We consider the problem of finding an optimal and sub-optimal allocation of program modules onto processors of a distributed computing system. A module causes two types of cost to be incurred at the processor to which it is allocated-an execution cost for processing the module, and a communication c
Multiple Instance Resource Allocation in Distributed Computing Systems
โ Scribed by R. Satyanarayanan; C.R. Muthukrishnan
- Publisher
- Elsevier Science
- Year
- 1994
- Tongue
- English
- Weight
- 531 KB
- Volume
- 23
- Category
- Article
- ISSN
- 0743-7315
No coin nor oath required. For personal study only.
โฆ Synopsis
A token-based distributed algorithm to perform the allocation of any one from a set of multiple instances of a shared resource type has been presented in this note. The number of messages generated by the algorithm per critical section execution for accessing a resource instance varies from 0 to (2(n-1)), where (n) is the number of nodes in the distributed system. An advantage of the algorithm presented here is that the upper bound on the number of messages generated per critical section execution is independent of the number of instances of the resource type covered by the algorithm. 1994 Academic Press, Inc.
๐ SIMILAR VOLUMES
In this paper, we examine three general classes of space-sharing scheduling policies under a workload representative of large-scale scientific computing. These policies differ in the way processors are partitioned among the jobs as well as in the way jobs are prioritized for execution on the partiti