Abstract  Desktop  grid  is  an  exciting  discipline  for  high  thr oughput  applications  but  due  to  inherent  resource  v olatility,  desktop  grids  are  not  feasible  for  short  lived  appl ications  that  require  rapid  turnaround  time.  Efficient  and  more  knowledgeable  resource  selection  mechanism  can  make   it  possible.  In  this  paper,  we  propose  a  group  based  resource  scheduling mechanism. The groups are made by using  three measures: Collective impact of CPU and RAM, spot checking and  task  completion  history.  We  evaluated  the  proposed  mechanism  over  a  network  of  900  nodes  having  varied   resources  and  behavior and found that excluding desktop resources  on the basis of just clock rates is not a good idea and RAM should also,  be considered as a collective parameter besides spo t checking and task completion history. We also, show that the appropriate  scheduling  mechanisms  can  only  be  implemented  after   the  grouping  of  resources  on  computing  strength  and  behavior.  The  proposed  mechanism  ensures  that  tasks  are  allocated   to  hosts  with  higher  probability  of  tasks  completion  that  reduces  tasks  failures and improves fault tolerance.   

References

[1] Anderson P., Emulating Volunteer Computing Scheduling Policies, in Proceedings of IEEE International Symposium on Parallel and Distributed Processing Workshops and Phd Forum , Shanghai, pp. 1839-1846, 2011.

[2] Anderson D. and Fedak G., The Computational and Storage Potential of Volunteer Computing, in Proceedings of the 6 th IEEE International Symposium on Cluster Computing and the Grid , pp. 73-80, 2006.

[3] Berkeley Open Infrastructure for Network Computing, available at: https://boinc.berkeley.edu, last visited 2013.

[4] Choi S., Baik M., Hwang C., Gil J., and Yu H., Mobile Agent based Adaptive Scheduling Mechanism in Peer to Peer Grid Computing, in Proceedings of the International Conference on Computational Science and its Applications , Singapore, pp. 936-947, 2005.

[5] Compute Against Cancer, available at: http://www.computeagainstcancer.org, last visited 2013.

[6] Fight Aids at Home, available at: http://www.fightaidsathome.org, last visited 2013.

[7] Great Internet Mersenne Prime Search, available at: http://www.mersenne.org, last visited 2013.

[8] Hanandeh F., Khazaaleh M., Ibrahim H., and Latip R., CFS: A New Dynamic Replication Strategy for Data Grids, the International Arab Journal of Information Technology , vol. 9, no. 1, pp. 94-99, 2012.

[9] Khan K., Hyder I, Chowdhry B., Shafiq F., and Ali H., A Novel Fault Tolerant Volunteer Selection Mechanism for Volunteer Computing, Sindh University Research Journal-Science Series , vol. 44, no. 3, pp. 501-506, 2012.

[10] Kondo D., Chien A., and Casanova H., Scheduling Task Parallel Applications for Rapid Turnaround on Enterprise Desktop Grids, Journal of Grid Computing , vol. 5, no. 4, pp. 379-405, 2007.

[11] Search for Extraterrestrial Intelligence., availabl e at: http://setiathome.ssl.berkeley.edu, last visited 2013. 280 The International Arab Journal of Information Tech nology VOL. 13, NO. 2, March 2016

[12] Silaghi G., Domingues P., Araujo F., Silva L., and Arenas A., Defeating Colluding Nodes in Desktop Grid Computing Platforms, in Proceedings of IEEE International Symposium on Parallel and Distributed Processing, Florida, USA, pp. 1-8, 2008.

[13] Toth D. and Finkel D., Improving the Productivity of Volunteer Computing by Using the Most Effective Task Retrieval Policies, Journal of Grid Computing , vol. 7, no. 4, 2009.

[14] Watanabe K., Fukushi M., and Horiguchi S., Optimal Spot-checking for Computation Time Minimization in Volunteer Computing, Journal of Grid Computing , vol. 7, no. 4 , 2009. Mohammed Khan is the Director College of Computing and Information Sciences at PAF- Karachi Institute of Economic and Technology, Pakistan. He holds Post Graduation in Computer Science as well as in management science. He is experienced in academic leadership and has launched several bachelors programs at PAF-KIET. He has also completed several consultancy and training assignment at leading organizations. His research interests include distributed systems, intelligent and multi agent systems and performance evaluation. Irfan Hyder is working as the Dean College of Business Management and College Engineering and Sciences at Institute of Business Management, Pakistan. He holds a PhD and MS in computer science from University of Texas at Austin, USA. He has a wide experience in academic leadership, consultancy, entrepreneurial ventures, trainings, teaching, research, design and implementation of innovative programs and as a motivational speaker. In his career spanning over 1 8 years, he has worked in leadership positions at var ious organizations. He was deputy director at IBA and De an and VP at PAF-KIET. He has extensive industry consultancy experience for private and public secto r organizations. Ghayas Ahmed is serving as Deputy Controller in Federal Urdu University, Pakistan. He holds MS Computer Science from PAF- Karachi Institute of Economic and Technology. He also, possesses close to 11 years of software development experience. His research interests incl ude software engineering development patterns and distributed systems. Saira Begum is serving as Lecturer in Computer Science Department at Jinnah University for Women, Pakistan. Due to more than 6 years if software development experience, she is also heading the fin al year project committee. She holds MS Computer Science from PAF-Karachi Institute of Economic and Technology. Her research interests include programming paradigms and databases.