The control of response times in multi-class systems by memory allocation
Communications of the ACM
Feedback coupled resource allocation policies in the multiprogramming-multiprocessor computer system
Communications of the ACM
Decomposability, instabilities, and saturation in multiprogramming systems
Communications of the ACM
A comparative analysis of disk scheduling policies
Communications of the ACM
A learning program which plays partnership dominoes
Communications of the ACM
Clustering Algorithms
Performance Improvement by Feedback Control of the Operating System
Proceedings of the Third International Symposium on Modelling and Performance Evaluation of Computer Systems: Performance of Computer Systems
Automatic load adjustment in time-sharing systems
Proceedings of the SIGOPS workshop on System performance evaluation
Experiments & measurements in computing
SIGME '73 Proceedings of the 1973 ACM SIGME symposium
An adaptive policy driven scheduler
SIGMETRICS '74 Proceedings of the 1974 ACM SIGMETRICS conference on Measurement and evaluation
A model for learning systems
A method for adaptive performance improvement of operating systems
A method for adaptive performance improvement of operating systems
A simulation study of adaptive scheduling policies in interactive computer systems
ACM SIGMETRICS Performance Evaluation Review
Perceptrons: An Introduction to Computational Geometry
Perceptrons: An Introduction to Computational Geometry
Insights into providing dynamic adaptation of operating system policies
ACM SIGOPS Operating Systems Review
Hi-index | 0.00 |
This paper presents a method for dynamic modification of operating system control parameters to improve system performance. Improved parameter settings are learned by experimenting on the system. The experiments compare the performance of alternative parameter settings in each region of a partitioned load-performance space associated with the system. The results are used to modify important control parameters periodically, responding to fluctuations in system load and performance. The method can be used to implement adaptive tuning, to choose between alternative algorithms and policies, or to select the best fixed settings for parameters which are not modified. The method was validated and proved practical by an investigation of two parameters governing core quantum allocation on a Sperry Univac 1100 system. This experiment yielded significant results, which are presented and discussed. Directions for future research include automating the method, determining the effect of simultaneous modifications to unrelated control parameters, and detecting dominant control parameters.