1st Intenational IEEE Workshop on Information Assurance Middleware for Communications

Research Article

Method using Modified Chord Algorithm to Balance Pool Element Ownership among Registrars in a Reliable Server Pooling Architecture

  • @INPROCEEDINGS{10.1109/COMSWA.2007.382489,
        author={C.S. Chandrashekaran and  Walter  L. Johnson and  Abhijit Lele},
        title={Method using Modified Chord Algorithm to Balance Pool Element Ownership among Registrars in a Reliable Server Pooling Architecture},
        proceedings={1st Intenational IEEE Workshop on Information Assurance Middleware for Communications},
        publisher={IEEE},
        proceedings_a={IAMCOMM},
        year={2007},
        month={7},
        keywords={Collaboration  Computer architecture  Degradation  Fault tolerance  Internet  Peer to peer computing  Redundancy  Reliability engineering  Transport protocols  Web server},
        doi={10.1109/COMSWA.2007.382489}
    }
    
  • C.S. Chandrashekaran
    Walter L. Johnson
    Abhijit Lele
    Year: 2007
    Method using Modified Chord Algorithm to Balance Pool Element Ownership among Registrars in a Reliable Server Pooling Architecture
    IAMCOMM
    IEEE
    DOI: 10.1109/COMSWA.2007.382489
C.S. Chandrashekaran1,*, Walter L. Johnson2,*, Abhijit Lele1,*
  • 1: Motorola Labs, Bagmane Technology Park, Bangalore 560 093
  • 2: Motorola Labs, 1301 E. Algonquin Road, Schaumburg, Illinois 60196
*Contact email: chandra.cs@motorola.com, walterjohnson@motorola.com, abhijit.lele@motorola.com

Abstract

The Internet Engineering Task Force (IETF) reliable server pooling (RSerPool) group provides a protocol framework that supports server redundancy and session failover providing fault tolerance. The RSerPool protocol framework allows a pool of redundant information sources to be viewed as a single transport end point and is thereby able to provide persistent connections for end users. A performance evaluation of the RSerPool protocols show that early arriving registrars are disproportionately loaded as compared to late arriving registrars. This results in an unbalanced load across registrars and leads to performance degradation of the overall reliable server pooling system. In this paper, we propose a methodology based on the chord algorithm for balancing the load across all registrars. A performance evaluation is carried out to showcase the effectiveness of the algorithm.