Allocation of Real-Time Computations under Fault Tolerance Constraints

Allocation of resources in \next-generation" real-time operating systems requires some important features in addition to those demonstrated by current systems, resulting in an increased complexity of each system. The allocation is closely related to the scheduling, and the two are based on time considerations, rather then on a static priority scheme. The allocation is fault tolerance motivated, to cope with the application's reliability goals. Distributed system issues and adaptive behavior requirements increase the complexity and signi cance of the allocation approach. The allocation scheme we propose here accomplishes the hard real-time goal of guaranteeing a deadline satisfaction in case the job is accepted. In addition, this allocation scheme supports fault tolerance objectives in both damage containment and resiliency requirements. It does this in cooperation with a schedulability veri cation mechanism, and with an object architecture in which for each object there exists a calendar that maintains the time of its execution. A nice feature of this scheme is the way in which it can be used for reallocation while increasing the resiliency.