Abstract
In the LHC operations era, analysis of the multi-petabyte ATLAS data sample by globally distributed physicists is a challenging task. To attain the required scale the ATLAS Computing Model was designed around the concept of Grid computing, realized in the Worldwide LHC Computing Grid (WLCG), the largest distributed computational resource existing in the sciences. The ATLAS experiment currently stores over 140 PB of data and runs about 140,000 concurrent jobs continuously at WLCG sites. During the first run of the LHC, the ATLAS Distributed Analysis (DA) service has operated stably and scaled as planned. More than 1600 users submitted jobs in 2012, with 2 million or more analysis jobs per week, peaking at about a million jobs per day. The system dynamically distributes popular data to expedite processing and maximally utilize resources. The reliability of the DA service is high and steadily improving; Grid sites are continually validated against a set of standard tests, and a dedicated team of expert shifters provides user support and communicates user problems to the sites. Both the user support techniques and the direct feedback of users have been effective in improving the success rate and user experience when utilizing the distributed computing environment. In this contribution a description of the main components, activities and achievements of ATLAS distributed analysis is given. Several future improvements being undertaken will be described.
Highlights
During the first LHC run, the ATLAS experiment [1] collected more than 140 PB of data
More than 400M analysis jobs were executed in the time period January 2011-August 2013, for a daily average of half a million jobs
Based on Run 1 experiences, ATLAS is planning several improvements in its computing model to cope with the increased amount of data expected during LHC Run 2
Summary
During the first LHC run, the ATLAS experiment [1] collected more than 140 PB of data. 2. Distributed Analysis activities during LHC Run 1 The recommended way for users to access distributed ATLAS resources is through the PanDA workload management system [4]. The output of the analysis jobs is temporarily stored on the scratch disks of the sites, from which it can be retrieved by the users using DQ2 client tools, or automatically transferred to group space disks. More than 1600 users submitted jobs during the first run of the LHC.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
