Speaker
Dr
Francesco Delli Paoli
(INFN Padova)
Description
The improvements of the peak instantaneous luminosity of the Tevatron Collider will
give CDF up to 2 fb-1 of new data every year, forcing the collaboration to increase
proportionally the amount of Monte Carlo data it produces. This is in turn forcing
the CDF collaboration to move beyond the dedicated resources it is using today and
start exploiting Grid resources. Monte Carlo production was the first obvious step in
that direction, since there are fewer technical problems associated with it: A portal
to the LCG was implemented with an user interface identical to the CAF (CDF Analysis
Farm) infrastructure already used for submission to dedicated CDF resources. CDF code
distribution is performed by an AFS filesystem mounted on each worker node while the
CDF calibration database is accessed through a local SQUID-based cache located at the
CNAF Tier 1 center. The job output is stored on either CDF specific fileservers via
kerberos authentication or LCG Storage Elements. We used a few stable sites to
produce several millions events for studies of B physics background and tagging
efficiencies in top measurements. The observed LCG/gLite performancies and possible
improvements will be discussed.
Primary author
Dr
Francesco Delli Paoli
(INFN Padova)
Co-authors
Dr
Armando FELLA
(INFN, Pisa)
Dr
Daniel JEANS
(INFN, Frascati)
Dr
Donatella LUCCHESI
(INFN, Padova)
Dr
Igor SFILIGOI
(INFN, Frascati)
Dr
Shih-chieh HSU
(Fermi National Accelerator Laboratory)
Dr
Subir SARKAR
(INFN, Bologna)
