A Multiple Resource Model for a Batch-Processing Multiprogramming System.

Abstract

A batch-processing multiprogramming system is modeled as a multiple resource system. The interactions and relationships among the methods of allocating individual resources are examined. The basic model is a closed, cyclic three-stage model with a fixed number of jobs. The three stages are the permanent resource allocation stage, a CPU stage, and an I/O stage. Permanent resources are those which are assigned to a job, or job step, for the total time the job is in the system. The CPU and I/O resources are termed demand resources. A set of transition probabilities, representing the transition from the permanent resource stage to the demand resource stage, is determined from the discrete probability distributions of resource requests. Then using fixed, but unknown, service time distributions, CPU dispatching rules, and I/O scheduling rules, an equilibrium mean-value rate analysis is given. The principal result shows the relationship between the CPU and I/O stages. Finally the model is analyzed under hyperexponential service time distributions, a geometric CPU request distribution, and three specific CPU dispatching algorithms. Numerical examples are given and discussed. (Author)

Document Details

Document Type
Technical Report
Publication Date
Mar 01, 1971
Accession Number
AD0722332

Entities

People

  • Donald L. Boyd

Organizations

  • University of Iowa

Tags

DTIC Thesaurus Topics

  • Algorithms
  • Automata
  • Batch Processing
  • Determinants (Mathematics)
  • Mathematics
  • Multiprogramming
  • Probability
  • Probability Distributions
  • Scheduling (Production)
  • Transitions

Fields of Study

  • Engineering

Readers

  • Computer Networking
  • Mathematical Modeling and Probability Theory.
  • Systems Analysis and Design