A Simulation Environment for WS-Agreement Negotiation Compliant Strategies

A Simulation Environment for WS-Agreement Negotiation Compliant Strategies

Abstract

Today, the dominant market mechanism for trading cloud resources is the static supermarket approach where consumers buy resources from providers without negotiation. The recent development of Amazon’s EC2 spot market shows that dynamic market mechanisms are gaining popularity. Dynamic market mechanism using multi-round bilateral negotiations - aka Bazaar negotiations - are a promising approach for trading cloud resources. On such a Bazaar-market consumers and providers negotiate cloud resources which are specified in SLA’s. To foster the development of Bazaar-markets the Open Grid Forum released the WS-Agreement Negotiation standard. Currently, WS-Agreement Negotiation compliant strategies are missing. Hence, in the paper at hand we introduce a time dependent negotiation strategy which is compliant to the WS-Agreement Negotiation standard. For a systematic analysis of the strategy a market simulation environment is necessary. As existing scientific simulation environments are not able to simulate Bazaar-markets we extended the CloudSim framework by developing the so called Bazaar-Extension which implements the WS-Agreement Negotiation standard. Using the Bazaar-Extension cloud service consumers and providers can be created and strategies can be assigned to them. The so created market can be simulated and the resulting resource allocation can be analyzed. The architecture of the simulation framework as well as its evaluation are introduced in this paper.

Grafik Top
Authors
  • Mach, Werner
Grafik Top
Shortfacts
Category
Paper in Conference Proceedings or in Workshop Proceedings (Paper)
Event Title
The 19th International Conference on Information Integration and Web-based Applications & Services (iiWAS2017)
Divisions
Workflow Systems and Technology
Subjects
Informatik Allgemeines
Computersimulation
Event Location
Salzburg
Event Type
Conference
Event Dates
4-6 December 2017
Date
December 2017
Export
Grafik Top