Models and algorithms for reliability-oriented task-allocation in redundant distributed-computer systems
- Illinois Univ., Chicago, IL (USA)
One often quoted advantage of distributed-computer systems is their potential for improved system reliability. Hardware redundancy is a common technique to help achieve this potential. When the hardware configuration is fixed, system reliability depends mainly on resource allocation, viz, software allocation. For a distributed-computer system, the software application is called a task and is a set of cooperating modules. This paper discusses reliability-oriented task-allocation for heterogeneous distributed-computer systems with hardware redundancy. The authors consider redundancy of both processors and communication links. A quantitative allocation model is derived and used to present and discuss models and algorithms for systems with level-2 or level-3 redundancy. The results contribute to the body of knowledge on task allocation and provide an alternative to performance-oriented methods.
- OSTI ID:
- 5345674
- Journal Information:
- IEEE (Institute of Electrical and Electronics Engineers) Transactions on Reliability; (USA), Journal Name: IEEE (Institute of Electrical and Electronics Engineers) Transactions on Reliability; (USA) Vol. 38:1; ISSN 0018-9529; ISSN IEERA
- Country of Publication:
- United States
- Language:
- English
Similar Records
An automata theory approach to solving the dynamic task allocation problem in distributed computer systems
TASK ALLOCATION IN GEO-DISTRIBUTED CYBER-PHYSICAL SYSTEMS