Fault management in ECLIPSE

ECLIPSE is a next-generation virtual telecommunication network that provides multimedia services that integrate voice, video, text and images. ECLIPSE facilitates the modular decomposition of new telecommunication services. In this paper, we sketch the challenges we face in making ECLIPSE highly available when running on top of a heterogenous and widely distributed system. We describe how we approach these challenges using the modular decomposition of services provided by ECLIPSE, together with its novel failure detection and recovery strategies.