The following are issues that currently affect the Batch Processor:
When the Java Batch Processor starts multiple worker processes, the sub-processes are invoked using "java". In order for execution to be successful, a suitable Java run-time must be available from the operating system path.
Inferred relationships and instances are not supported as outputs in this version of the Batch Processor. While they can be used in a rulebase, any inferred instances will not appear in the output. Also, inferred relationships cannot be specified as outputs; only attributes of existing instances may be specified as outputs.
The number of cases the Batch Processor can write to the database (if database output is specified) within a single transaction is controlled by the --blocksize configuration option. If these database chunks contain many entity instances, you may need to increase the memory allocation of the batch processor process to avoid out of memory errors.
If there is an error updating the database, the whole update chunk will be rolled back.