Speaker
Mr
Artem Petrosyan
(JINR)
Description
Implementation of COMPASS data processing in the distributed environment has started in 2015. Since the summer of 2017, data processing system works in production mode, distributing jobs to two traditional Grid sites: CERN and JINR. There are two storage elements, both at CERN: disk-storage EOS for short-term storage and tape-storage Castor for long-term storage. Processing management services, including MySQL server, PanDA servers, APF/Harvester server, monitoring server, and production management server, are deployed in JINR Cloud Service. Thus, the system, which manages distributed data processing of the experiment, is also distributed. The production management system is based on the principles of service-oriented architecture. Each service of the system is maximally isolated from the others, is executed independently, and usually performs only one function, for example: sends jobs, checks their statuses, archives results, and so on. During last year, the system was replenished by task archiving mechanism, FTS and Harvester services, and Monte Carlo processing chain. Status, statistics, workflow, data management, and infrastructure overview are presented in this report.
Primary author
Mr
Artem Petrosyan
(JINR)
