Predicting queue times on space-sharing parallel computers
- 22 November 2002
- conference paper
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 209-218
- https://doi.org/10.1109/ipps.1997.580894
Abstract
We present statistical techniques for predicting the queue times experienced by jobs submitted to a space-sharing parallel machine with first-come-first-served (FCFS) scheduling. We apply these techniques to trace data from the Intel Paragon at the San Diego Supercomputer Center and the IBM SP2 at the Cornell Theory Center. We show that it is possible to predict queue times with accuracy that is acceptable for several intended applications. The coefficient of correlation between our predicted queue times and the actual queue times from simulated schedules is between 0.65 and 0.72.Keywords
This publication has 7 references indexed in Scilit:
- A comparison of workload traces from two production parallel machinesPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- A Model for Speedup of Parallel ProgramsPublished by Defense Technical Information Center (DTIC) ,1997
- Using Queue Time Predictions for Processor AllocationPublished by Defense Technical Information Center (DTIC) ,1997
- Exploiting process lifetime distributions for dynamic load balancingACM SIGMETRICS Performance Evaluation Review, 1996
- Models and algorithms for coscheduling compute-intensive taks on a network of workstationsJournal of Parallel and Distributed Computing, 1992
- Predictability of process resource usage: a measurement-based study on UNIXIEEE Transactions on Software Engineering, 1989
- Load-balancing heuristics and process behaviorPublished by Association for Computing Machinery (ACM) ,1986