 |
    |
Resource Planning for Massive Number of Process Instances
Jiajie Xu, Chengfei Liu, and Xiaohui Zhao
Centre for Complex Software System and Services Faculty of Information and Communication Technologies Swinburne University of Technology Melbourne, Australiajxu@groupwise.swin.edu.au
cliu@groupwise.swin.edu.au
xzhao@groupwise.swin.edu.au
Abstract. Resource allocation has been recognised as an important topic for business process execution. In this paper, we focus on planning resources for a massive number of process instances to meet the process requirements and cater for rational utilisation of resources before execution. After a motivating example, we present a model for planning resources for process instances. Then we design a set of heuristic rules that take both optimised planning at build time and instance dependencies at run time into account. Based on these rules we propose two strategies, one is called holistic and the other is called batched, for resource planning. Both strategies target a lower cost, however, the holistic strategy can achieve an earlier deadline while the batched strategy aims at rational use of resources. We discuss how to find balance between them in the paper with a comprehensive experimental study on these two approaches.
LNCS 5870, p. 219 ff.
lncs@springer.com
© Springer-Verlag Berlin Heidelberg 2009