Dynamic configuration of the computing nodes of the ALICE $O^2$ system

摘要

The ALICE (A Large Ion Collider Experiment) Collaboration is preparing major upgrades for the detectors in 2020 in order to take advantage of the increased LHC Pb-Pb collision rate of up to 50 kHz. Together with these upgrades, the ALICE Online and Offline computing systems are being redesigned and upgraded to a new common system called O2. The system is made of a software framework and a computing facility. The concept of the framework consists of online reconstruction of the data from the ALICE detector and archival of the reconstructed data from the O2 system to permanent data storage. The design of the Control, Configuration, and Monitoring (CCM) is one of the key components of the O2 system. This paper focuses on the configuration module that will allow dynamic configuration of processes and environment parameters during runtime. Based on the current status of each machine used as a computing node, parameters can be tuned either manually or automatically in order to achieve optimum performance, utilization and efficiency of the O2 system.