ISBN-13: 9783639003826 / Angielski / Miękka / 2008 / 200 str.
ISBN-13: 9783639003826 / Angielski / Miękka / 2008 / 200 str.
At the Large Hadron Collider (LHC), a large number of events containing Z bosons will be available and the calibration of the absolute jet energy scale with these events becomes feasible for the first time. Such a calibration is important as the uncertainty on this observable will be the dominant systematical uncertainty on most LHC analyses. This work demonstrates the feasibility of such a calibration within the CMS experiment through a comparison of the momentum of a jet with the precisely reconstructed momentum of a Z boson decaying into two muons. The four LHC experiments will produce about 15 petabytes of data annually which are stored and processed using grid technologies. This, however, requires additional services which do not fully load server machines at smaller grid sites. The virtualisation of operating systems enables the consolidation of different services on a few server machines. Extending this concept to batch queuing systems allows overcoming intrinsic limitations of shared computing infrastructures. Both, the consolidation of an institute's computing and grid infrastructure and a prototype implementation of a virtualised batch queuing systems are presented.
At the Large Hadron Collider (LHC), a large number of events containing Z bosons will be available and the calibration of the absolute jet energy scale with these events becomes feasible for the first time. Such a calibration is important as the uncertainty on this observable will be the dominant systematical uncertainty on most LHC analyses. This work demonstrates the feasibility of such a calibration within the CMS experiment through a comparison of the momentum of a jet with the precisely reconstructed momentum of a Z boson decaying into two muons. The four LHC experiments will produce about 15 petabytes of data annually which are stored and processed using grid technologies. This, however, requires additional services which do not fully load server machines at smaller grid sites. The virtualisation of operating systems enables the consolidation of different services on a few server machines. Extending this concept to batch queuing systems allows overcoming intrinsic limitations of shared computing infrastructures. Both, the consolidation of an institutes computing and grid infrastructure and a prototype implementation of a virtualised batch queuing systems are presented.