Javelin++: scalability issues in global computing

Abstract

SUMMARYJavelin is a Java-based infrastructure for global computing. This paper presents Javelin++, an extension of Javelin, intended to support a much larger set of computational hosts. Contributions to scalability and fault tolerance are presented. This is the focus of the paper. Two scheduling schemes are presented: probabilistic work stealing and deterministic work stealing. The distributed deterministic work stealing is integrated with a distributed deterministic eager scheduler, which is one of the paper's primary original contributions. An additional fault tolerance mechanism is implemented for replacing hosts that have failed or retreated. A Javelin++ API is sketched, then illustrated on a raytracing application. Performance results for the two schedulers are reported, indicating that Javelin++, with its broker network, scales better than the original Javelin.

Browser-based Harnessing of Voluntary Computational Power

“…In Javelin++ [62] two scheduling policies were proposed, both based on work-stealing (where when host is idle, it "steals" task from other hosts). Both policies require that tasks can be easily divided into smaller chunks.…”

A New Approach to Resource Discovery in Grid Computing

“…Centralized resource discovery approaches (Berman et al, 2003;Chien et al, 2003;Foster & Kesselman, 1997;Germain et al, 2000;Mutka & Livny, 1987;Neary et al, 1999) are one of the mechanisms which suffer from single point of failure and bottlenecks.…”

Formal Models and Methods on Volunteer Computing

“…19, No.5, May 2008 Charlotte [6] ,Bayanihan [7] ,Javelin [8] ,GUCHA [9] ,XtremWeb [10] ,WebCom [11] 等 等 . 其 中 以 BOINC(Berkeley open infrastructure for network computing) [12] • 组合:gup(Nd1.RC)=gup(Nd2.RC).由 Def.…”