Large scale distributed computing support

LHC experiments use geographically distributed computing centers for data processing, storage and simulations. Participating sites are categorized according capacity and level of supports in several levels Tier0 – Tier3. Institute of Physics (FZU) operates a Tier2 site supporting ATLAS and ALICE experiments. The site is steeply growing and in several years will reach a capacity of current Tier1 centers. We are now solving problems with installations and operations of significantly larger number of computing nodes and storage servers with several times higher capacity. These problems were already faced in CC IN2P3 and we want to take use of experience of the Tier1 site. Because the computer hardware develops and changes rapidly, we cannot just copy existing solutions. Both institutes participate also in the Pierre Auger Observatory. CC IN2P3 is the main computing center for the project; every collaboration member has a computer account and local access to data. FZU in collaboration with CESNET manages the Auger virtual organization for grid computing. We plan to copy results of simulations from distributed grid storage elements to CC IN2P3. This will make a backup copy and it will also enable data access for non grid users.