Last update:
Apr 1, 2000
|
|
Claude Pruneau1,
Jeff Porter2,
Mei-li Chen2,
Torre Wenaus2
- Wayne State University
- BNL
Speaker:
Claude Pruneau
The Star Online system is designed to operate top-level
configuration, control, and monitoring of the Star experiment.
It involves components for subsystem control, configuration, run control,
monitoring, and condition logging of the detector status and conditions.
It also includes an event pool service that feeds directly from the data
acquisition event builder, for constant monitoring of the Star detector
performance based on analyses of raw detector data as well as a full
reconstruction of collisions. The Online system is based on
heteregeneous architecture and computing languages. Computing
platforms include a small cluster of Microsoft NT workstations
for detector subsystems and run control, SUN Solaris machines used as
file, database, and event pool servers, a small farm of Pentium III machines, running LINUX, for
the online analysis of live data. Connection and communication between the various processes
running on these machines are accomplished using CORBA services. The core experiment control
components consist of reactive servers written in C++. A graphical user interface, called NeXus,
and implemented in JAVA, provides modular client control of the servers. A command line interface
implemented with ROOT is also provided for additional flexibility. The Online system implements
data storage and archival using XML and mySql databases.
Presentation: |
|