Title :
Load distributing for locally distributed systems
Author :
Shivaratri, Niranjan G. ; Krueger, Phillip ; Singhal, Mukesh
Author_Institution :
Dept of Comput. & Inf. Sci., Ohio State Univ., Columbus, OH, USA
Abstract :
The problem of judiciously and transparently redistributing the load of the system among its nodes so that overall performance is maximized is discussed. Several key issues in load distributing for general-purpose systems, including the motivations and design trade-offs for load-distributing algorithms, are reviewed. In addition, several load-distributing algorithms are described and their performances are compared. These algorithms are sender-initiated algorithms, receiver-initiated algorithms, symmetrically initiated algorithms, and adaptive algorithms. Load-distributing policies used in existing systems are examined, and conclusions about which algorithm might help in realizing the most benefits of load distributing are drawn.<>
Keywords :
distributed algorithms; distributed processing; performance evaluation; adaptive algorithms; load-distributing algorithms; locally distributed systems; receiver-initiated algorithms; sender-initiated algorithms; symmetrically initiated algorithms; Algorithm design and analysis; Availability; Communication networks; Computer networks; Costs; Delay; Distributed computing; Fluctuations; Heuristic algorithms; Microprocessors;