Tuning databases and queries on Oracle RAC

Database load patterns

Leading database engines are far too complex for one-dimensional evaluations. They have different features and architectural principles, e.g. "shared nothing" vs. "shared all". Advantages and disadvantages can be assessed in the context of the specific application area.

Once the database technology is selected, the database performance is usually a question of sizing: assuming some typical load patterns, how much and what type of resources are required in order to cope with the expected load.

There are some environment variables which are out of the ordinary evaluation scope but still highly impact performance. One such example is the know-how of the personnel authorized to run ad-hoc queries. Complex joins on large amount of data can overstrain even the most sophisticated built-in optimizer.

Our contribution

We offer expert know-how in database and statement tuning. Our primary focus is Oracle RAC but also IBM DB2 and Teradata. We assess the load pattern, resource contention, database tuning parameters, stored procedures and identify the bottlenecks. We elaborate on a classification of long-running queries and identify the applications or ad-hoc queries causing trouble. Subsequently, in collaboration with the DBA, we suggest short and medium-term actions to fight the issues and monitor the resulting changes.