Abstract :
The BTeV experiment was a collider based high energy physics (HEP) B-physics experiment proposed at Fermilab. It included a large-scale, high speed trigger/data acquisition (DAQ) system, reading data off the detector at 500 Gbytes/sec and writing to mass storage at 200 Mbytes/sec. The online design was considered to be highly credible in terms of technical feasibility, schedule and cost. This paper gives an overview of the overall trigger/DAQ architecture, highlight some of the challenges, and describe the BTeV approach to solving some of the technical challenges. At the time of termination in early 2005, the experiment had just passed its baseline review. Although not fully implemented, many of the architecture choices, design, and prototype work for the online system (both trigger and DAQ) were well on their way to completion. Other large, high-speed online systems may have interest in the some of the design choices and directions of BTeV, including (a) a commodity-based tracking trigger running asynchronously at full rate, (b) the hierarchical control and fault tolerance in a large real time environment, (c) a partitioning model that supports offline processing on the online farms during idle periods with plans for dynamic load balancing, and (d) an independent parallel highway architecture
Keywords :
data acquisition; fault tolerant computing; high energy physics instrumentation computing; particle detectors; resource allocation; 500 GByte/s; B-physics experiment; BTeV trigger; collider based high energy physics; commodity-based tracking trigger; data acquisition system; dynamic load balancing; fault tolerance; hierarchical control; high-speed online systems; independent parallel highway architecture; Costs; Data acquisition; Detectors; Fault tolerant systems; Large-scale systems; Load management; Prototypes; Real time systems; Road transportation; Technological innovation;
