Jens G Jensen e-Science Centre hepsysmanix HEPiX report for hepsysman RAL, 10 May 2006.

Slides:



Advertisements
Similar presentations
30-31 Jan 2003J G Jensen, RAL/WP5 Storage Elephant Grid Access to Mass Storage.
Advertisements

Jens G Jensen CCLRC/RAL AHM Nottingham 2005Storage Middleware HEPiX Meta-Summary (and Storage Middleware) Jens Jensen GridPP16 QMUL June 2006.
Jens G Jensen Atlas Petabyte store Supporting Multiple Interfaces to Mass Storage Providing Tape and Mass Storage to Diverse Scientific Communities.
J Jensen CCLRC RAL Data Management AUZN (mostly about SRM though) GGF 16, Athens J Jensen.
Steve Traylen Particle Physics Department Experiences of DCache at RAL UK HEP Sysman, 11/11/04 Steve Traylen
Jens G Jensen CCLRC/RAL hepsysman 2005Storage Middleware SRM 2.1 issues hepsysman Oxford 5 Dec 2005.
Storage: Futures Flavia Donno CERN/IT WLCG Grid Deployment Board, CERN 8 October 2008.
Secure Off Site Backup at CERN Katrine Aam Svendsen.
TRIUMF Site Report for HEPiX, SLAC, October 10-14,2005 TRIUMF SITE REPORT Corrie Kost Update since Hepix Spring 2005.
CERN IT Department CH-1211 Genève 23 Switzerland t Next generation of virtual infrastructure with Hyper-V Michal Kwiatek, Juraj Sucik, Rafal.
I/O Systems ◦ Operating Systems ◦ CS550. Note:  Based on Operating Systems Concepts by Silberschatz, Galvin, and Gagne  Strongly recommended to read.
Copyright © 2010 Platform Computing Corporation. All Rights Reserved.1 The CERN Cloud Computing Project William Lu, Ph.D. Platform Computing.
Lecture 4 Operating System & Application Software.
HEPiX Orsay 27 th April 2001 Alan Silverman HEPiX Large Cluster SIG Report Alan Silverman 27 th April 2001 HEPiX 2001, Orsay.
IRODS performance test and SRB system at KEK Yoshimi KEK Building data grids with iRODS 27 May 2008.
Introduction and Overview Questions answered in this lecture: What is an operating system? How have operating systems evolved? Why study operating systems?
A comparison of distributed data storage middleware for HPC, GRID and Cloud Mikhail Goldshtein 1, Andrey Sozykin 1, Grigory Masich 2 and Valeria Gribova.
October, Scientific Linux INFN/Trieste B.Gobbo – Compass R.Gomezel - T.Macorini - L.Strizzolo INFN - Trieste.
StoRM Some basics and a comparison with DPM Wahid Bhimji University of Edinburgh GridPP Storage Workshop 31-Mar-101Wahid Bhimji – StoRM.
Storage Tank in Data Grid Shin, SangYong(syshin, #6468) IBM Grid Computing August 23, 2003.
CERN IT Department CH-1211 Geneva 23 Switzerland t Storageware Flavia Donno CERN WLCG Collaboration Workshop CERN, November 2008.
Batch Scheduling at LeSC with Sun Grid Engine David McBride Systems Programmer London e-Science Centre Department of Computing, Imperial College.
D C a c h e Michael Ernst Patrick Fuhrmann Tigran Mkrtchyan d C a c h e M. Ernst, P. Fuhrmann, T. Mkrtchyan Chep 2003 Chep2003 UCSD, California.
Introduction to dCache Zhenping (Jane) Liu ATLAS Computing Facility, Physics Department Brookhaven National Lab 09/12 – 09/13, 2005 USATLAS Tier-1 & Tier-2.
RAL Site Report Andrew Sansum e-Science Centre, CCLRC-RAL HEPiX May 2004.
Δ Storage Middleware GridPP10 What’s new since GridPP9? CERN, June 2004.
Your university or experiment logo here GridPP Storage Future Jens Jensen GridPP workshop RHUL, April 2010.
HEPix April 2006 NIKHEF site report What’s new at NIKHEF’s infrastructure and Ramping up the LCG tier-1 Wim Heubers / NIKHEF (+SARA)
Optimisation of Grid Enabled Storage at Small Sites Jamie K. Ferguson University of Glasgow – Jamie K. Ferguson – University.
Jefferson Lab Site Report Sandy Philpott Thomas Jefferson National Accelerator Facility Jefferson Ave. Newport News, Virginia USA 23606
Light weight Disk Pool Manager experience and future plans Jean-Philippe Baud, IT-GD, CERN September 2005.
Owen SyngeTitle of TalkSlide 1 Storage Management Owen Synge – Developer, Packager, and first line support to System Administrators. Talks Scope –GridPP.
Distributed Data Management Graeme Kerr Oracle in R&D Programme.
USATLAS dCache System and Service Challenge at BNL Zhenping (Jane) Liu RHIC/ATLAS Computing Facility, Physics Department Brookhaven National Lab 10/13/2005.
Forschungszentrum Karlsruhe in der Helmholtz-Gemeinschaft Implementation of a reliable and expandable on-line storage for compute clusters Jos van Wezel.
Status SC3 SARA/Nikhef 20 juli Status & results SC3 throughput phase SARA/Nikhef Mark van de Sanden.
CERN Database Services for the LHC Computing Grid Maria Girone, CERN.
Jens G Jensen RAL, EDG WP5 Storage Element Overview DataGrid Project Conference Heidelberg, 26 Sep-01 Oct 2003.
14 th April 1999CERN Site Report, HEPiX RAL. A.Silverman CERN Site Report HEPiX April 1999 RAL Alan Silverman CERN/IT/DIS.
EGEE-II INFSO-RI Enabling Grids for E-sciencE EGEE Site Architecture Resource Center Deployment Considerations MIMOS EGEE Tutorial.
HEPiX 2 nd Nov 2000 Alan Silverman Proposal to form a Large Cluster SIG Alan Silverman 2 nd Nov 2000 HEPiX – Jefferson Lab.
Scientific Storage at FNAL Gerard Bernabeu Altayo Dmitry Litvintsev Gene Oleynik 14/10/2015.
HEPiX report Helge Meinhard, Harry Renshall / CERN-IT Computing Seminar / After-C5 27 May 2005.
CERN - IT Department CH-1211 Genève 23 Switzerland t High Availability Databases based on Oracle 10g RAC on Linux WLCG Tier2 Tutorials, CERN,
Storage and Data Movement at FNAL D. Petravick CHEP 2003.
Grid Technology CERN IT Department CH-1211 Geneva 23 Switzerland t DBCF GT Upcoming Features and Roadmap Ricardo Rocha ( on behalf of the.
Office of Science U.S. Department of Energy NERSC Site Report HEPiX October 20, 2003 TRIUMF.
HEPIX Backup Survey David Asbury CERN/IT/FIO HEPIX, Rome, 6 April 2006.
HEPiX/HEPNT report Helge Meinhard, Alberto Pace, Denise Heagerty / CERN-IT Computing Seminar 05 November 2003.
EGI-Engage Data Services and Solutions Part 1: Data in the Grid Vincenzo Spinoso EGI.eu/INFN Data Services.
GridPP storage status update Joint GridPP Board Deployment User Experiment Update Support Team, Imperial 12 July 2007,
SESEC Storage Element (In)Security hepsysman, RAL 0-1 July 2009 Jens Jensen.
BNL dCache Status and Plan CHEP07: September 2-7, 2007 Zhenping (Jane) Liu for the BNL RACF Storage Group.
Western Tier 2 Site at SLAC Wei Yang US ATLAS Tier 2 Workshop Harvard University August 17-18, 2006.
1 5/4/05 Fermilab Mass Storage Enstore, dCache and SRM Michael Zalokar Fermilab.
Next Generation of Apache Hadoop MapReduce Owen
Martina Franca (TA), 07 November Installazione, configurazione, testing e troubleshooting di Storage Element.
An Introduction to GPFS
Grid Technology CERN IT Department CH-1211 Geneva 23 Switzerland t DBCF GT Standard Protocols in DPM Ricardo Rocha.
GPFS Parallel File System
Scientific Linux Connie Sieh CSAM Meeting May 2, 2006.
CERN IT Department CH-1211 Genève 23 Switzerland t DPM status and plans David Smith CERN, IT-DM-SGT Pre-GDB, Grid Storage Services 11 November.
High Performance Storage System (HPSS) Jason Hick Mass Storage Group HEPiX October 26-30, 2009.
WLCG IPv6 deployment strategy
James Casey, IT-GD, CERN CERN, 5th September 2005
dCache “Intro” a layperson perspective Frank Würthwein UCSD
Introduction to Data Management in EGI
Artem Trunov and EKP team EPK – Uni Karlsruhe
Enabling High Speed Data Transfer in High Energy Physics
Data Management cluster summary
Presentation transcript:

Jens G Jensen e-Science Centre hepsysmanix HEPiX report for hepsysman RAL, 10 May 2006

Jens G Jensen e-Science Centre Grid Storage for beginners SRM is a control protocol –Version 1.1 and version 2.1 One head node, many pool nodes Data transfer –WAN protocol: GridFTP –LAN protocol: RFIO or DCAP Implementations: –dCache from DESY, its SRM from FNAL –DPM from CERN/LCG –StoRM from INFN (SRM 2 only)

Jens G Jensen e-Science Centre Grid Storage for admins XFS improves SE performance –See Greig’s talk –Support issues Tape vs disk –“Adapt to customer demand more easily with tape” (FNAL) –Easy to expand – cost of media –HSMs can facilitate utilisation –Performs differently, mechanical

Jens G Jensen e-Science Centre CPU 32 bit vs 64 bit (FNAL) –32 bit OS, 32 bit apps, on 64 bit proc. –64 bits OS/drivers, 32 bit apps, on 64 bit proc % slower than 32/32 –64 bit OS, 64 bit apps, on 64 bit proc % faster than 32/32 Extra features un(der)used – CERN –Security, SSE, power saving,…

Jens G Jensen e-Science Centre CPU, cores Dual core –Good performance/price – FZK –Almost same power consumption – FNAL More, more –Quad cores early ’07? –Octo cores ’09? Fast parallelism, IPC –Debugging, algorithms, paradigms –

Jens G Jensen e-Science Centre Power / Power ratio Maximise computing power, minimise electrical power IN2P3 tests –Opteron > Xeon –Dual cores > hyperthreading –Big names better –Blade systems better

Jens G Jensen e-Science Centre Optimisation Optimised by compiler or programmer? Platform and compiler (and language and versions) dependencies –How to set the right options (easy-ish), –Improve algorithms, –Performance Analysis (correctness, regression test) –Understand machine code, cache, execution units,… Inlining: speedup, but code dup; cache misses can slow down Tools to diagnose – premature opt’n: rule; diminishing returns: rule

Jens G Jensen e-Science Centre Scientific Linux SL: FNAL-repackaged RedHat – SLC: SL CERN –With added CERN bits (e.g. kernel) –XFS support Version 3 vs version 4 –3.0.x widely used in LCG… –…but its days are numbered –…there should be a 3.0.8, at least

Jens G Jensen e-Science Centre SL4 issues XFS support –Not upstream, but maybe(?) in SLC –Cannot use XFS for pools now, need to upgrade to SL4 later… 64 bit processors –Likely to support x86_64 but not IA64 –SLC might support IA64?

Jens G Jensen e-Science Centre Monitoring MonAMI – Glasgow –Overview, Alerts, Gfx, Logs, Details –Target  Sensor  Display –MonAMI universal sensor framework

Jens G Jensen e-Science Centre Filesystems (CERN) ext3, XFS, JFS, ReiserFS Blocks and inodes –Can be clumsy for large files –Extents can prevent fragmentation Performance –Much depends on controller and disk –PCs bad at mixing R/W IO, and concurrent writes –See also Greig’s talk

Jens G Jensen e-Science Centre Journaling Filesystems in RHEL4 SLOCIncl. since ExtentsStatus EXT310K2000NoWidely used Reiser v3 30K2001NoSmall files. Fragile? XFS100K2001YesLarge files. Widely used JFS30K2002NoNot widely used, slow devel.

Jens G Jensen e-Science Centre Distributed Filesystems (RZG) NFS –Slow for large files –Cluster only for non-security AFS –Global GPFS –Commercial –Fast, striping data across disks –Defective or malicious client can disrupt the system…

Jens G Jensen e-Science Centre Distributed Filesystems (RZG) SAN filesystems… –Fast, but infrastructure expensive –Cluster only (local uid) –Client can corrupt system Object filesystems – LUSTRE, Panasas –Clients can not corrupt the server –Only few in linux –Metadata server req’d (also for access ctrl)

Jens G Jensen e-Science Centre Backup at HEP sites 9 Home made (INFN, RAL-HPC) 7 TSM (CERN, DESY-H, FZK, IN2P3, 3xINFN) –“Functionality, wide range of clients, good support” 3 Legato (DESY-Z, 2xINFN) –“Good offer for our Solaris requirements when started” 2 Time Navigator (2xINFN) 3 HP Data Protector (3xINFN) 1 TiBS(FNAL) –“Scalable, easily partitioned, met requirements” 1 AMANDA (TRIUMF) –“works and is cheap” 1 Oracle Reliaty (JLAB) –“Best fit to requirements in 2001”

Jens G Jensen e-Science Centre Authentication ssh, grid certs, Kerberos, passwords OTP deployed at LBNL –Easy to integrate –Slightly expensive hw tokens SSO project at CCLRC –Grid access via MyProxy, integrated with site AD and site user mgmt –Deployed for NGS

Jens G Jensen e-Science Centre Misc CERN uses BOINC –( to utilise extra resources (see Data Transfer –FTS seems more robust than RFT atm (see Graeme’s talk) Procure by specint (FZK) –Then economy (number, power, …) Helpdesk –SLAC (also) switched from Remedy to RT