Probabilistic allocation of tasks on desktop grids
- 1 April 2008
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE) in 2008 IEEE International Symposium on Parallel and Distributed Processing
Abstract
While desktop grids are attractive platforms for executing parallel applications, their volatile nature has often limited their use to so-called "high-throughput" applications. Checkpointing techniques can enable a broader class of applications. Unfortunately, a volatile host can delay the entire execution for a long period of time. Allocating redundant copies of each task to hosts can alleviate this problem by increasing the likelihood that at least one instance of each application task completes successfully. In this paper we demonstrate that it is possible to use statistical characterizations of host availability to make sound task replication decisions. We find that strategies that exploit such statistical characterizations are effective when compared to alternate approaches. We show that this result holds for real-world host availability data, in spite of only imperfect statistical characterizations.Keywords
This publication has 9 references indexed in Scilit:
- On Resource Volatility in Enterprise Desktop GridsPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2006
- Modeling Machine Availability in Enterprise and Wide-Area Distributed Computing EnvironmentsPublished by Springer Nature ,2005
- Characterizing and evaluating desktop grids: an empirical studyPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2004
- Generalized communicators in the Message Passing InterfacePublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- Experience with the Condor distributed batch systemPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- An annotated overview of system-reliability optimizationIEEE Transactions on Reliability, 2000
- Reliability optimization of series-parallel systems using a genetic algorithmIEEE Transactions on Reliability, 1996
- A statistical identity linking folded and censored distributionsJournal of Economic Dynamics and Control, 1995
- The interaction of parallel and sequential workloads on a network of workstationsPublished by Association for Computing Machinery (ACM) ,1995