CERN Accelerating science

Experiments that are held at the LHC at CERN, the European Organization for Nuclear Research, are known for a huge amount of data which are recorded at a very high rate. Moreover, the experiments are required to maintain a high data taking efficiency while the beam at the Large Hadron Collider (LHC) is present. This requires dedicated tools which enable to monitor online the quality of the recorded data, allowing for rapid interventions in case of any problems. Such actions may be for example the need for recalibration of certain detectors, exclusion of individual detectors from data taking, and many more. The goal of this work is to develop a concept of streaming data generated by the ALICE detector in real time to end users. The scope of the project is to investigate and create a prototype (a set of computer programs), demonstrating the advantages of the proposed solution over the existing framework. Therefore, the tasks required to accomplish the goals set for this project are the following: survey of existing computing infrastructure and QA solutions, preparing requirements for extension of the infrastructure for the new solution, implementation of the newly developed framework and preparation of libraries and API for end-users. The proposed solution is based on the use of proxy servers. The ZeroMQ framework was used for data transmission, which allows the use of different environments and programming languages on individual machines. Prototypes use multithreaded architecture and were written in C / C ++. The work includes short performance tests of the resulting solution. The target environment for the created prototype will be the so-called "Run 3" – the next running period of the LHC, which will start in the first quarter of the year 2021. The test environment is "Run 2", which has been operating until now, which was launched in the emulation mode on machines at the Warsaw University of Technology. The proposed solution can be easily used in a similar physical experiment, as well as in other areas of the industry, wherever real-time data transfer between different platforms is required using various programming languages.