   |  |
Computing Technical Design Report
| | |
Computing Technical Design Report
The primary purpose of this chapter is to present the current ideas on the steady-state ATLAS offline Computing Model and to detail the tests performed to validate that model. The current best estimate of the resources implied is given in Chapter 7. The model extends from the primary event store, which is where events selected by the trigger are recorded, to the analyst at a remote university. Consideration is also given to computing issues in the DAQ and the High Level Trigger (HLT) farm, that is prior to the primary event store.
Also considered is the model for the commissioning of the computing system with real data. This will require enhanced access to raw and nearly-raw data for calibration, algorithm development etc, with resultant implications for the resource providers.
The ideas and estimates presented here have evolved from previous studies, including the ATLAS computing resource and cost estimates developed using the MONARC hierarchical model [2-1] for the CERN LHC Computing Review [2-2]. The advent of the World Wide Grid triggered new ideas on how to organize the ATLAS system as a virtual worldwide distributed computing facility. This was presented to a subsequent CERN LHC Computing Review [2-3].
The main requirements on the Computing Model are to enable all members of the ATLAS Collaboration speedy access to all reconstructed data for analysis during the data-taking period, and appropriate access to raw data for organized monitoring, calibration and alignment activities. The model presented here makes substantial use of Grid Computing concepts, thereby allowing the same level of data access, and making available the same amount of computing resources, to all members of the ATLAS Collaboration.
The Computing Model embraces the Grid paradigm and a high degree of decentralization and sharing of computing resources. However, as different computer facilities are better suited to different roles, a degree of hierarchy, with distinct roles at each level, remains. This should not obscure the fact that all of the roles described are vital and must receive due weight. The required level of computing resources means that off-site facilities will be vital to the operation of ATLAS in a way that was not the case for previous CERN-based experiments.
The primary event processing occurs at CERN in a Tier-0 Facility. The RAW data is archived at CERN and copied (along with the primary processed data) to the Tier-1 facilities around the world. These facilities archive the RAW data, provide the reprocessing capacity, provide access to the various processed versions and allow scheduled analysis of the processed data by physics analysis groups. Derived datasets produced by the physics groups are copied to the Tier-2 facilities for further analysis. The Tier-2 facilities also provide the simulation capacity for the experiment, with the simulated data housed at Tier-1s. In addition, Tier-2 centres will provide analysis facilities and some will provide the capacity to produce calibrations based on processing some raw data. A CERN Analysis Facility provides an additional analysis capacity, with an important role in the data-intensive calibration and algorithmic development work.
ATLAS will negotiate relationships between Tier-1s and Tier-2s, and also among Tier-1s themselves, to try to optimize the smooth running of the system in terms of data transfer, balanced storage and network topologies. It is not assumed that all Tier-1s or Tier-2s will be of the same size. However, the ratio of disk, tape and CPU resources required is the same in each case.
| | |
Copyright © CERN 2005