Skip to main content

TS

Optimizing the performance of automated business processes executed on virtualized resources

Authors

Christian Janiesch, Ingo Weber, Michael Menzel and Jörn Kuhlenkamp

University of Karlsruhe

NICTA

Abstract

With few exceptions, to date the opportunities cloud computing offers to BPM technologies have been neglected. This paper investigates opportunities and challenges of implementing a cloud-aware BPM architecture for the benefit of process runtime optimization. Specifically, certain processes with predominantly automated tasks, such as in data transformation and data integration, may not allow reducing the overall process turnaround time any further through organizational projects. Improvement projects in this context usually target waiting times and change-over times, but do not necessarily optimize the actual processing time of individual tasks, especially if they are automated. We assume that in certain scenarios there may be a relation between a task’s processing time and its available computational resources. Hence, we argue that the overall process turnaround time can be optimized by making use of the elasticity of a cloud-aware BPM architecture. In theory, load balancers and horizontal scaling mechanisms can provide as much computational resources as necessary for a process to execute. However, in order to optimize process instances’ processing time as a whole, it is important not simply to over-provision resources but to meet the processes’ SLAs as precisely as possible. We present a model and method of cloud-aware business process optimization which aims at providing computational resources based on process knowledge. We describe a performance measurement experiment and evaluate it against the performance of standards load balancers and auto-scaling.

BibTeX Entry

  @inproceedings{Janiesch_WMK_14,
    author           = {Janiesch, Christian and Weber, Ingo and Menzel, Michael and Kuhlenkamp, Jörn},
    month            = jan,
    year             = {2014},
    keywords         = {business process management, cloud computing, optimization},
    title            = {Optimizing the Performance of Automated Business Processes Executed on Virtualized Resources},
    booktitle        = {Hawaii International Conference on System Sciences},
    pages            = {3818-3826},
    address          = {Big Island, HI, USA}
  }

Download

Served by Apache on Linux on seL4.