Information Technology Reference
In-Depth Information
can be used for the automation of a process where documents, information,
or tasks are passed from one participant to another to be processed, accord-
ing to a set of procedural rules. In the context of service-oriented architec-
ture, service-based approach is often employed to build grid workl ow
where legacy code, business logic, and grid-computation- related tasks are
wrapped or deployed as Web services that can then be integrated into
workl ow process. In this section, the basics of workl ow technology and
workl ow used in the grid environment are introduced. The start-of-the-art
service component architecture (SCA) that can be used to create a work-
l ow is discussed.
11. 3.1
Introduction to Workflow Technology
The workl ow technology will be introduced from the following aspects:
(1) service-oriented workl ow, (2) workl ow language, (3) workl ow engine,
and (4) workl ow system.
11.3.1.1
Service-Oriented Workfl ow
According to the Workl ow Management Coalition (WfMC), an interna-
tional organization for workl ow vendors, users, and research, the work-
l ow can be dei ned as “the automation of a business process, in whole or
parts, where documents, information or tasks are passed from one partici-
pant to another to be processed, according to a set of procedural rules” [49].
This dei nition can be extended to scientii c workl ow, which is a set of com-
ponents and relations between them used to dei ne a complex process from
simple building blocks [50]. Relations may be in the form of data links,
transferring information from the output of one component to the input
of another, or in the form of control links that state some conditions on
the execution of a component [50]. The component is a reusable building
block that performs some well-dei ned function within a process, and may
consume information and may produce output to provide information and
knowledge. These components can be implemented as Web services and
hosted by computing resources where external users can access.
Two approaches to compose the workl ow have been identii ed: (1)
job-based approach, and (2) service-based approach [51,53]. In the
job-based approach, the workl ow manager is responsible for the actual
processing of data by programs on physical resources. As this approach
is much close to the grid infrastructure, it allows the optimization of
submission rate, the dispatch rate, the scheduling rate, and so on [51, 54].
For example, Condor DAGMan [54] is a workl ow manager that adopts
this job-based approach where a child node will not start until its parents
have successfully completed. In the service-based approach, the work-
l ow manager is responsible for the transmission of data to remote
services and for the collection of the results. The workl ow manager is
 
Search WWH ::




Custom Search