The ALICE experiment is the heavy-ion detector designed to study the physics of strongly interacting matter and the quark-gluon plasma at the LHC. After 10 years of operations, a new computing system called O2 (Online-Offline) is currently under development and is scheduled for production in 2021. Built on a microservices architecture and comprised of +100k processes running on 2000 nodes, it will require highly integrated infrastructure services in order to be operated by a small shift crew with high efficiency and limited training. To control the quality of the physics data being collected, and to ensure an in-memory data reduction by a factor of 35, several Machine Learning techniques are currently under investigation. This talk will present some of these topics and the ongoing work in the context of the O2 project.
About the speaker:
Vasco Barroso joined CERN in 2005 as a Software Engineer, first in the Information Technology department and later in the Experimental Physics department. He led the development of the bookkeeping facility for the ALICE experiment and later moved to the topics of control, configuration and monitoring. Currently he leads the design and development of the infrastructure services for the new ALICE Online-Offline computing farm.