On the modeling of parallel access to shared data

ABSTRACT

A model is constructed of a database that can be accessed and modified concurrently by a number of users, and an approximate solution is presented. The resource allocation policies considered involve dynamic acquisition of entities and locking; deadlock is avoided by limiting the number of consecutive attempts to acquire a particular entity. The accuracy of the approximation is evaluated by simulations. Several generalizations aimed at improving the practicality of the model are described.

REFERENCES

• 1: E. G. Coffman, Jr. , E. Gelenbe , B. Plateau, Optimization of the number of copies in a distribution data base, Proceedings of the 1980 international symposium on Computer performance modelling, measurement and evaluation, p.257-263, May 28-30, 1980, Toronto, Ontario, Canada
• 2: Gelenbe, E. and Sevcik, K.C. Analysis of update synchronization for multiple copy databases. In Proc. 3rd Berkeley Workshop on Distribution, Data Management, and Computer Networks, San Francisco, 1978.
• 3: Mohan, C. Distributed database management: Progress, problems, and future directions. Working paper WP-7802, Computer Science Dept., Univ. Texas, Austin, 1978.
• 4: D. Potier , Ph. Leblanc, Analysis of locking policies in database management systems, Communications of the ACM, v.23 n.10, p.584-593, Oct. 1980 [doi>10.1145/359015.359021]
• 5: Daniel R. Ries , Michael Stonebraker, Effects of locking granularity in a database management system, ACM Transactions on Database Systems (TODS), v.2 n.3, p.233-246, Sept. 1977 [doi>10.1145/320557.320566]

INDEX TERMS

Primary Classification:
  D. Software
  D.4 OPERATING SYSTEMS

Additional Classification:
  D. Software
  D.4 OPERATING SYSTEMS

  H. Information Systems
  H.2 DATABASE MANAGEMENT

  I. Computing Methodologies
  I.6 SIMULATION AND MODELING

General Terms:
Management, Performance

Keywords:
concurrency, database systems, parallel process, performance models