Model-driven self-management of legacy applications

  • Authors:
  • Markus Debusmann;Markus Schmid;Reinhold Kroeger

  • Affiliations:
  • Department of Computer Science, Distributed Systems Lab, Fachhochschule Wiesbaden – University of Applied Sciences, Wiesbaden, Germany;Department of Computer Science, Distributed Systems Lab, Fachhochschule Wiesbaden – University of Applied Sciences, Wiesbaden, Germany;Department of Computer Science, Distributed Systems Lab, Fachhochschule Wiesbaden – University of Applied Sciences, Wiesbaden, Germany

  • Venue:
  • DAIS'05 Proceedings of the 5th IFIP WG 6.1 international conference on Distributed Applications and Interoperable Systems
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

Increasing complexity of todays applications and services leads to the emerging trend of self-managing systems. Legacy applications often offer interfaces for manual management and very rarely do provideself-management features. Therefore it is very important to reuse existing management interfaces for achieving self-manageability. This paper presents our model-driven self-management approach of legacy applications. We introduce a framework for model-driven service level management which transforms abstract SLAs defined in UML into concrete SLA descriptions and deploys them for management. These SLAsare used to define the goals for our self-management agent which is responsible for providing feedback control. Its management knowledge is transformed from UML models using also a model-driven approach.