John Heidemann

Defining and Measuring Conflicts in Optimistic Replication

TitleDefining and Measuring Conflicts in Optimistic Replication
Publication TypeTechnical Report
Year of Publication1995
AuthorsJ. Heidemann, A. Goel, and G. Popek
Date Publishedsep

Optimistic replication is often viewed as essential for large scale systems and for supporting mobile computing. In optimistic replication, updates can be made concurrently to different file replicas, resulting in multiple versions of the file. To recover from these \emphconflicting updates, after-the fact \emphconflict resolution actions are required to recombine multiple versions into one. This paper defines these concepts and discusses approaches to measure them in optimistically replicated systems. Measurement of the number of conflicting updates and conflict resolution is important to judge the practicality of optimistic replication. An environment where conflicting updates are frequent will not be attractive since users cannot assume they have up-to-date data. Although many conflicts can be automatically resolved, some conflicts require user intervention; such conflicts cannot be too common. This paper shows an approach to measure the number of conflicting updates. From this measurement we derive the actual amount of work done by the user or system to resolve conflicts and the minimum amount of work required to resolve conflicts.