Author: Chen, J.
Paper Title Page
THCPL03 A Success-History Based Learning Procedure to Optimize Server Throughput in Large Distributed Control Systems 1182
  • Y. Gao, T.G. Robertazzi
    Stony Brook University, Stony Brook, New York, USA
  • K.A. Brown
    BNL, Upton, Long Island, New York, USA
  • J. Chen
    Stony Brook University, Computer Science Department, Stony Brook, New York, USA
  Funding: Work supported by Brookhaven Science Associates, LLC under Contract No. DE-SC0012704 with the U.S. Department of Energy.
Large distributed control systems typically can be modeled by a hierarchical structure with two physical layers: Console Level Computers (CLCs) and Front End Computers (FECs). The controls system of the Relativistic Heavy Ion Collider (RHIC) consists of more than 500 FECs, each acting as a server providing services to a potentially unlimited number of clients. This can lead to a bottleneck in the system. Heavy traffic can slow down or even crash a system, making it momentarily unresponsive. One mechanism to circumvent this is to transfer the heavy communications traffic to more robust higher performance servers, keeping the load on the FEC low. In this work, we study this client-server problem from a different perspective. We introduce a novel game theory model for the problem, and formulate it into an integer programming problem. We point out its difficulty and propose a heuristic algorithms to solve it. Simulation results show that our proposed schemes efficiently manage the client-server activities, and result in a high server throughput and a low crash probability.
video icon Talk as video stream:  
slides icon Slides THCPL03 [1.321 MB]  
DOI • reference for this paper ※  
Export • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)