RESEARCH INITIATION AWARD: Nonmasking Fault-tolerance in Distributed Systems

研究启动奖：分布式系统中的非屏蔽容错

• 批准号: 9308640
• 负责人: Anish Arora
• 金额: $ 9万
• 依托单位: Ohio State University Research Foundation -DO NOT USE
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1993
• 资助国家: 美国
• 起止时间: 1993-07-15 至 1997-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9308640&HistoricalAwards=false
• 关键词: RESEARCH; INITIATION; AWARD; Nonmasking; Fault

9398640 Arora This project addresses the design, implementation and application of distributed systems whose input-output relation is violated only temporarily when faults occur. Such "nonmasking" fault- tolerance is often desirable in distributed systems for reasons of efficiency, computability, or application. One goal of the research is to demonstrate that a certain nonmasking fault- tolerant system, namely a stabilizing reset subsystem, can be implemented efficiently and utilized effectively in a wide variety of distributed environments. To this end, (i) software will be developed that augments any given distributed system with a stabilizing reset subsystem, (ii) tools will be implemented that measure the efficiency of nonmasking fault-tolerance, and (iii) applications will be demonstrated where the software is useful. Another goal of the research is to demonstrate that nonmasking fault-tolerant distributed systems can be designed in a uniform and rigorous manner. To this end, methods will be developed for (i) system design based on distributed constraint satisfaction, (ii) systems design based on stepwise refinements that preserve nonmasking fault-tolerance, and (iii) reasoning about nonmasking fault-tolerance in the presence of real-time actions and constraints. The research will build upon the investigator's prior work on a uniform and rigorous framework for reasoning about fault- tolerance. ***

9398640 ARORA此项目解决了分布式系统的设计，实现和应用，其输入输出关系仅在发生故障时暂时违反。出于效率，可计算性或应用的原因，这种“非卸载”可容忍度通常在分布式系统中是可取的。该研究的一个目标是证明，可以在各种分布式环境中有效地实施某种非卸载容耐耐受性系统，即稳定的重置子系统。为此，（i）将开发软件，以增强具有稳定重置子系统的任何给定的分布式系统，（ii）将实施工具，以衡量不兑现故障耐受性的效率，并且（iii）应用程序将在软件有用的情况下证明。 该研究的另一个目的是证明可以统一和严格的方式设计不熔化故障分布式系统。为此，将为（i）基于分布式约束满意度的系统设计开发方法，（ii）基于逐步改进的系统设计，以保留不承担性能的耐受性耐受性，以及（iii）在实时操作和约束的情况下对不掩盖断层耐受性的推理。 这项研究将基于调查人员先前在统一而严格的框架上为可容忍推理的工作。 ***