A dynamic rescheduling algorithm for resource management in large scale dependable distributed systems

Scheduling is the key to distributed applications performance in large scale heterogeneous environments. For such systems resilience in case of faults can be approached at the level of rescheduling mechanisms. The performance of rescheduling is very important in the context of large scale distributed systems and dynamic behavior. The paper proposes a generic rescheduling algorithm, which can be used with a wide variety of scheduling heuristics which can be selected by users in advance, depending on system’s structure. The dynamic behavior of the rescheduling component is issued by a monitor, which periodically provides information about the systems’ states. The rescheduling component is designed as a middleware service that aims to increase the dependability of large scale distributed systems. The monitor component, which triggers the rescheduler when an error occurs, is also a middleware component. The rescheduling service was implemented for a real Grid environment. The proposed approach supports fault tolerance and offers an improved mechanism for resource management. The evaluation of the proposed rescheduling algorithm was performed using modeling and simulation. We present results confirming its performance and capabilities.