Tuning Database PerformanceNext Topic: Monitoring Facilities


Monitoring Guidelines

Why You Need to Monitor

Eventually, a database may begin to outgrow its initial allocation of space, with resulting increased I/O and poor response time. If you don't monitor your databases on a regular basis, these conditions can become critical, forcing you to take emergency actions at an inconvenient time.

Suggested Monitoring Schedule

Consider using the following schedule as the basis for monitoring database performance:

Monitoring tool

Monitoring frequency

Information provided

JREPORT 004

Daily

Summary information on the database processing activities for each program recorded in the journal file

IDMSDBAN report 2

Weekly

Area detail statistics, such as number of logically full pages and number of relocated records

IDMSDBAN report 5

Monthly

Set detail statistics, such as the number of pages needed to store a chained set

PRINT SPACE

Daily

Area space utilization statistics

IDMSDBAN (all reports)

Monthly or as needed

Set statistics, including broken chains, record data, and area data

Keep a History of Meaningful Statistics

Keep a history of meaningful statistics so that you can identify abnormal conditions when they arise.

SQL Considerations

Most of the information in this chapter applies to both SQL and non-SQL defined databases. Text that applies to only one or the other will be noted. In addition, much of the chapter applies to the physical structures that underlie the database definition. Therefore, one set of terms will be used for these physical entities. For example, chain sets are the physical structure used to implement both SQL linked constraints and non-SQL sets defined with the MODE IS CHAIN clause.