Combining Advantages - Unified Data Stores in Global Enterprises

Abstract

Nowadays, data is heterogeneously stored across various data silos in global enterprises. For instance, OLAP profits in terms of performance from data immanent optimizations within the silo.We motivate the use of unified, enterprise-wide data stores which built the basis for combining data from different sources in arbitrary ways while eliminating redundancies. Without replicas all data operations, e.g. ad-hoc operational reporting, perform directly on real-time business data. The need for extracting data from various silos in preparation for reporting vanishes. A unified store consumes less system resources while hard disk and main memory capacity are used more efficiently.We analyze demands of applications in global enterprises and introduce a layered model. Our architecture enables rapid development of optimized applications. The data persistence becomes part of the infrastructure comparable to the network stack in network applications: it is no longer implemented solely for every application.

Data Loading and Caching Strategies in Service-Oriented Enterprise Applications

“…Hybrid reverse caching strategies are promising, because they cache data structures rather than static responses. This caching type can be built upon unified data stores to eliminate redundant and duplicate data [17]. Our approach to cache ES responses can be situated in the above described reverse proxy schema.…”