As increases in data volume lengthen batch processing time, accelerating the speed of batch processing has become crucial. A batch job distributed execution system accelerates batch processing by dividing a large volume of input data and then processing the divided data on multiple hosts in parallel. For details about batch job distributed execution systems, see the manual uCosminexus Grid Processing Server User's Guide.

By storing data in HiRDB that you wish to process using a batch job distributed execution system, you can configure a system that uses a batch job distributed execution system together with HiRDB. This section provides a system configuration example in which a batch job distributed execution system is used with HiRDB, and explains an operation mode (operation mode that adds units while HiRDB is active) that works well in such a configuration.

Note that the operation mode that combines a batch job distributed execution system with HiRDB, and that adds units while HiRDB is active, can be used only with the Linux version of a HiRDB parallel server configuration.

Organization of this section

29.1.1 System configuration example that uses a batch job distributed execution system with HiRDB

29.1.2 Parallel execution of batch jobs by adding units to HiRDB