Online prediction of the running time of tasks: Summary

Peter A Dinda*

*Corresponding author for this work

Research output: Contribution to journalConference articlepeer-review

20 Scopus citations


The running time advisor (RTA), which can predict the running time of a compute-bound task on a shared host running a variant of the Unix operating system, was described. These predictions are valuable for scheduling the soft real-time tasks of distributed interactive applications such as scientific visualizations. The RTA predicts running time as a confidence interval computed to the application's requested confidence level. Confidence levels provide a simple abstraction to the application and even provide information to enable valid statistical reasoning in the scheduling process.

Original languageEnglish (US)
Pages (from-to)336-337
Number of pages2
JournalPerformance Evaluation Review
Issue number1
StatePublished - 2001
EventJoint International Conference on Measurement and Modeling of Computer Systems - Cambridge, MA, United States
Duration: Jun 16 2001Jun 20 2001

ASJC Scopus subject areas

  • Software
  • Hardware and Architecture
  • Computer Networks and Communications


Dive into the research topics of 'Online prediction of the running time of tasks: Summary'. Together they form a unique fingerprint.

Cite this