Information Technology Reference
In-Depth Information
work here to help understand the scientii c workl ow background. More
projects can be referred to in [3].
•
Swinburne Decentralized Workl ow for Grid (SwinDeW-G) is a
decentralized workl ow system [1]. Since in e-Science, a workl ow
is naturally computation and/or data intensive, it is believed by
SwinDeW-G that decentralized/P2P-based execution with modest
centralized management is one of the good solutions. In this
sense, SwinDeW-G was developed. In SwinDeW-G, workl ows
are executed by distributed peers. In grid computing, a peer can
be just a grid service. The interaction between peers is performed
in a P2P fashion [9] rather than via the centralized engine.
The Chimera system implements support for establishing virtual
•
catalogs that can be used to describe and manage information
on how a data product in an application has been derived from
other data [10]. This information can be queried, and data trans-
formation operations can be executed to regenerate the data
product.
The Pegasus project develops systems to support mapping and
•
execution of complex workl ows in a grid environment. The
Pegasus framework uses the Chimera system for abstract work-
l ow description and Condor DAGMan and schedulers for work-
l ow execution. It allows construction of abstract workl ows and
mappings from abstract workl ows to concrete workl ows that are
executed in the grid.
The Kepler project develops a scientii c workl ow management
•
system based on the notion of actors. Application components can
be expressed as actors that interact with each other through
channels. The actor-oriented approach allows the application
developer to reuse components and compose workl ows in a
hierarchical model.
Adapting Computational Data Streams (ACDS) is a framework
•
proposed by Isert and Schwan [11]. It addresses construction and
adaptation of computational data streams in an application. A
computational object performs i ltering and data manipulation,
and data streams characterize data l ow from data servers or from
running simulations to the clients of the application.
Dv is a framework proposed by Aeschlimann et al. [12]. It
•
is based on the notion of active frames. An active frame is an
application-level mobile object that contains application data,
called frame data, and a frame program that processes the data.
Active frames are executed by active frame servers running on
the machines at the client and at remote sites.
Search WWH ::
Custom Search