Globus Toolkit Support for Distributed Data—Intensive Science

In high-energy physics,terabyte and soon petabyte-scale data collections are emerging as critical community resources.A new class of “Data Grid“ infrastructure is required to support distributed access to and analysis of these datasets by potentially thousands of users.Data Grid technology is being deployed in numerous experiments through collaborationssuch as the EU DataGrid,the Grid Physics Network,and the Particel Physics Data Grid[1],The Globus Toolkit is a widely used set of services designed to support the creation of these Grid infrastrunctures and applications,In this paper we survey the Globus technologies that will play a major role in the development and deployment for these Grids.     

Design and Evaluation of Dynamic Replication Strategies for a High—Performance Data Grid

Physics experiments that generate large amounts of data need to be able to share it with researchers around the world .High performance grids facilitate the distribution of such data to geographically remote places.Dynamic replication can be used as a technique to reduce bandwidth consumption and access latency in accessuing these huge amounts of data.We describe a simulation framework that we have developed to model a grid scenario,which enables comparative studies of alternative dynamic replication strategies.We present preliminary results obtained with this simulator,in which we evaluate the performance of six different replication strategies for three different kinds of access patterns.The simulation results show that the best strategy has significant savings in latency and bandwidth consumption if the access patterns contain a moderate amount of gerographical locality.     

Simulation Studies in Data Replication Strategies

The aim of this work is to present the simulation studies in evaluating different data replication strategies between Regional Centers.The simulation Framework developed within the “Models of Networked Analysis at Rgional Centers”(MONARC) project,as a design and optimization tool for large scale distributed systems,has been used for these modeling studies.Remote client-serer access to database servers as well as ftp-like data transfers have been ralistically simulated and the performance and limitations are presented as a function of the characteristics of the protocol used and the network parameters.     

StureGate：a Data Model for the ATLAS Software Architecture

ATLAS[1] has recently joined Gaudi,an open project to develop a data processing framework for HEP experiments[2],The data model is one of the areas where ATLAS has extended more the original Gaudi design to meet the experiment‘s own requirments.This paper describes StoreGate,the first implementation of the ATLAS Data Model.     

A Generic Digitization Framework for the CDF Simulation

Digitization from GEANT tracking requires a predictable sequence of steps to produce raw simulated detector readout information.We have developed a software framework that simplifies the development and integration of digitizers by separating the coordination activities(sequencing and dispatching)from the actual digitization process.This separation allows the developers of digitizers to concentrate on digitization.The framework provides the sequencing infrastructure and a digitizer model,which means that all digitizers are required to follow the same sequencing rules and provide an interface that fits the model.     

Nuclear magnetic moments for J~π=2_1~+ state of ~(10)Be with ab initio Monte Carlo shell model calculation

The magnetic moment of 2+1 state for 10Be are calculated and investigated in terms of single particle orbits for protons and neutrons under the framework of ab initio Monte Carlo shell model method in an emax = 3 model space. The reduced matrix elements of orbital and spin angular momentum are evaluated. It is found that the orientations of orbital angular momentum in diferent single particle orbits are consistent. Conversely,the orientations of spin in diferent single particle orbits tend to be chaotic. The nuclear magnetic moment of 2+1 state for 10Be is obtained as 1.006N and is discussed in regards to the contribution of orbital and spin angular momentum both for protons and neutrons. The corresponding g-factor is also given.     

Improvement of mobility edge model by using new density of states with exponential tail for organic diode

The mobility edge(ME) model with single Gaussian density of states(DOS) is simplified based on the recent experimental results about the Einstein relationship. The free holes are treated as being non-degenerate, and the trapped holes are dealt with as being degenerate. This enables the integral for the trapped holes to be easily realized in a program. The J–V curves are obtained through solving drift-diffusion equations. When this model is applied to four organic diodes, an obvious deviation between theoretical curves and experimental data is observed. In order to solve this problem, a new DOS with exponential tail is proposed. The results show that the consistence between J–V curves and experimental data based on a new DOS is far better than that based on the Gaussian DOS. The variation of extracted mobility with temperature can be well described by the Arrhenius relationship.     

Querying Large Physics Data Sets Over an Information Grid

Optimising use of the Web(WWW) for LHC data analysis is a complex problem and illustrates the challenges arising from the integration of and computation across massive ampunts of information distributed worldwide.Finding the right piece of information can,at times,be extremely time-consuming,if not impossible,SO-called Grids have been proposed to facilitate LHC computing and many groups have embarked on studies of data replication,data migration and netwroking plhilosophies.Other aspects such as the role of moddleware‘ for Grids are emerging as requiring research.This paper positions the need for appropriate middleware that enables users to resolve physics queries across massive data sets.It identifies the role of meta-data for query resolution and the importance of Information Grids for high-energy physics analysis rather than just computational or Data Grids,This paper identifies software that is being implemented at CERN to enable the querying of very large collaborating HEP data-sets,initially being employed for the construction of CMS detectors.     

A Distributed Agent—based Architecture for Dynamic Services

Aprototype system for agent-based distributed dynamic services that will be applied to the development of Data Grids for high-energy physics is presented.The agent-based systems we are designing and develogping gather,disseminate and coordinate configuration ,time-dependent state and other information in the Grid system as a whole.These systems are being developed as an enabling technology for workflow-management and other forms of end-to-end Grid system monitoring and management.This prototype is being developed in Java and is based on the JINI,Mobile Agents,Self-Organizing Neural Networks.     

Evolution analysis of the states of the EZ model

Based on suitable choice of states, this paper studies the stability of the equilibrium state of the EZ model by regarding the evolution of the EZ model as a Markov chain and by showing that the Markov chain is ergodic. The Markov analysis is applied to the EZ model with small number of agents, the exact equilibrium state for N = 5 and numerical results for N=18 are obtained.     

Two-Dimensional Self-Propelled Fish Motion in Medium： An Integrated Method for Deforming Body Dynamics and Unsteady Fluid Dynamics

We present （1） the dynamical equations of deforming body and （2） an integrated method for deforming body dynamics and unsteady fluid dynamics, to investigate a modelled freely self-propelled fish. The theoretical model and practical method is applicable for studies on the general mechanics of animal locomotion such as flying in air and swimming in water, particularly of free self-propulsion. The present results behave more credibly than the previous numerical studies and are close to the experimental results, and the aligned vortices pattern is discovered in cruising swimming.     

CDF Run Ⅱ Data File Catalog

The CDF experiment started data taking in April 2001,The data are organized into datasets which contain events of similar physics properties and reconstruction version.the information about datasets is stored in the Data File Catalog,a relational database.This information is presented to the data processing framework as objects which are retrieved using compound keys.The objects and the keys are designed to be the algorithms‘ view of information stored in the database.Objects may use several DB tables.A database interface management layer exists for the purpose of managing the mapping of persistent data to transient objects that can be used by the framework.This layer exists between the algorithm code and the code which reads directly from datanbase tables.At the user end,it places get/put interface on a top of a transient class for retrieval or storage of objects of this class using a key.Data File Catalog code makes use of this facility and contains all the code needed to manipulate CDF Data File Catalog from a C program or from the command prompt,It supports an Oracle interface using OTL,and a mSQL interface,This code and the Oravcle implementation of Data File Catalog were subjected to test during CDF Commissioning Run last fall and during first weeks of Run II in April.It performed exceptionally well.     

Dynamical study of light scalar mesons below 1 GeV in a flux-tube model

The light scalar mesons below 1GeV configured as tetraquark systems are studied in the framework of the flux-tube model. Comparative studies indicate that a multi-body confinement,instead of the additive two-body confinement, should be used in a multiquark system.The σ and κ mesons could be well accommodated in the diquark-antidiquark tetraquark picture, and could be colour-confinement resonances. The a0(980) and fo(980) mesons are not described as KK molecular states and ns diquark-antidiquark states.However, the mass of the first radial excited state of the diquark-antidiquark state, nn is 1019 MeV,is close to the experimental data of the fo (980).     

Analytical Solution for the SU（2） Hedgehog Skyrmion and Static Properties of Nucleons

An analytical solution for symmetric Skyrmion is proposed for the SU（2） Skyrme model, which takes the form of the hybrid form of a kink-like solution, given by the instanton method. The static properties of nucleons is then computed within the framework of collective quantization of the Skyrme model, in a good agreement with that given by the exact numeric solution. The comparisons with the previous results as well as the experimental values are also presented.     

Charged pion condensation in anti-parallel electromagnetic fields and nonzero isospin density

The formation of charged pion condensate in anti-parallel electromagnetic fields and in the presence of the isospin chemical potential is studied in the two-flavor Nambu-Jona-Lasinio model.The method of Schwinger proper time is extended to explore the quantities in the off-diagonal flavor space,i.e.the charged pion.In this framework,π^± are treated as bound states of quarks and not as point-like charged particles.The isospin chemical potential plays the role of a trigger for charged pion condensation.We obtain the associated effective potential as a function of the strength of the electromagnetic fields and find that it contains a sextic term which possibly induces a weak first order phase transition.The dependence of pion condensation on model parameters is investigated.     

Mechanisms of Avalanche Dynamics in a Stochastic Four-State Sandpile Model

We study the stochastic four-state sandpile model on the square lattice. The static and dynamical properties of the model are investigated and compared with the deterministic sandpile model of Bak, Tang, and Wiesenfeld [Phys. Rev. Lett. 59 (1987) 381]. The numerical results show that the stochastic model defines a new universality class with respect to the deterministic sandpile. We also find that the waves in an avalanche are uncorrelated in the stochastic model (in the BTW model, the waves in an avalanche are correlated). The physical origin of the critical behaviour of the stochastic model being different from that of the BTW model is ascribed to the ordering and deterministic property of the toppling law in the BTW model.     

Projected Shell Model Analysis of Multi—Quasiparticle High—K Isomers in ^174Hf

Multi-quasiparticle high-K states in ^174Hf are studied in the framework of the projected shell model.The calculation reproduces well the observed ground-state band as well as most of the two-and four-quasiparticle rotational bands.Some as yet unobserved high-K isomeric states in ^174Hf are predicted.Possible reasons for the existing discrepancies between calculation and experiment are discussed.It is suggested that the projected shell model may be a useful method for studying multi-quasiparticle high-K isomers and the K-mixing phenomenon in heavy deformed nuclei.     

Tuning and Validation of the Lundcharm Model with J/ψ Decays

Using J/ψ decays collected with the BES Ⅲ detector, Lundcharm model parameters are optimized with J/ψlight hadron decays. The dependence of the response function on model parameters is approximated up to the quadratic term, and the model parameters are optimized by simultaneously fitting J/ψ inclusive charged track distributions and event shapes. The Monte Carlo simulations show that optimal parameters yield satisfactory MC distributions as compared to both the J/ψ and ψ（2S） data distributions. These optimal values are suggested for the Lundcharm model to produce J/ψ and ψ（2S） decays to light hadrons.     

Fermilab Distributed Monitoring System（NGOP）

A Distributed Monitoring System(NGOP)that will scale to the anticipated requirements for RUn II computing has been under development at Fermilab.NGOP [1] provides a framework to create Monitoring Agents for monitoring the overall state of computers and software that are running on them.Several Monitoring Agents are available within NGOP that are capable of analyzing log files,and checking existence of system daemons,CPU and memory utilization,etc,NGOP also provides customizable graphical hierarchical representations of these monitored systems.NGOP is able to generate events when serious problems have occurred as well as raising alarms when potential problems have been detected.NGOP allows performing correctiv actions or sending notifications,NGOP provides persistent storage for collected events,alarms and actions.A first implementation of NGOP was recently deployed at Fermilab.This is a fully functional prototype that satisfies most of the existing requirements.For the time being the NGOP prototype is monitoring 512 nodes.During the first few months of running NGOP has proved to be a useful tool.Multiple problems such as node resets,offline CPUs,and dead system daemons have been detected.NGOP provided system administrators with information required for better system tuning and configuration.The current state of deployment and future steps to improve the prototype and to implement some new features will be presented.