In the last few years, many emulation systems have been developed to help researchers evaluate the effectiveness of new protocols and applications in realistic network scenarios. NEPTUNE (Network Emulation for Protocol TUNing and Evaluation) is a exible and scalable system developed at University of Napoli for the emulation of different network scenarios by means of a cluster of workstations. Setting up an emulation experiment in a cluster-based system requires, firstly, the ability to map virtual resources requested by an experimenter onto available physical resources and, secondly, the ability to exert a precise control over the allocated physical resources. These two requirements have much in common with resource management issues already addressed by the Grid computing community. Hence, we decided to exploit the Virtual Workspace concept at the foundation of the design of the NEPTUNE architecture. In this paper, we illustrate the peculiar virtualization requirements of a cluster-based emulation system and discuss how a Globus VirtualWorkspace based on Xen virtual machines can be used as the basis for implementing a distributed network emulation system.
Network emulation on Globus-based grids: mechanisms and challenges / Canonico, Roberto; DI GENNARO, Pasquale; Manetti, Vittorio; Ventre, Giorgio. - STAMPA. - Signals and Communication Technology Series:(2008), pp. 455-468.
Network emulation on Globus-based grids: mechanisms and challenges
CANONICO, ROBERTO;DI GENNARO, PASQUALE;MANETTI, VITTORIO;VENTRE, GIORGIO
2008
Abstract
In the last few years, many emulation systems have been developed to help researchers evaluate the effectiveness of new protocols and applications in realistic network scenarios. NEPTUNE (Network Emulation for Protocol TUNing and Evaluation) is a exible and scalable system developed at University of Napoli for the emulation of different network scenarios by means of a cluster of workstations. Setting up an emulation experiment in a cluster-based system requires, firstly, the ability to map virtual resources requested by an experimenter onto available physical resources and, secondly, the ability to exert a precise control over the allocated physical resources. These two requirements have much in common with resource management issues already addressed by the Grid computing community. Hence, we decided to exploit the Virtual Workspace concept at the foundation of the design of the NEPTUNE architecture. In this paper, we illustrate the peculiar virtualization requirements of a cluster-based emulation system and discuss how a Globus VirtualWorkspace based on Xen virtual machines can be used as the basis for implementing a distributed network emulation system.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.