|
|
||||
|
Bulletin 199 - 2010 March 12
1. HPCCC and CSIRO ASC Helpdesk Updates The solar userguide is available at: http://www.hpccc.gov.au/hpccc/userguides/solar/ Users requiring assistance with solar should log an incident report using iSupport (aka cSupport):
For guidance on using iSupport see the FAQ at http://www.hpccc.gov.au/hpccc/userguides/faq/isupport.php 1.2 NEC SX-6, LSDSS and SAM Help Users requiring assistance with the NEC SX-6, LSDSS and SAM can:
Note: hpchelp@csiro.au should no longer be used for requesting assistance for the NEC SX-6. Please use the address. 1.3 CSIRO ASC Help - burnet, cherax and gpu cluster RT (Request Tracker) is our new system for managing user support requests.
For guidance on using RT see the FAQ at Users requiring assistance with burnet and cherax can:
Users requiring assistance with the gpu cluster can:
[ page top ] 2. History - CSIRO Supercomputing and Scientific Data Storage This month marks the 20th Anniversary of the commissioning of CSIRO's first Cray Research Supercomputer. The 1990 system, a Cray Research Y-MP 2/216, was housed in Port Melbourne and managed by Leading Edge Technologies, which used a share of the system for seismic data processing. The system was named cherax, being a rather weak pun on the name Cray, since cherax is the genera name for the yabby, the Australian Cray. We initially pronounced cherax with a starting sound like "chips", but in the opening ceremony on 23rd March 1990, the Hon Barry Jones AO remarked that "since the name is derived from the Greek, the pronunciation is cherax, as in chemistry" (with a hard "k" sound). We stood corrected. CSIRO went on the use two other Crays, and then two SGI Altix systems all bearing the name cherax. When the original Cray system went live, it used large washing-machine sized DD-49 disc units, each holding 1.2 Gbyte with 9.8 Mbyte/s transfer rates. These soon proved to have inadequate capacity, and in November 1991, a hierarchical storage system called the Data Migration Facility (DMF) was brought into operation, to extend the file system storage beyond the disc and on to magnetic tape. The home file system set up for users in 1990 has been carried forward and still runs on the CSIRO IM&T Advanced Scientific Computing's Altix system today, as the CSIRO ASC Data Store, and still uses DMF. This file system is on its fourth site, its fifth host, and its ninth type of tape technology, and now holds 1 petabyte of data (with two or more copies kept of all the data). Indeed, within that file system, we have been able to identify code written in the mid-1960s by staff from the CSIRO Division of Computing Research, no doubt originally on punched cards! So, the ASC Data Store has a long history, and now supports users doing computation on the NCI National Facility at ANU, the HPCCC systems at the Bureau of Meteorology, and on various ASC systems. This month, the National Computing Infrastructure's Sun Constellation system will be upgraded to its full size. CSIRO has a 25% share of the new system. For further information about current facilities, see http://intra.hpsc.csiro.au/ Here is a comparison between the Cray Y-MP and the NCI National Facility Sun Constellation system which is due to come into full production this month.
Here is a comparison between the Cray Y-MP storage and the new storage facilities on the CSIRO ASC Data Store, after a pending upgrade this year.
[ page top ] 3. CSIRO ASC - User set job priority for cherax, burnet, gpu cluster If you are submitting many jobs and need to ensure that some of them have a higher priority than others, you can now set a user priority with the qsub '-p' option. It takes an integer argument between -1023 and 1023 (larger number for more priority). [ page top ] 4. CSIRO ASC New and Upgraded software
[ page top ]
|
||||||||||||||||||||||||||||||||||||||||||||||
|
Comments to: © Copyright 2010, CSIRO Australia Use of this web site and information available from it is subject to our Legal Notice and Disclaimer and Privacy Statement |