Shared-memory Metacomputing

这项研究的重点是开发技术和算法，这将允许空闲的工作站被利用的并行和分布式共享内存应用程序的机会。目标环境的范围从局域网（LAN）到广域网（WAN）。主要的挑战是来自其他应用程序对工作站和网络资源的竞争，不同系统速度引起的负载不平衡，以及异构性。其目的是设计和构建尽可能透明的系统。苏查系统将需要新的方法来获得动态共享和在线处理器容量信息，新的负载平衡和线程迁移策略来最小化动态环境中的通信和负载不平衡，以及易于使用的API，它将提供解决异构性所需的类型信息。这项工作的最终结果将是一组技术、策略和代码，这些技术、策略和代码将允许共享内存应用程序透明地跨一组非专用工作站迁移。在允许其他未使用的资源被利用这个新的应用领域，这项研究将显着扩大增量部署的工作站集的价值。

The focus of this research is on developing techniques and algorithms thatwill allow idle workstations to be opportunistically exploited for paralleland distributed shared-memory applications. The target environments rangefrom local-area networks (LAN's) to wide-area networks (WAN's). The centralchallenges are competition from other applications for workstation andnetwork resources, load imbalance arising from differing system speeds, andheterogeneity.The intent is to design and build as transparent a system as possible. Sucha system will need new methods of deriving dynamic sharing and processorcapacity information online, new load-balancing and thread-migrationheuristics that minimize both communication and load imbalance in dynamicenvironments, and easy-to-use API's that will provide the type informationneeded to address heterogeneity. The end result of this work will be a setof techniques, policies, and code that will allow shared-memoryapplications to be transparently migrated across a set of non-dedicatedworkstations. In allowing otherwise-unused resources to be exploited bythis new application domain, this research will significantly expand thevalue of incrementally-deployed sets of workstations.