Speaker
Description
The Trigger and Data Acquisition (TDAQ) system of the ATLAS experiment the Large Hadron Collider (LHC) at CERN currently is composed of a large number of distributed hardware and software components (about 3000 machines and more than 25000 applications) which, in a coordinated manner, provide the data-taking functionality of the overall system.
During data taking runs, a huge flow of operational data is produced in order to constantly monitor the system and allow proper detection of anomalies or misbehaviors. The Persistent Back-End for the ATLAS Information System of TDAQ (P-BEAST) is a system based on a custom-built time-series database and it is used to archive and retrieve for applications any operational monitoring data. P-BEAST stores about 18 TB of highly compacted and compressed raw monitoring data per year acquired at 200 KHz average information update rate during ATLAS data taking periods.
Since P-BEAST has been put into production, 4 years ago, several promising database technologies for fast access to time-series and column-oriented data have become available. InfluxDB and ClickHouse were the most promising candidates for improving the performance and functionality of the current implementation of P-BEAST.
This paper presents a short description of main features of both technologies and a description of the synthetic tests ran on both database systems which try to leverage the best possible options for storage of the P-BEAST data using their respective data model capabilities. Then, the results of the performance testing that has been performed using a subset of archived ATLAS operational monitoring data are presented. Finally, a comparison of the results is presented.
