Two more best practices have recently been released for IBM Smart Analytics Systems. Thanks to Serge Boivin for this information!
This document describes IBM Smart Analytics Systems scenarios where DB2 software manages multiple database partitions in a cluster. The scenarios describe optimization methods that can help improve performance in a short time.
The two main tasks for query optimization are:
- Provide the DB2 optimizer with rich statistical information. The DB2 optimizer uses the statistical information to perform its analysis and find a good execution plan. This article describes the best practices for gathering statistics.
- Determine the best query execution plan. If a query needs further tuning, you must inspect the explain plan for the query. This article provides a list of patterns to look for in explain plans. Each pattern is paired with recommended actions.
The key design goals of ingest application design are to balance the required rate of ingest with the availability of processing resources, to ingest data without affecting data availability, and to maintain flexibility in the volumes of data ingested. Identifying service level objectives for the speed and volume of data ingested into the data warehouse will determine the architecture, design, and development of the data ingest application.
See also my previous posts on Best Practices: