go directly to the contents
EUChinaGRID

application repository

  • name: ARGO
  • domain: Astroparticle Physics
  • country: Italy-China
  • author:
  • institute:
  • contacts:
  • description: The ARGO detector will provide a detailed space-time picture of the showers front, initiated by primaries of energies in the range 10GeV- 500 TeV. ARGO-YBJ telescope is optimised for the detection of small size air showers. Data gathered with ARGO will face the gamma astronomy at 100GeV threshold, allowing the monitoring of more than 300 sources, gamma ray burst physics, the study of diffuse gamma ray from galactic plane, the Anti-p/p ratio at energies 300GeV- TeV, the primary proton spectrum in the 10-200 TeV region and many other different physics items.
  • functionalities: The ARGO application in EUChinaGRID focuses on data management (data transfer,  backup and data processing) and is covered by 2 modules:
    • Data Transfer: the ARGO experimental data are collected at rates up to 7.5 Mbyte/sec and should be available to reconstruction and data analysis process as fast as possible.
    • Data Processing: reconstruction algorithms applied to space-time information of the showers front identify the primaries, their direction and characteristics of the showers. To the same module belongs the ARGO MonteCarlo production, that simulates the showers and the respond of the RPC carpet.
    • middleware requirements:
      • Mirrored data catalogues - The experiment requirements imply the presence of a copy of the raw experimental data files and of the processed data files in the Storage Element of each computing centre and inside the data catalogues. These data files should be managed using authomatically mirrored data catalogues, instead of doing that by proprietary scripts.

      • Mirrored metadata catalogues - The data files are logically grouped in RUNs and some characteristics are valid for all the files in the group. This information is kept in metadata catalogues, which should also be automatically mirrored as are the data catalogues maintaining data files information.
      • Shared jobs control - The control of computation process should be distributed to several persons. This require the possibility to control the access rights to jobs (list access right, cancellation and get-output access right) for different users of a same VO. [See also: EGEE PTF #100809. (Status: none)]
      • Job listing for a user - The problem of monitoring the current situation of computational process is very important and loosing the information about the job, if its ID is lost, is not acceptable. Therefore, a user should be enable to list submitted jobs, even if the jobs Ids is not known. All uncompleted jobs (running or output not retrieved) should be displayed. [See also #100535  (Status: satisfied)]
      • Recuperating job status - In some  cases  we can loose the link to the CE while the assigned jobs are still running. The automatic recuperation of the job status when the link is reactivated will avoid the resubmission of the jobs.
      • Compatibility of the applications with GRID computing environment - the programs used for MonteCarlo production, Corsika e GEANT3, where already successfully used in GRID by other experiments.
    • resources requirements:
      • The data transfer module imply the transfer over the network of 200 Tbyte of raw data from the experimental site to the processing sites in Italy and China.  At the end of the data processing we have to the transfer also  the reconstructed data files between these two sites, for a total amount of 36 Tbyte.  The procedures are PERL scripts.
      • The data processing module requests a computing power of 200,000 SPECint2000 only for the experimental data processing. Assuming the reuse of the simulated showers, we can esteem the requested computing resource for MonteCarlo production in at least 250,000 SPECint2000. Taking into account the possibility to share the ARGO computing resources  between the Italian and Chinese computing centers we halve the resources requested on each side and the time  needed for the reconstructed data to be available for the physics analysis.

    Back to list of applications >>

powered by
Consortium GARR