Mathematics Colloquia and Seminars

Return to Colloquia & Seminar listing

Description

Data Centers have grown significantly in both size and complexity. Some Data Centers today consist of tens of thousands of servers with high-speed network connections for both inter- and intra-IDC communications. Managing both infrastructure and applications in such large and complex environments raises many challenges since Data Center operating costs can escalate very quickly, damaging the profitability of these businesses. The key to effective management of these large data complexes is to continually evaluate and refine business models and processes, and to employ decision support and automation tools for assessing performance and making good operational and strategic decisions. In this talk, I address the resource allocation problem (RAP) for large scale Data Centers using mathematical optimization techniques. The RAP problem is defined as follows: given a physical topology of resources in a large Data Center, and an application with certain architecture and requirements, we want to determine which resources in the physical topology should be assigned to the application architecture such that application requirements and bandwidth constraints in the network are satisfied, while communication delay between assigned servers is Minimized.