Long-term archiving of relational databases with Chronos

Abstract

CSP Chronos Archiving provides solutions for many of the core issues raised by the PresDB’07 workshop. Chronos enables serious and highly scalable long-term archiving of relational databases for companies, organizations, science data centers, and public archives. The dependency of the data on original production environments is removed, but the original semantics, structure and integrity of the data is retained and preserved in an open Text/XML-based format that does not require any database management system to maintain, access and retrieve archive data. Chronos can collectively maintain any number of database archives, and provides easy web-based access to any number of users from any location. Furthermore, Chronos can be seamlessly integrated with the life cycle management of production systems: Although the database schema of a production system may evolve over time, Chronos can continuously and incrementally extract data subsets in many subsequent archiving runs and yet ensure coherent and collective accessibility and manageability for all archived data. This is possible because of Chronos’ unique ability to detect, describe and manage the semantic and structural changes in the production database schema between any two subsequent executions of the archiving process.