Scheduling Divisible Loads on Bus Networks with Arbitrary Processor Release Time and Start-Up Costs: XRMI

We present a novel algorithm to schedule divisible loads using multi-round installment load distribution. This algorithm considers additional real-world factors, such as task initialization overhead and arbitrary processor release times. We analyze this algorithm's properties and demonstrate its behaviors through multiple illustrative examples. The analysis provides both a deeper understanding of the divisible load scheduling difficulties when considering more real-world factors and allows us to design more efficient algorithms. We show that this algorithm generate better solutions for practical applications.

[1]  Richard P. Martin,et al.  Modeling communication pipeline latency , 1998, SIGMETRICS '98/PERFORMANCE '98.

[2]  Henri Casanova,et al.  Scheduling divisible loads on star and tree networks: results and open problems , 2005, IEEE Transactions on Parallel and Distributed Systems.

[3]  Bharadwaj Veeravalli,et al.  Scheduling divisible loads on heterogeneous linear daisy chain networks with arbitrary processor release times , 2004, IEEE Transactions on Parallel and Distributed Systems.

[4]  V Bharadwaj,et al.  Scheduling divisible loads in bus networks with arbitrary processor release times , 1996 .

[5]  Thomas G. Robertazzi,et al.  Divisible Load Scheduling for Grid Computing , 2003 .

[6]  Jacek Blazewicz,et al.  Scheduling Divisible Jobs on Hypercubes , 1995, Parallel Comput..

[7]  Bharadwaj Veeravalli,et al.  On the Influence of Start-Up Costs in Scheduling Divisible Loads on Bus Networks , 2000, IEEE Trans. Parallel Distributed Syst..

[8]  Henri Casanova,et al.  Multiround algorithms for scheduling divisible loads , 2005, IEEE Transactions on Parallel and Distributed Systems.

[9]  Maciej Drozdowski,et al.  Selected problems of scheduling tasks in multiprocessor computer systems , 1997 .

[10]  Hai Jin,et al.  Challenges of Grid Computing , 2005, WAIM.

[11]  Debasish Ghose,et al.  Large matrix-vector products on distributed bus networks with communication delays using the divisible load paradigm: performance analysis and simulation , 2001, Math. Comput. Simul..

[12]  Thomas G. Robertazzi,et al.  Ten Reasons to Use Divisible Load Theory , 2003, Computer.

[13]  Debasish Ghose,et al.  Foreword (Special Issue of Cluster Computing on Divisible Load Scheduling) , 2004, Cluster Computing.

[14]  Emmanuel Jeannot,et al.  On the distribution of sequential jobs in random brokering for heterogeneous computational grids , 2006, IEEE Transactions on Parallel and Distributed Systems.

[15]  Debasish Ghose,et al.  Scheduling Divisible Loads in Parallel and Distributed Systems , 1996 .

[16]  Debasish Ghose,et al.  Divisible Load Theory: A New Paradigm for Load Scheduling in Distributed Systems , 2004, Cluster Computing.