ProActive Caching: Generating Caching Heuristics for Business Process Environments

Today's complex and multi-layered enterprise systems demand fine-grained access control mechanisms supporting dynamic security policies for large and distributed repositories. Thus, the efficient evaluation of security policies becomes an important factor for the overall system performance, specifically with respect to systems with a high degree of user interaction like workflow systems. Caching approaches may help to address this situation. We propose ProActive Caching, a two-phased caching approach: in an offline phase, we automatically determine a workflow-specific heuristic for pre-computing cache entries. In an online phase, we use the previously determined heuristic for the cache management. The latter includes also the pre-computation of cache entries which already provides a performance improvement while evaluating a policy object for the first time. In this paper, we present a method for the automatic generation of a workflow specific caching heuristic, i.e., the offline phase.