Degree of multiprogramming in page-on-demand systems
Communications of the ACM
Sim/61: A simulation measurement tool for a time-shared, demand paging operating system
Proceedings of the SIGOPS workshop on System performance evaluation
CP-67 measurement and analysis overhead and throughput
Proceedings of the SIGOPS workshop on System performance evaluation
A probabilistic framework for system performance evaluation
Proceedings of the SIGOPS workshop on System performance evaluation
RESOURCE ALLOCATION IN MULTIPROCESS COMPUTER SYSTEMS
RESOURCE ALLOCATION IN MULTIPROCESS COMPUTER SYSTEMS
Measurement and performance of a multiprogramming system
SOSP '69 Proceedings of the second symposium on Operating systems principles
Two approaches for measuring the performance of time-sharing systems
SOSP '69 Proceedings of the second symposium on Operating systems principles
The instrumentation of multics
SOSP '69 Proceedings of the second symposium on Operating systems principles
Multiprogramming system performance measurement and analysis
AFIPS '68 (Spring) Proceedings of the April 30--May 2, 1968, spring joint computer conference
Measurement and analysis of large operating systems during system development
AFIPS '68 (Fall, part I) Proceedings of the December 9-11, 1968, fall joint computer conference, part I
AFIPS '75 Proceedings of the May 19-22, 1975, national computer conference and exposition
An interactive software engineering tool for memory management and user program evaluation
AFIPS '74 Proceedings of the May 6-10, 1974, national computer conference and exposition
| Hi-index | 0.00 |
The techniques of multiprogramming originated in an attempt to better utilize a computer system's resources. Multiprogramming supervisory systems are usually rather complicated and their performance is still poorly understood. Some of the reasons why performance should be analyzed are given in Reference 1. It is possible to monitor the system with hardware devices, but analysis must often wait several hours or days before it can be performed. Also, there are quantities that are impossible to reach with conventional hardware monitoring devices. In particular, process identities are lost. This means that some kind of software monitoring method must be envisaged.