Labs of the World, Unite!!!

Abstract

eScience is rapidly changing the way we do research. As a result, many research labs now need non-trivial computational power. Grid and voluntary computing are well-established solutions for this need. However, not all labs can effectively benefit from these technologies. In particular, small and medium research labs (which are the majority of the labs in the world) have a hard time using these technologies as they demand high visibility projects and/or high-qualified computer personnel. This paper describes OurGrid, a system designed to fill this gap. OurGrid is an open, free-to-join, cooperative Grid in which labs donate their idle computational resources in exchange for accessing other labs' idle resources when needed. It relies on an incentive mechanism that makes it in the best interest of participants to collaborate with the system, employs a novel application scheduling technique that demands very little information, and uses virtual machines to isolate applications and thus provide security. The vision is that OurGrid enables labs to combine their resources in a massive worldwide computing platform. OurGrid is in production since December 2004. Any lab can join it by downloading its software from http://www.ourgrid.org.

A Communication Framework for Fault-Tolerant Parallel Execution

“…Condor [1] is a workload management system that can effectively harness wasted CPU power from otherwise idle desktop workstations. Other systems that build desktop computing grids include Entropia [5], iShare [6], and OurGrid [7]. Mechanisms applied for fault tolerance in PC grids, such as redundancy in BOINC and checkpointing in Condor [8] are important for long running sequential and bag-of-task codes, but are generally not sufficient for communicating parallel programs.…”

Scheduling Strategies for Cycle Scavenging in Multicluster Grid Systems

“…OurGrid [15] is an open platform that enables different research labs to share their idle computational resources when needed. OurGrid relies on a peer-to-peer incentive mechanism, called Network of Favors, which aims to make it in each participant's best interest to donate idle cycles, along with preventing free riding.…”

“…The implemented cycle scavenging mechanism runs alongside the regular grid scheduling, being unobtrusive to the jobs of higher priority (both local and grid jobs). Although cycle scavenging infrastructures do exist [11,15,22], our mechanism obviates the need for additional software installations on the compute nodes or any modifications to the resource managers of the clusters (e.g., SGE [9]), both of which would be administrative obstacles in multicluster grid systems. We exclusively target large-scale applications that can be modeled as Parameter Sweep Applications (PSAs) to run as cycle scavenging (CS) jobs.…”

LBG-SQUARE Fault-Tolerant, Locality-Aware Co-Allocation in P2P Grids

“…A Peer manages a set of resources on the behalf of user agents. Peers can share their resources with bartering [4,2,3] or market-based methods [3]. Bartering consists of fully distributed, non-monetary exchanges of computing time, and thus does not need a central bank.…”