Paper detail

A Relational Database Model for Managing Accelerator Control System Software At Jefferson Lab

The operations software group at the Thomas Jefferson National Accelerator Facility faces a number of challenges common to facilities managing a large body of software developed in-house. Developers include members of the software group, operators, hardware engineers and accelerator physicists. One management problem has been ensuring that all software has an identified maintainer who is still working at the lab. In some cases, locating source code for 'orphaned' software has also proven to be difficult. Other challenges include enforcing minimal standards for versioning and documentation, segregating test software from operational software, encouraging better code reuse, consolidating input/output file storage and management, and tracking software dependencies. This paper will describe a relational database model for tracking the information necessary to solve the problems above. The instantiation of that database model provides the foundation for various productivity- and consistency- enhancing tools for automated (or at least assisted) building, versioning, documenting and installation of software.