|
|
||||
|
Bulletin 176 - 2008 March 17
Note: "CSIRO" items can apply to BoM users of cherax and burnet 1. HPCCC SX-6/TX7 system continuation The HPCCC has extended the maintenance on the SX-6/TX7 system with NEC for another year beyond the end of the current contract on 31st March 2008. [ page top ] 2. HPCCC NEC Applications Support The HPCCC has not extended applications support with NEC beyond the end of the current contract on 31st March 2008. From 1st April 2008, would users please not contact NEC for applications support, but channel any requests through the HPCCC request system. [ page top ] 3. SX-6 scheduling As advised in HPCbull 175, we wish to trial a new way of allocating jobs to CPUs on the SX-6 nodes, as has been under trial on the CSIRO nodes for several months. To enable the trial to be carried out, please immediately reduce the maximum number of CPUs per node requested by jobs from 8 to 7. #PBS -l cpunum_prc=7 #PBS -l cpunum_job=7 Please also stop requesting more than one CPU for jobs going to the bm queue. Trials are then scheduled to be carried out from Tuesday 18th March. [ page top ] 4. Change in policy for the $WORKDIR flush areas on the SX-6/TX7 system USERS SHOULD NOTE THE CONSEQUENCES OF THIS CHANGE. Following further incidents where flushable file systems on the SX-6/TX7 system filled, and led to both TX7s going down, the policy advertised in HPCbull 174 will be implemented. The major change will be that Bureau $WORKDIR files will be retained for a minimum of only 7 days. (CSIRO $WORKDIR files are already subject to flushing down to 7 days). We propose simplifying the procedure outlined in the draft policy. When usage reaches 95%, the system will proceed in a single pass, removing files in order from the oldest to youngest, until usage is reduced to 80% (in which case the procedure will stop), or until it has removed all files older than 7 days, rather than a two-stage process with operator intervention. If the procedure stops without reducing the usage down to 80%, the operators will be notified as well as HPCCC staff. Reports will be placed in /bm/flush*/flush.status If you have objections to this policy, please contact the HPCCC immediately to discuss your requirements. Otherwise, this change will be implemented on 1st April. Please do not rely on files staying in the flushable areas for longer than 7 days. Please remember that there is no backup of files in the $WORKDIR and $DATADIR areas. Please allow everyone to make good use of these areas, by removing files no longer needed. [ page top ] 5. CSIRO - Burnet Cluster Refresh
[ page top ] 6. Milestone for batch jobs on burnet We have reached a significant milestone for burnet - one million jobs have been submitted! And the lucky user to submit the millionth job is ..... Wes Barris from Livestock Industries - well done! Note that the SX-6/TX7 system handles over 200,000 jobs per month! [ page top ] 7. More space for short jobs on burnet We have recently added more compute nodes to the pool that can run very short jobs - 2 hours or less. Also, there are now more reservations for jobs that fall into 6, 12 and 23 hours of walltime. You are likely to get better turn-around if you break up your work to fit into these slots. [ page top ] 8. New Software on burnet and cherax
[ page top ] 9. New Software on cherax
[ page top ] 10. CSIRO - New ASC software To satisfy user demand we recently purchased Intel C++ for Windows v10. There are two floating licenses that can be used anywhere within CSIRO. For instructions on obtaining and installing the software visit the APAC software map at http://nf.apac.edu.au/facilities/software/index.php?l=&site=CSIRO [ page top ] 11. CSIRO - ASC Roadshow The The ASC Roadshow, presented by ASC Senior Manager Dr. Alf Uhlherr, continues this month in Queensland. The 90 minute presentation will include time for questions, and is designed to help CSIRO users understand and take full advantage of CSIRO Advanced Scientific Computing capabilities. Venues are given below; further locations can be added on request.
The ASC team also welcome the opportunity to meet on the day with individual groups for more in-depth discussion concerning the ASC requirements for their research. For further information, please contact Justin Baker on 03 8601 3801, or [ page top ] 12. CSIRO - New ASC positions advertised internally There are some exciting opportunities to join the Advanced Scientific Computing team. A number of roles have just been advertised internally to CSIRO. We are seeking to fill four roles.
Systems Manager, to co-ordinate the administration of
ASC's central computing systems,
Grid Systems Administrator, to maintain CSIRO's grid
gateway system as part of our participation in the ARCS
partnership,
Application and User Support Specialists (two similar roles),
to work with scientists to help them make effective use of
Advanced Scientific Computing resources. Applications close on the 19th March. The positions will be advertised externally if they are not filled by internal applicants. [ 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 |