The second batch performance mechanism provided by the application addresses the issue of database I/O contention. Improvements to database I/O in batch processing are always worth making, especially as batch windows reduce and the case and client loads increase. To this end a number of in-memory caches have been introduced for core Cúram entities which are available for re-use by Cúram batch processes.
Note that data which is accessed repeatedly during the eligibility processing is not limited to that stored in core Cúram entities. As a result, consideration should be given by customers to the caching of custom entities (e.g. evidence) which are accessed as part of this process.
It is also worth noting that these caches have been constructed so that they cannot be used in on-line mode. When on-line, because the application server is in control of the thread scheduling, the consistency between the cached data and that on the database cannot be guaranteed.