Abstract | ||
---|---|---|
We propose a processor-level technique called Selective Replication, by which the application can choose where in its application stream and to what degree it requires replication. Recent work on static analysis and fault-injection-based experiments on applications reveals that certain variables in the application are critical to its crash- and hang-free execution. If it can be ensured that only the computation of these variables is error-free, then a high degree of crash/hang coverage can be achieved at a low performance overhead to the application. The Selective Replication technique provides an ideal platform for validating this claim. The technique is compared against complete duplication as provided in current architecture-level techniques. The results show that with about 59% less overhead than full duplication, selective replication detects 97% of the data errors and 87% of the instruction errors that were covered by full duplication. It also reduces the detection of errors benign to the final outcome of the application by 17.8% as compared to full duplication. |
Year | DOI | Venue |
---|---|---|
2007 | 10.1109/DSN.2007.75 | DSN |
Keywords | Field | DocType |
selective replication detects,high degree,processor-level selective replication,current architecture-level technique,index terms—replication,critical code sections,complete duplication,low performance overhead,selective replication technique,full duplication,selective replication,reconfiguration,application stream,processor-level technique,instruction sets,prototypes,redundancy,high performance computing,computer networks,duplication,detectors,critical variable,indexing terms,error detection,hardware,static analysis | Crash,Computer science,Parallel computing,Static analysis,Real-time computing,Error detection and correction,Redundancy (engineering),Hang,Gene duplication,Fault injection,Computation,Distributed computing | Conference |
ISSN | ISBN | Citations |
1530-0889 | 0-7695-2855-4 | 17 |
PageRank | References | Authors |
0.78 | 22 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Nithin Nakka | 1 | 90 | 6.14 |
Karthik Pattabiraman | 2 | 1030 | 55.17 |
Ravishankar Iyer | 3 | 720 | 35.52 |