Resolution Based on MRAA for Generalized Distributed Deadlocks

A deadlock occurs when there is a cycle of processes waiting for the resources held by others. This deadlock can be resolved only when the cycle is broken i.e. when one of the processes is victimized. None of the previous resolution algorithms choose a victim based on the work done by it before it is aborted. When it aborts, it cancels its pending requests and releases all the resources it had. The aborted process has to restart all the requests to complete its work. The performance of the system deteriorates as a process that has most of its resources allocated is often chosen as a victim. Hence in the proposed algorithm, a victim is chosen based on the Minimum Resource Allocated Already (MRAA) principle. This minimizes the wasted work done and hence the performance of the distributed systems is improved.

[1]  Soojung Le,et al.  Fast detection and resolution of generalized distributed deadlocks , 2002, Proceedings 10th Euromicro Workshop on Parallel, Distributed and Network-based Processing.

[2]  Walter A. Burkhard,et al.  Resolution of Deadlocks in Object-Oriented Distributed Systems , 1989, IEEE Trans. Computers.

[3]  Neelam Soundarajan,et al.  Efficient distributed deadlock detection and resolution using probes, tokens, and barriers , 1997, Proceedings 1997 International Conference on Parallel and Distributed Systems.

[4]  Akim Demaille,et al.  Correctness of a distributed deadlock resolution algorithm for the single request model , 1995, Proceedings Euromicro Workshop on Parallel and Distributed Processing.

[5]  Federico Fariña,et al.  A safe distributed deadlock resolution algorithm for the OR request model , 1998, Proceedings of the Sixth Euromicro Workshop on Parallel and Distributed Processing - PDP '98 -.

[6]  Federico Fariña,et al.  A Safe Algorithm for Resolving OR Deadlocks , 2003, IEEE Trans. Software Eng..

[7]  Federico Fariña,et al.  An O(n) distributed deadlock resolution algorithm , 2006, 14th Euromicro International Conference on Parallel, Distributed, and Network-Based Processing (PDP'06).

[8]  Soojung Lee,et al.  Efficient generalized deadlock detection and resolution in distributed systems , 2001, Proceedings 21st International Conference on Distributed Computing Systems.

[9]  Naphtali Rishe,et al.  Efficient algorithms for detection and resolution of distributed deadlocks , 1995, Proceedings.Seventh IEEE Symposium on Parallel and Distributed Processing.

[10]  Nacer Farajzadeh,et al.  Optimal detection and resolution of distributed deadlocks in the generalized model , 2006, 14th Euromicro International Conference on Parallel, Distributed, and Network-Based Processing (PDP'06).

[11]  Laura M. Haas,et al.  Distributed deadlock detection , 1983, TOCS.

[12]  J. R. Garitagoitia,et al.  A Distributed Deadlock Resolution Algorithm for the AND Model , 1999, IEEE Trans. Parallel Distributed Syst..

[13]  Soojung Lee,et al.  Fast, Centralized Detection and Resolution of Distributed Deadlocks in the Generalized Model , 2004, IEEE Trans. Software Eng..

[14]  Nacer Farajzadeh,et al.  An Efficient Generalized Deadlock Detection and Resolution Algorithm in Distributed Systems , 2005, The Fifth International Conference on Computer and Information Technology (CIT'05).