Quality-of-service-aware fault tolerance for grid-enabled applications

Abstract

This study first reviews how grid-enabled applications can be provided with fault tolerance. Existing methods, implemented either in the grid application/middleware or in a Generalized Multi-Protocol Label Switching (GMPLS)-based network, are outlined. Then, the paper shows the advantages of integrating application/middleware fault-tolerant schemes, such as service replication, with GMPLS network-layer fault-tolerant schemes, such as path restoration. An integrated fault-tolerant scheme is capable of providing flexible QoS-aware fault tolerance while minimizing the necessary computational and network resources. In the end, the implementation of the proposed integrated scheme in a Video-on-Demand (VoD) application is experimentally validated.