TY - GEN

T1 - Efficient parallel algorithms can be made robust

AU - Kanellakis, Paris C.

AU - Shvartsman, Alex A.

PY - 1989

Y1 - 1989

N2 - The efficient parallel algorithms proposed for many fundamental problems, such as list ranking, computing preorder numberings and other functions on trees, or integer sorting, are very sensitive to processor failures. The requirement of efficiency (commonly formalized using Parallel-time x Processors as a cost measure) has led to the design of highly tuned PRAM algorithms which, given the additional constraint of simple processor failures, unfortunately become inefficient or even incorrect. We propose a new notion of robustness, that combines efficiency with fault tolerance. For the common case of fail-stop errors, we develop a general (and easy to implement) technique to make robust many efficient parallel algorithms, e.g., algorithms for all the problems listed above. More specifically, for any dynamic pattern of fail-stop errors with at least one surviving processor, our method increases the original algorithm cost by at most a multiplicative factor polylogarithmic in the input size.

AB - The efficient parallel algorithms proposed for many fundamental problems, such as list ranking, computing preorder numberings and other functions on trees, or integer sorting, are very sensitive to processor failures. The requirement of efficiency (commonly formalized using Parallel-time x Processors as a cost measure) has led to the design of highly tuned PRAM algorithms which, given the additional constraint of simple processor failures, unfortunately become inefficient or even incorrect. We propose a new notion of robustness, that combines efficiency with fault tolerance. For the common case of fail-stop errors, we develop a general (and easy to implement) technique to make robust many efficient parallel algorithms, e.g., algorithms for all the problems listed above. More specifically, for any dynamic pattern of fail-stop errors with at least one surviving processor, our method increases the original algorithm cost by at most a multiplicative factor polylogarithmic in the input size.

UR - http://www.scopus.com/inward/record.url?scp=0024942821&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0024942821&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0024942821

SN - 0897913264

T3 - Proceedings of the Annual ACM Symposium on Principles of Distributed Computing

SP - 211

EP - 221

BT - Proc Eighth ACM Symp Princ Distrib Comput

PB - Publ by ACM

T2 - Proceedings of the Eighth Annual ACM Symposium on Principles of Distributed Computing

Y2 - 14 August 1989 through 16 August 1989

ER -