DB2 Version 10.1 for Linux, UNIX, and Windows

Performance enhancements

Version 10.1 contains numerous SQL performance enhancements that continue to make the DB2® data server an industrial-strength data server solution that is suitable for any size of organization.

Enhanced query performance for common SQL statements
A number of performance improvements have been included in DB2 V10.1 to improve the speed of many queries.
RUNSTATS and database statistics improvements
There are a number of improvements to the RUNSTATS command to make the gathering of statistics faster in some cases. The command parameters also are simplified.
Optimization profile supports registry variables and inexact matching
The optimization profile can now be used to set certain registry variables and supports inexact matching. Inexact matching can be used for better matching when compiling query statements.
Statistical views enhance statistics and statistic gathering for the query optimizer
There are new features of statistical views which the DB2 query optimizer can now use to generate better access plans, improving the performance of certain queries.
Intrapartition parallelism improvements
One goal of the DB2 query optimizer is to choose parallel execution strategies that maintain data balance among subagents and keep them equally busy. In this release, the parallelization capabilities of the optimizer have been further enhanced to enable more workloads to better use multi-core processors.
Enhanced memory sharing on large POWER7 systems running AIX
A DB2 V10.1 system can now determine the hardware topology on AIX® systems running POWER7® to potentially enhance query performance by effectively sharing memory.
Improved query performance through more efficient data and index prefetching
DB2 V10.1 features smart data prefetching and smart index prefetching, which improves query performance and reduces the need to reorganize tables and indexes.
Improved performance for queries on tables with composite indexes
The DB2 query optimizer can now create additional access plans that might be more efficient for queries with index gaps in their start-stop keys by using a jump scan operation.
Performance of star schema based queries is improved
Star schema performance improvements include an improved detection algorithm and a new join method.
FP2: Recovery history file improvements might increase performance
The process of writing to the recovery history file in DB2 Version 10.1 Fix Pack 2 is now more efficient. This improvement might result in general speed increases, especially as the recovery history file size increases. These speed improvements occur automatically; no change to the system or queries are required.
FP4: Optimizing data warehouse queries
A number of improvements have been made to optimize the performance of data warehouse queries.