Improving Cache Global Consistency and Hit Ratio in Dependency Objects with Semantic Spatial Locality Correlations

On requesting cache data in disk, the distribution of spatial locality is critical to access performance. Unfortunately, spatial locality properties of cached data is largely ignored, and only temporal locality is considered. Besides, an individual disk object might induce different dependency relations in different applications and possibly partial dependency to several distributed original data source (ODS). This interesting property might be neglected. Normally, these situations can be improved by effectiveness of storage caching, prefetching, and prediction of the user navigation behavior, data layout of storage systems and global consistency storage replicas. In this paper, we consider the problem and solve it using data mining techniques and a service routable consistency framework (Global Distributed Hierarchical Cache Consistent Model; GDHCCM). The model based on a scalable routing service algorithm that dynamic reconfiguration forwarding data path within hierarchical enterprise region portals. A novel hypergraph scheme was also proposed to represent the complex object relations among the applications. Instead of a local measure that depends only on common objects among patterns, we propose a global measure process based on the semantic properties of these patterns in the overall data set. The experiments show the effectiveness of the proposed framework. Apply scenario can reduce the global patch service cost; improve performances and minimum the turnaround time in access the scope of computer games or virtual environments (VE). Key-Words: cache, spatial locality, global consistency, hypergraph, prefetching, service routing, virtual environments