SmartProduction for DB2 can detect a wide range of production inefficiencies (over 50) which, when resolved, can greatly reduce system resources utilization (e.g., CPU and I/O) and significantly cut job elapsed times.

SmartProduction for DB2 applies the following four key strategies to improve the performance of the DB2 production batch workload:

Improve DB2 Application Efficiency

Many site-developed DB2 applications are not as efficient as they could be. This causes a problem when the degree of inefficiency is significant.

Sample points: Large sorts could be eliminated via additional index; improper SQL commands performed under batch.

Increase DB2 Parallelism

Many DB2 applications do not make use of the DB2 parallelism methods which allow to perform concurrent I/O and CPU. Using DB2 parallelism can significantly reduce elapsed time.

Sample points: An application does not make use of parallelism; parallelism has been requested, but is not honored by DB2 due to various problems.

Improve Efficiency of DB2 Housekeeping

Many utility executions are performed in order to backup, restore, copy or reorganize the DB2 database. These utility executions often do not use all performance options available.

Sample points: Multiple utility executions could be combined into one utility execution; utility execution does not use proper SORT optimization techniques.

Eliminate DB2 Contentions

Batch tasks are frequently delayed, or even aborted, due to DB2 contention problems. Eliminating (or at least greatly reducing) the occurrences of contentions can significantly reduce elapsed time.

Sample points: Deadlocks, timeouts, lock suspensions, lock escalations, rid pool failures, buffer shortages.