Proactive detection and repair of data corruption: towards a hassle-free declarative approach with Amulet

AbstractOccasional corruption of stored data is an unfortunate byproduct of the complexity of modern systems. Hardware errors, software bugs, and mistakes by human administrators can corrupt important sources of data. The dominant practice to deal with data corruption today involves administrators writing ad hoc scripts that run data-integrity tests at the application, database, file-system, and storage levels. This manual approach, apart from being tedious and error-prone, provides no understanding of the potential system unavailability and data loss if a corruption were to occur. We have developed the Amulet system that can verify the correctness of stored data proactively and continuously. This demonstration focuses on the uses of Amulet and its technical innovations: (i) a declarative language for administrators to specify their objectives regarding the detection and repair of data corruption; (ii) optimization and execution algorithms to meet the administrator's objectives robustly and with least cost using pay-as-you-go cloud resources; and (iii) timely notification when corruption is detected, allowing proactive repair of corruption before it impacts users and applications.