Please note: the text below is from this NGS press release. From the 5th April 2004, the UK National Grid Service (NGS) enters its initial production phase which is funded to provide a significant resource for the UK e-Science Community. The four new clusters, supplied and maintained by ClusterVision and hosted by the University of Manchester, CCLRC (Rutherford Appleton Laboratory), University of Oxford and the White Rose Grid (Leeds) are now available for early adopter users to register. As part of the initial 'core' the NGS also includes access to the resources of the HPCx and CSAR services. Building on the experiences of the UK e-Science community and in particular the work achieved by the Engineering Task Force (ETF) in building the UK e-Science 'Level 2 Grid' last year, Tony Hey initiated the UK NGS to provide not only a significant resource for research that allows users to 'do science using grid technologies', but also an infrastructure to which new resources can be added in the future. Following the initial phase the UK National Grid Service will enter its full production stage. At present, access to the service is guided by the NGS management board. The aim of this is to maximise the potential of these new resources and encourage existing or new projects to experiment and utilise grid technology in their work. Available to users are 2 compute and 2 data facilities (the services of HPCx and CSAR are treated separately as they have existing funding mechanisms for users). The compute facilities comprise 64 dual CPU Intel 3.06GHz nodes, whilst the data facilities comprise 20 dual CPU Intel 3.06GHz nodes and 18TB Fibre SAN storage, giving a total of 36TB of storage across both sites available to users of the NGS. Further information on the UK National Grid Service and how to register as a user can be found at www.ngs.ac.uk by emailing support@grid-support.ac.uk. If you have resources that you would like to integrate into the NGS then contact a.j.richards@dl.ac.uk. Notes: - The UK National Grid Service (NGS) comprising two data clusters and two compute clusters is funded by JISC and CCLRC.
- The size of the clusters is such that they offer a significant resource for the UK e-Science community.
Compute Clusters - 64 dual CPU Intel 3.06 GHz (1MB cache) nodes
- 2GB memory per node
- 2 x 120GB IDE disks (1 boot, 1 data)
- Gigabit network
- Myrinet M3F-PCIXD-2
- Front end node
- Disk server node with 2 x Infortrend 2TB U160 SCSI Arrays (147GB SCSI disks)
- PGI compilers
- Intel compilers, MKL
- PBS Pro
- TotalView debugger
- ClusterVisionOS based on Red Hat Enterprise 3.0
Data Cluster - 20 dual CPU Intel 3.06 GHz (1MB cache) nodes
- 4GB memory per node
- 2 x 120GB IDE disks (1 boot, 1 data)
- Gigabit network
- Myrinet M3F-PCIXD-2
- Front end node
- 18TB Fibre SAN (Infortrend F16F 4.1TB Fibre Arrays (147GB FCAL disks)
- PGI compilers
- Intel compilers, MKL
- PBS Pro
- TotalView debugger
- Oracle 9i RAC & Application server
- ClusterVisionOS based on Red Hat Enterprise 3.0
- The clusters are located at Manchester (data cluster), Oxford (compute cluster), CCLRC (data cluster) and the White Rose Grid (compute cluster).
- JISC are funding one staff post at each site to ensure that the clusters are professionally managed and to enable a production grid software stack to implemented. An additional post at CCLRC-RAL is the Grid Coordinator who will oversee the project and ensure that it is a success.
- CSAR and HPCx are also part of the core NGS but they have their own funding arrangements and user policies. Other sites are already in negotiation to commit resources to the NGS.
- Further information, conditions of use and registration information can be obtained from www.ngs.ac.uk.
For more information
Dr ir Matthijs van Leeuwen
ClusterVision BV
Nieuw-Zeelandweg 15B
1045 AL Amsterdam
Netherlands
Tel NL: +31 20 407 7550
Tel UK: +44 870 080 1980
Fax: +31 84 759 8389
www.clustervision.com
| 
