Dr
Helge Meinhard
(CERN-IT)
26/10/2009, 10:30
Site Reports
Summary of important news at CERN since the spring 2009 meeting
Sandy Philpott
(JLAB)
26/10/2009, 10:50
Site Reports
Status of Scientific Computing at JLab, including experimental physics and high performance computing for Lattice QCD.
Mr
philippe olivero
(CC-IN2P3)
26/10/2009, 11:30
Site Reports
News, changes, upgrades occured last year at CC-IN2P3.
Jeffrey Altman
(Your File System Inc.)
26/10/2009, 13:15
Storage
The U.S. Department of Energy has awarded Your File System Inc. a US$648,000 Small Business Innovative Research Phase II grant to support the development of a next generation globally distributed file system that is compatible with AFS. This talk will describe the technologies that Your File System Inc. will be implementing and contributing to OpenAFS through August 2011.
Christopher J Walker
(Queen Mary, University of London)
26/10/2009, 13:45
Storage
Queen Mary, University of London has been using the StoRM SRM in front
of a Lustre filesystem. We present the results of benchmarks on the Lustre filesystem, and the throughput from simulated analysis using the hammercloud framework.
Walter Schoen
(GSI)
26/10/2009, 14:15
Storage
lustre@gsi: A Petabyte file system for the analysis farm - status and outlook
Artem Trunov
(Karlsruhe Institute of Technology)
26/10/2009, 14:45
Storage
This work is continuation of storage solution testing performed by HEPiX storage workgroup on it's testbed at FZK. Hadoop, an Apache project, offers a cluster file system called HDFS inspired by Google File System and designed to run on commodity hardware. It has gained some popularity in OSG, where is has become a supported storage solution, and is currently in production at a few T2 sites....
Dr
Ulrich Schwickerath
(CERN)
26/10/2009, 15:30
Virtualization
Between March and August 2009 a project has been set up at CERN with
the aim to evaluate possibilities to use virtualization at a large scale, with
the focus on batch computing. Two key issues have been identified for this specific
application: the placing of virtual machines on an appropriate hyper-visor, and the
selection of an appropriate image which should be driven by the actual...
Dr
Chadwick Keith
(Fermilab)
26/10/2009, 16:00
Virtualization
The current virtualization infrastructure in use within FermiGrid and the operational experience will be presented.
Mr
Brent Draney
(NERSC/LBNL)
26/10/2009, 16:30
Virtualization
NERSC and the Argonne LCF have been funded by DOE to acquire test systems to explore cloud computing technologies. We present an overview of the Cloud Computing Project at NERSC.
Mr
Jay Srinivasan
(Lawrence Berkeley National Lab. (LBNL)-Unknown-Unknown)
27/10/2009, 09:30
Site Reports
We present the current status of PDSF and updates since the last HEPiX meeting.
Dr
Douglas McNab
(University of Glasgow)
27/10/2009, 09:45
Site Reports
This presentation will provide an overview of the very successful UK ATLAS Tier-2, ScotGrid, as we fast approach LHC data taking and data analysis. This will cover a variety of topics, ranging from an overview of the fabric and middleware, the site's current readiness, success during STEP, storage issues, site optimisations all the way through to disaster planning and site security. The...
Michele Michelotto
(Univ. + INFN)
27/10/2009, 10:45
Benchmarking
Performances of the last intel DP processor Nehalem 55xx and last amd DP processor Instanbul 24xx using the HEP-SPEC06 benchmark.
Dr
Ulrich Schwickerath
(CERN)
27/10/2009, 11:15
Benchmarking
The recent generation of Intel XEON CPUs comes with support for
symmetric multi processing, formally known as hyperthreading. In
addition, a new CPU feature has been added, the Intel Turbo mode.
The influence of these features on the system performance has been
tested using the HEPSPEC06 benchmark suite, with enhanced statistics,
and a study of the scaling behaviour has been done.
The...
Helge Meinhard
(CERN-IT)
27/10/2009, 13:15
Storage
This talk will present an update on R&D activities around storage at CERN. The main focus will be various
activities around iSCSI technology, but an update will also be presented on the Lustre evaluation project.
Pierre Emmanuel Brinette
(CNRS-CCIN2P3)
27/10/2009, 13:40
Storage
IN2P3 Computing Center has been using HPSS as a Mass Storage System since 1999. There has been no major system upgrade since 2005 and IN2P3 still runs HPSS 5.1. This version is no more supported by IBM and doesn't include the T10K-B drive support.
In June 2009, the system has been upgraded to HPSS 6.2.2.2. This upgrade implies major changes (DCE removal, DB2 and systems upgrade).
This...
Jason Hick
(LBNL)
27/10/2009, 14:05
Storage
The High Performance Storage System (HPSS) has served the DOE community for high performance archival storage for the past fifteen years. It specifically serves the HEP community at LBNL/NERSC by providing archival storage from the PDSF system. This presentation will provide a brief overview of how HPSS works, what its current more unique features are, and what our plans our for our next...
Jonathan Schaeffer
(CCIN2P3)
27/10/2009, 14:30
Storage
TReqS is our Tape Request Scheduler. Based on BNLBatch, its goal is to get between dCache and HPSS and to reorder the files requests.
Since May of this year, a first implementation hit our production system.
We will present here :
- the problematic of tape access for LHC experiments
- the solution we implemented
- TReqS in its production environement, our 5 month experiment with the beast
Peter van der Reest
(DESY)
27/10/2009, 14:55
At DESY we have installed a testbed for Grid and storage related issues. One of the first trials we did was running a ALICE Proof job against data held in a distributed NFS 4.1 service based on dCache. Also trails were made against an industrial system, of which I am not allowed to speak yet. Comparisons we made with regard to scaling in PROOF and IO performance in comparison to NFS3 To make...
Mr
Juraj Sucik
(CERN)
27/10/2009, 15:30
Virtualization
The Internet services group provides the infrastructure and sophisticated management tools for virtual machine provisioning based on Hyper-V, Microsoft Virtual Machine Manager, management SOAP web services and a user web interface. This virtualisation service has already confirmed its reliability and efficiency by wide range of satisfied users. The infrastructure which was presented at the...
Andrea Chierici
(INFN-CNAF)
27/10/2009, 16:00
Virtualization
INFN-T1 implemented a solution to get Worker-nodes on demand using virtualization technology. This solution is allowing us extreme flexibility providing dynamic virtual execution environments and integrating seamlessly into our production grid. Currently 200 VMs slots are available and will increase further.
Dr
Tony Cass
(CERN)
27/10/2009, 16:30
Virtualization
This talk presents a possible roadmap for the use of virtualisation across WLCG sites to deliver improved computing services for the experiments and users.
Dr
Miguel Oliveira
(LIP - Laboratório de Instrumentação e Física Experimental de Partículas)
28/10/2009, 09:00
Site Reports
Site report for all sites (LIP-Lisbon,LIP-Coimbra,NCG-INGRID-PT) and activities at LIP.
Andrea Chierici
(INFN-CNAF)
28/10/2009, 09:20
Site Reports
I will present the status report of the Italian Tier1 site
Kelvin Raywood
(TRIUMF)
28/10/2009, 09:40
Site Reports
An external review of TRIUMF computing took place and some changes have been recommended. The Tier-1 Center has completed its acquisition for the 2009/2010 upgrade. We continue to use a mix of Xen, OpenVZ and Hyper-V for virtualisation with OpenVZ being preferred for hosted servers. For Linux desktops, we provide a repository of TRIUMF rpms which customise a standard Scientific Linux installation.
Mr
Wojciech Lapka
(Unknown)
28/10/2009, 10:30
Monitoring Infrastructure and Tools
Since 2005 Worldwide LHC Computing Grid (WLCG) services have been monitored by the Service Availability Monitoring (SAM) system which has been the main source of information for the monthly WLCG availability and reliability calculations.
During this time SAM framework gained popularity amongst site and service managers and was very useful in building robust grid infrastructure.
Experience...
Mr
Thomas Davis
(NERSC/LBNL)
28/10/2009, 11:00
Monitoring Infrastructure and Tools
We present a method of monitoring the environment and
performance using open source tools such as Nagios, Ganglia and Cacti to collect and display performance data as well as availability information for various components of large computing systems in an integrated fashion. We will present information on how the data is collected, viewed and analyzed, with specific examples from NERSC's Cray system.
Frédéric AZEVEDO
(CC-IN2P3)
28/10/2009, 11:30
Due to the continuous load and intensive usage on our robotics, we regularly face some hardware
issues with tapes and tape drives. A recurrent issue concerns possible data loss which leads to go
through a long recovery process.
In order to improve our reliability, we have studied commercial solutions to avoid permanent
write/read errors, or at least foresee occurring errors. We've tested...
Mr
Sebastian Lopienski
(CERN)
28/10/2009, 13:15
Network, Security
This talk gives an update on security issues affecting computers, software applications and networks during the last months. It includes information on emerging types of vulnerabilities and recent attack vectors, and provides an insight into the cyber-crime economy of 2009. This talk is based on contributions and input from the CERN Computer Security Team.
Dr
David Kelsey
(RAL)
28/10/2009, 13:45
Network, Security
An update on Grid security in WLCG, EGEE and EGI, concentrating on progress in operational and policy issues.
Miguel Oliveira
(LIP - Laboratório de Instrumentação e Física Experimental de Partículas)
28/10/2009, 14:45
Benchmarking
Benchmarking is a key activity on all computer centers not only for tender procedures but also to optimize resources. At LIP we underwent recently major upgrades of all sites and deployed a new one.
We report on CPU HEP-SPEC and storage benchmarking results.
Dr
Tony Chan
(Brookhaven National Laboratory)
28/10/2009, 15:30
This presentation will describe the expansion of the RHIC/ATLAS Computing Facility
(RACF) to accomodate its commitments to the computational needs of the scientific
programs at Brookhaven National Laboratory. The expansion has nearly tripled the
footprint of the facility over the past 2+ years and allows the RACF to adequately
meet our computing and storage requirements for the foreseeable...
Mr
Shesha Krishnapura
(Intel)
28/10/2009, 16:30
Silicon design technical complexity is increasing every year due to several new features and process technology shrinks. Additionally, the business drivers such as shorter product development time, reduced headcount, and lower cost is increasing pre-silicon verification, high degree of design automation, and global multi-site design teams. These two factors (technological and business) are...
Mr
Sebastian Lopienski
(CERN)
29/10/2009, 09:00
CERN hosts a large number of Web sites (CERN-related, but also private), both on central Web Services, as well as on machines managed by particular Web site owners. Some of these Web sites are actually interactive Web applications developed with languages like PHP, ASP, Java, Perl, Python etc. - and unavoidably a fraction of them have bugs making them vulnerable to attacks such as Cross Site...
Maarten Litmaath
(CERN)
29/10/2009, 09:30
Network, Security
The Worldwide LHC Computing Grid (WLCG) infrastructure has been built up
for the storage and analysis of the very large data volumes that will be
recorded by the LHC experiments. Its existing security mechanisms and
policies are foreseen to evolve in various respects, for example with an
increasing use of virtual machines, pilot jobs, clouds, enhancements to
data storage and access...
29/10/2009, 10:15
Monitoring Infrastructure and Tools
The UK Tier 1 Centre at RAL will increase in size significantly in the
coming year. The need for better automation of both system deployment and
ongoing configuration management has prompted a survey of possible
solutions. Since deciding to adopt Quattor earlier this year, we have
successfully deployed our new SL5 batch service using the system and are
already seeing better consistency...
marc hausard
(CC-IN2P3)
29/10/2009, 10:45
Monitoring Infrastructure and Tools
At CC-IN2P3, Nagios have taken over from the previous system to become the main monitoring tool used by the operation. This presentation will introduce its configuration in a Tier-1 environment and will present various extra features developed at CC-IN2P3 in order to customize the notification system and to provide multi-user development and failover mechanism.
Mr
Andrew Uselton
(NERSC/LBNL)
29/10/2009, 11:15
Monitoring Infrastructure and Tools
The Lustre Monitoring Tool (LMT) provides a useful view of the server-side behavior of the Lustre parallel file system. This talk presents a brief overview of the architecture of the tool and explores several use cases including tracking system health, server-side performance tuning, applications-side performance tuning, and incident evaluation, among others.
Mr
Tobias Koenig
(Karlsruhe Institute of Technology (KIT))
29/10/2009, 14:15
ITIL
The Karlsruhe Institute of Technology (KIT) was founded at 1st October 2009, merging the University of Karlsruhe and the Forschungszentrum Karlsruhe. As a first new organizational unit of the KIT the Steinbuch Centre for Computing (SCC) was established, combining the former Institute for Scientific Computing of Forschungszentrum Karlsruhe and the Computing Centre of the Technical University...
Dr
Chadwick Keith
(Fermilab)
29/10/2009, 14:45
ITIL
The software acceptance process that is used by FermiGrid together with the operational experience of operating using this process will be presented.
Dr
Tony Cass
(CERN)
29/10/2009, 15:15
ITIL
This talk will cover progress at CERN to improve service organisation throtugh adoption of ITIL principles.
Juraj Sucik
(CERN)
29/10/2009, 16:00
Desktop Management
Windows 7, the next version of Windows OS, is scheduled to be available worldwide on 22 October 2009. CERN IT-IS group has been working with it ever since its beta release in January 2009. The purpose of this talk is to discuss this experience and to share the plans for deployment of Windows 7 at CERN.
Christopher Huhn
(GSI Helmholtzzentrum für Schwerionenforschung GmbH)
29/10/2009, 16:30
Desktop Management
The standard Linux desktops at GSI received their operating system via NFS and not from the local hard drive since the last millennium.
Since then this policy has been enhanced to work with shared read-only OS images that provide advanced security and fast OS deployment and upgrades. Image generation and configuration management of NFS-root desktops is completely integrated into our...
Mr
Alan Silverman
(CERN)
29/10/2009, 17:00
Desktop Management
We will review the Desktop Computing Support across some HEP sites.
Joe Burrescia
(ESnet)
30/10/2009, 09:00
Network, Security
This talk will cover an introduction to the Energy Sciences Network (ESnet). It will outline both ESnet as a program in DOE's Office of Science and the current network implementation known as ESnet4. The roles of both components of ESnet4, the IP network and Science Data Network (SDN) will be discussed. The presentation will also touch upon several international network collaborations that...
Mr
Brian Tierney
(NERSC/LBNL)
30/10/2009, 09:30
Roberto Gomezel
(INFN)
30/10/2009, 10:00
Site Reports
An overview about INFN computing and networking activities
John Bartelt
(SLAC)
30/10/2009, 10:20
Site Reports
Report on new personnel and projects, and status of IT and HPC at SLAC.
Jiri Chudoba
(Institute of Physics, Prague)
30/10/2009, 10:40
Site Reports
Prague Tier2, its current status and plans for nearest future.
Mattias Wadenstein
(NDGF)
30/10/2009, 11:00
Site Reports
News and overview of the whats happening in the NDGF region, as well as some small updates on previously covered topics in NDGF-related matters.
