Online and Offline Computing systems in the PHENIX experiment

TL;DR

The paper describes the upgrades to the PHENIX experiment's computing systems, enhancing data acquisition, processing speed, and infrastructure to improve real-time and offline data analysis at RHIC.

Contribution

It provides a comprehensive overview of the hardware and software upgrades implemented in PHENIX's online and offline computing systems.

Findings

Data rate increased to 100 MB/s

Fast offline reconstruction achieved within weeks

Successful transition from Solaris to Linux servers

Abstract

PHENIX is one of two large experiments at Brookhaven National Laboratory's Relativistic Heavy-Ion Collider (RHIC). At the time of the conference, the PHENIX experiment was about halfway through the 2003 run, which started in January 2003. In preparation for the run, the PHENIX data acquisition, the computing infrastructure, and the software have undergone several upgrades. Those upgrades boost the recorded data rate to about 100 MB/s and allow for a fast reconstruction only a few weeks after the data have been taken. As part of the upgrade, essentially all servers in the Online System have been converted from Solaris to Linux, and a new Linux computing farm has been commissioned at the experimental site that is used to prepare for a rapid offline reconstruction pass. This paper presents a general overview of PHENIX computing. We will explain the current status, the changes, choices of software and hardware, and discuss our experience with the new setup.