MUTEAM: an experience in the design of robust multi-microprocessor systems

A multimicroprocessor prototype, called MuTEAM, is presented. Its architecture is oriented towards the exploitation of decentralized control policies for both error treatment and resource management The hardware architecture is composed of a set of clusters, each cluster comprising a tightly coupled multiprocessor. Within a cluster, specialized hardware is provided for memory protection and management, and for interprocessor communication. The kernal implements the run-time support for a CSP-like message-based concurrent programming language. The prototype includes a set of fault-tolerance mechanisms. These are inserted in the system programming language, the run-time support and the hardware configuration. They are being used to develop a fault-treatment policy, consisting of separate phases of diagnosis, reconfiguration and backward error recovery. Each phase is executed by a set of parallel processes, without a supervisor entity. Diagnosis is provided by a set of dedicated processes, with one process running in each computer element, while reconfiguration and recovery are achieved by a correct organization of the application software.