ON GUARANTEEING HARD REAL-TIME TASKS

The development of complex real-time systems has become a reality with the move from uniprocessor to distributed systems. In general, these systems require adaptability in order to cope with many types of timing constraints. The most critical part in the development of these systems is the scheduler. Our approach is based on the view that if the system is adaptable then the reliability is easier to be attained. We are proposing the use of milestone notion within the Least Laxity Scheduler Algorithm and augmenting it with an ''alert'' mechanism that is generated using the information provided by the milestones. The scheduler is fully implemented in an industrially available kernel namely RTXC/MP1.