IBM® Informix® Warehouse Accelerator, Version 12.10
Queries that benefit from acceleration
The database server uses information about the existing data marts to determine which types of queries are sent to the accelerator server for processing.
The following types of queries benefit the most from being sent
to the accelerator server:
- Complex, ad hoc queries that look for trends or exceptions
- Queries that access a large subset of the database, often by using sequential scans
- Queries that involve aggregation functions such as COUNT, SUM, AVG, MAX, MIN, and VARIANCE
- Queries that often create reports that group data by time, product, geography, customer set, or market
- Queries that involve star joins or snowflake joins of a large fact table with several dimension tables
Only SELECT queries that refer to a fact table, or SELECT queries that join a fact table with one or more dimension tables are processed by the accelerator server.
The
same restrictions apply when a query has a query block that contains
an online analytical processing (OLAP) window function or UNION, or
UNION ALL set operators.
- When a query contains an OLAP window function, the database server uses Informix® Warehouse Accelerator for the underlying SELECT, JOIN, GROUP BY, and PROJECT operations.
- When a query contains a UNION or UNION ALL set operator, Informix Warehouse Accelerator evaluates each SELECT statement separately. If either SELECT statement is eligible for query acceleration, it is sent to Informix Warehouse Accelerator for processing.
The characteristics of a query or a query block determine if it
can be processed by the accelerator server. For a query to be accelerated
it must match the following conditions:
- The query must reference only the tables and columns that are included in the data mart definition.
- The query must use only the supported data types.
- The query must reference a table that is marked as a fact table in the data mart definition.
- The query must use only the supported join types.
- The query must use only supported aggregate and scalar functions.
Release date: March 2015