eModel: Addressing the Need for a Flexible Modeling Framework in Autonomic Computing

  • Authors:
  • C. H. Crawford;A. Dan

  • Affiliations:
  • -;-

  • Venue:
  • MASCOTS '02 Proceedings of the 10th IEEE International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunications Systems
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper describes a novel, flexible framework,eModel, designed to address the runtime requirementsof autonomic computing: on-line workload measurement,analysis, and prediction. The eModel architecture has beendeveloped using platform independent technology (XMLand Java) to allow for maximum portability while also allowingfor ease-of-integration with existing measurementand system management tools. The eModel toolkit consistsof a GUI based model builder tool, a data base deploymenttool, a runtime tool, and an analysis tool. In addition tothe toolkit, the eModel design provides a runtime architecturewhich can be deployed directly without using any interactionwith the GUI. The architecture is flexible enough toallow for incorporation with models of various complexity,including modeling techniques that require a hierarchicalapproach to attain reasonable accuracy based upon on-line,measured data. We present examples that illustrate eModelas a capacity planning tool as well as an augmentation toautonomic system management in an effort to highlight thetechnological gaps that the eModel framework is capable ofbridging.