Raw data from optical modules (OMs) and section master modules (SMMs) are transmitted via fibre-optic cables first to the Cluster Centre, then to the Shore Station.
The Cluster Control Software processes raw data and structures it into a special Baikal-data format, categorizing it into three distinct record types: master, service, and monitor records.
Master records contain digitized data from OMs, accompanied by metadata, such as timestamps, OM addresses, trigger information, etc.
Service records encompass the information about the static and dynamic configuration of the cluster. Static-configuration data include the number of clusters, IP-addresses of telescope elements, and the arrangement of sections and strings. Dynamic-configuration data describe current run numbers and the system of triggers applied.
Monitor records consist of OM operational parameters, including PMT voltage, detection thresholds, temperature inside OMs, and signal amplitude distributions.
A single cluster generates approximately 15 Gb of data per day. Data are compressed and transmitted over the Internet to servers at the Joint Institute for Nuclear Research in Dubna, Russia. Here, data are processed by BARS (Baikal Analysis and Reconstruction Software). The BARS core is based on the MARS framework, originally developed by the MAGIC Collaboration (used with permission), which, in turn, is built on the ROOT framework.
BARS is designed to run on the Linux operating system. It comprises primarily C++ classes and ROOT macros, each engineered to perform specific tasks. These tasks include:
The output of one macro is typically a ROOT file that serves as the input for the next one, forming a chain of processing tasks. This chain constitutes a complete data processing cycle. To handle the daily influx of new data, this cycle is designed to run automatically every day. A dedicated Python-based automation system executes this daily processing workflow.
