BNL, Upton, Long Island, New York, USA
Brookhaven National Laboratory (BNL) is one of the premier high energy and nuclear physics laboratories in the world and has been a leader in accelerator based physics research for well over half a century. For the past ten years experiments at the Relativistic Heavy Ion Collider (RHIC) have recorded data from collisions of heavy ions and polarized protons, leading to major discoveries in nuclear physics and the spin dynamics of quarks and gluons. BNL is also the site of one of the oldest alternating gradient synchrotrons, the AGS, which first operated in 1960. The accelerator controls systems for these instruments span multiple generations of technologies. In this report we will describe the current status of the Collider-Accelerator Department controls systems, which are used to control seven different accelerator facilities (from the LINAC and Tandem van de Graafs to RHIC) and multiple science programs (high energy nuclear physics, high energy polarized proton physics, NASA programs, isotope production, and multiple accelerator research and development projects). We will describe the status of current projects, such as the just completed Electron Beam Ion Source (EBIS), our R&D programs in superconducting RF and an Energy Recovery LINAC (ERL), innovations in feedback systems and bunched beam stochastic cooling at RHIC, and plans for future controls system developments.
BNL, Upton, Long Island, New York, USA
The power and scope of modern Control Systems has led to an increased amount of data being collected and stored, including data collected at high (kHz) frequencies. One consequence is that users now routinely make data requests that can cause gigabytes of data to be read and displayed. Given that a users patience can be measured in seconds, this can be quite a technical challenge. This paper explores one possible solution to this problem - the creation of remote data servers whose performance is optimized to handle context-sensitive data requests. Methods for increasing data delivery performance include the use of high speed network connections between the stored data and the data servers, smart caching of frequently used data, and the culling of data delivered as determined by the context of the data request. This paper describes decisions made when constructing these servers and compares data retrieval performance by clients that use or do not use an intermediate data server.
