Abstract
This chapter introduces network-centric applications including Grid and Cloud
computing applications and presents the diverse underlying infrastructure required
by these network-centric applications. Traditionally, resource management for Grid
and Cloud computing considers IT resources as the main element in delivering
services. However, the need to move towards a collaborative network and IT
service-oriented approach has been proposed to improve service delivery. Orches-
tration of the network and IT resources required for Grid and Cloud computing
services provides a more optimum solution since the network infrastructure is
treated as a first-class resource. In turn, this influences the performance of the
application.
In order to manage the services in this infrastructure environment, it is impor-
tant to look at an architectural model involving different layers such as the optical
transport, control and service/middleware and the application layer. In addition,
this chapter presents a description of service delivery architectural models
stretching from the traditional to collaborative solutions. Collaboration can be
implemented either on the service/middleware layer or at the control plane level.
At the service layer, the implementation can be addressed in two ways. The first
can be through middleware extensions and the second through a separate service
layer, which encompasses the traditional middleware and a novel service plane.
The service plane interfaces northbound with the middleware and southbound
with the network control plane. Both implementations consider network and IT
resources as equally as first-class resources. Two middleware extensions, G-
Lambda and EnLIGHTened, and one service layer implementation, SOAFI, are
reported. At the control plane layer, control plane extensions required to deliver
network services need to be aware of the IT resources required in the Grid and
Cloud computing environment. G2MPLS is reported in terms of its architecture as
well as the protocol extensions required to support and deliver a Grid-enabled
control plane.
Finally, the heterogeneity and diversity of Grid and Cloud computing application requirements demand a flexible optical transport network able to provide
flexible network services. Thus, this chapter introduces flexible multi-granular,
transport solutions as well as sub-wavelength transport networks to cover the
required diverse transport requirements.
computing applications and presents the diverse underlying infrastructure required
by these network-centric applications. Traditionally, resource management for Grid
and Cloud computing considers IT resources as the main element in delivering
services. However, the need to move towards a collaborative network and IT
service-oriented approach has been proposed to improve service delivery. Orches-
tration of the network and IT resources required for Grid and Cloud computing
services provides a more optimum solution since the network infrastructure is
treated as a first-class resource. In turn, this influences the performance of the
application.
In order to manage the services in this infrastructure environment, it is impor-
tant to look at an architectural model involving different layers such as the optical
transport, control and service/middleware and the application layer. In addition,
this chapter presents a description of service delivery architectural models
stretching from the traditional to collaborative solutions. Collaboration can be
implemented either on the service/middleware layer or at the control plane level.
At the service layer, the implementation can be addressed in two ways. The first
can be through middleware extensions and the second through a separate service
layer, which encompasses the traditional middleware and a novel service plane.
The service plane interfaces northbound with the middleware and southbound
with the network control plane. Both implementations consider network and IT
resources as equally as first-class resources. Two middleware extensions, G-
Lambda and EnLIGHTened, and one service layer implementation, SOAFI, are
reported. At the control plane layer, control plane extensions required to deliver
network services need to be aware of the IT resources required in the Grid and
Cloud computing environment. G2MPLS is reported in terms of its architecture as
well as the protocol extensions required to support and deliver a Grid-enabled
control plane.
Finally, the heterogeneity and diversity of Grid and Cloud computing application requirements demand a flexible optical transport network able to provide
flexible network services. Thus, this chapter introduces flexible multi-granular,
transport solutions as well as sub-wavelength transport networks to cover the
required diverse transport requirements.
| Original language | English |
|---|---|
| Title of host publication | Cross-Layer Design in Optical Networks |
| Editors | Suresh Subramaniam, Maite Brandt-Pearce, Piet Demeester, Chava Vijaya Saradhi |
| Publisher | Springer, New York, NY |
| Pages | 263-290 |
| Volume | 15 |
| ISBN (Electronic) | 1935-3847 |
| ISBN (Print) | 978-1-4614-5670-4 |
| DOIs | |
| Publication status | Published - 2013 |