Load Balancing in Multiprocessor Systems
Abstract
We present an algorithm for dynamic load balancing in a multiprocessor system that minimizes the number of accesses to the shared memory. The algorithm assumes no information, probabilistic or otherwise, regarding task arrivals or processing requirements. For k processors to process n tasks, the algorithm incurs O(k log k log n) potential memory collisions in the worst case. The algorithm itself is a simple variation of the strategy of visiting the longest queue. The key idea is to delay reporting task arrivals and completions, where the delay is a function of dynamic loading conditions.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1989
- Accession Number
- ADA210144
Entities
People
- Michael C. Loui
- Milind A. Sohoni
Organizations
- University of Illinois Urbana–Champaign