The Research And Implement Of Cloud Workflow Systems

Workflow systems originated from office automation which started in1970s to support the office information management for accomplishing simple business tasks. In the last decade, workflow systems become more oriented to the process automation of large scale business and scientific applications. Many workflow systems have been deployed on high performance computing infrastructures such as cluster, peer-to-peer (p2p), and grid computing. One of the driving forces is the increasing demand of large scale instance intensive and data/computation intensive workflow applications which are common in both e-business and e-science application areas. Typical examples include such as the instance intensive securities exchange process in a stock market, the flight booking process in a travel agency, and the data and computation intensive the weather forecast process in Meteorology. These large scale workflow applications normally require the support of powerful high performance computing infrastructures (e.g. advanced CPU units, large memory space and high speed network).To meet these high resource requirements, expensive computing infrastructures including such as supercomputers, data servers and fibre networks are purchased, installed, and maintained. However, the problems of resource scalability and elasticity still exist in the conventional computing paradigm. Since most of these resources are self-contained and organised in a heterogeneous way, resource scalability is very low. Due to such a problem, most of the computing resources during off-peak times are largely idle, and thus results in the low ROI (return on investment) and a giant waste of energy consumption.In recent years, cloud computing is emerging as the latest distributed computing paradigm and attracts increasing interests of researchers in the area of Distributed and Parallel Computing, Service Oriented Computing and Software Engineering. Compared with conventional computing paradigms, cloud computing can provide a pool of abstracted, virtualised, dynamically scalable, managed computing power, storage, platforms, and services are delivered on demand to external customers over the Internet. Therefore, cloud computing can provide scalable resources on demand to system requirement. Evidently, it is possible to utilise cloud computing to address the problems of resource scalability and elasticity for managing large scale workflow applications. Therefore, the investigation of workflow systems based on cloud computing, namely cloud workflow systems, is a timely issue and worthwhile for increasing efforts.This paper discusses the design of cloud workflow systems, and focuses on three fundamental aspects, viz. system architecture, functionality and quality of service (QoS management). Through the research of the new cloud computing paradigm and the conventional workflow systems, we emphasise on the reuse and adaptation of existing methodologies and strategies rather than build from the scratch. Finally, we implemented a cloud workflow systems based on meteorological data access services. And we prove that the design scheme of cloud workflow systems is feasible.