Configuring parallel processing for the DB2 connector with DB2 for Linux, UNIX, and Windows databases
You can configure the DB2® connector
to read or write data in parallel mode. The distribution of parallel
processes across hardware is based on the DB2 for Linux®, UNIX, and Windows partition distribution.
Defining DB2 for Linux, UNIX, and Windows nodes and ETL nodes
You must define the DB2 for Linux, UNIX, and Windows nodes or ETL nodes in the parallel engine configuration file, also called the APT configuration file. The DB2 connector uses information specified in the APT configuration file to determine the number of processing nodes based on the number of partitions that are identified in the target table.
Limiting the number of processing nodes for bulk load
The DB2 connector uses one player process for each database partition. However, you can limit parallelism for bulk loads by configuring the DB2 connector to run one player process for multiple partitions.
Configuring the DB2 connector to use direct connections
If you configure the DB2 connector to use direct connections, the connector sends data directly to the appropriate DB2 data node without first sending the data to the DB2 admin node.
High availability for DB2 systems
You can configure DB2 for high availability, to ensure that a failing DB2 node is replaced with a backup node.