Categorizing Computing Assets According to Communication Patterns

In today's dynamic information society, organizations critically depend on the underlying computing infrastructure. Tracking computing devices as assets and their usage helps in the provision and maintenance of an efficient, optimized service. A precise understanding of the operational infrastructure and its users also plays a key role during the negotiation of outsourcing contracts and for planning mergers and acquisitions. Building an accurate inventory of computing assets is especially difficult in unknown heterogeneous systems and networking environments without prior device instrumentation. User mobility and mobile, not-always-signed-on, computing devices add to the challenge. We propose to complement basic network-based discovery techniques with the combined log information from network and application servers to compute an aggregate picture of assets, and to categorize their usage with data-mining techniques according to detected communication patterns.