Mysql-Based Data Archiver: Preliminary Results

摘要

Following an evaluation of the archival requirements of the Jefferson Laboratory accelerators user community, a prototyping effort was executed to determine if an archiver based on MySQL had sufficient functionality to meet those requirements. This approach was chosen because an archiver based on a relational database enables the development effort to focus on data acquisition and management, letting the database take care of storage, indexing and data consistency. It was clear from the prototype effort that there were no performance impediments to successful implementation of a final system. With our performance concerns addressed, the lab undertook the design and development of an operational system. The system is in its operational testing phase now. This paper discusses the archiver system requirements, some of the design choices and their rationale, and presents the acquisition, storage and retrieval performance.