Daily Bulletin Archive

Aug. 21, 2012

The CISL User Services Section has published additional documentation to help users prepare for computing with the Yellowstone, Geyser, and Caldera resources that are being tested at the NCAR-Wyoming Supercomputing Center.

The documentation includes compilation commands for the Intel, PGI, PathScale, and GNU compilers to be used in the new system. It also describes the Yellowstone environment’s file format and mathematical libraries.

Feel free to use the Feedback link on our Support & Training menu to let us know what you think.

Aug. 16, 2012
Date and Time: 
Aug 16th, 2012, 10:00

Who should attend:

  • UCAR users who move big data…
    lots of small files, very large files,
    or anywhere in between
  • Users that access NCAR data nodes
    on XSEDE, DOE, or U.Colorado endpoints
  • Campus computing resource admins
  • When: Thursday, August 16, 10:00AM MT
  • Location: Webinar URL will be provided after registration
  • Presenter: Steve Tuecke, Deputy Director of the Computation Institute, University of Chicago and Argonne National Laboratory

Globus Online is a convenient interface for transferring files between two endpoints – for example, between NCAR resources, the University of Colorado, and XSEDE facilities or other sites. Globus Online also offers a feature called Globus Connect, which enables you to move files easily to and from your laptop or desktop computer and other endpoints.

The Consulting Services Group at NCAR will host a two-hour online workshop to help UCAR users set up and work with Globus Online. You are encouraged to use your own laptop and follow along with the presenter.

Click here to register for this workshop. Please contact Si Liu (siliu_at_mail_dot_ucar_dot_edu) if you have any questions or concerns.

Aug. 10, 2012

The NCAR-Wyoming Supercomputing Center (NWSC) has been the site of a flurry of activity in July, with IBM and CISL making significant progress on the complex installation of the Yellowstone environment. As of July 30, most hardware installation issues have been resolved, IBM has run performance tests on the full 11-PB GLADE system, and IBM and CISL have installed most key software and configured the administrative and user environments.

At this time, we estimate that Yellowstone will be ready for the first users in early September, at which point the Accelerated Scientific Discovery (ASD) projects will begin. The rest of the user community will have access as soon as practical after the ASD projects have started.

In addition, because the Yellowstone schedule has slipped, CISL has reviewed the plan for Bluefire. Barring further delays, CISL will extend Bluefire's lifetime at least through the end of October 2012. CISL intends to operate Bluefire through November, but in the event of a major system failure during the month, the system may not be returned to service. Bluefire users should plan their work accordingly.

Between now and the beginning of September, Yellowstone will be subjected to a battery of tests. The system is now entering the "functional testing" period during which IBM will begin to put Yellowstone through a series of tests designed to confirm that all components are functioning as an integrated system and delivering the expected performance.

Following functional testing and after resolving any remaining hardware issues, Yellowstone will enter the three-week Acceptance Testing Period, during which IBM and CISL will jointly conduct benchmarks and run the system with a simulated workload around the clock. When Yellowstone successfully passes the acceptance tests, CISL will take over the system from IBM, and we are allowing a few days for NCAR-specific configuration and customization.

We will keep you posted as we proceed, particularly about significant changes in this schedule. Please contact cislhelp@ucar.edu with any questions.

Aug. 9, 2012
Date and Time: 
Aug 8th - 10th, 2012, 9:00-12:00, 13:00-16:00
Location: 
ML Damon Room
Speaker: 
Damian Rouson, Karla Morris, and Salvatore Filippone

Workshop Introduction:

Modern Fortran explicitly supports object-oriented programming (OOP). OOP aims to increase a program's maintainability in part by reducing cross-module data dependencies and to increase a program's reusability in part by providing for extensible derived types. Emerging compiler support for Fortran 2003/2008 inspires a more modern program design and implementation style. This course provides the requisite skills.

  • Day 1 introduces OOP in Fortran 2003.
  • Day 2 introduces patterns of best practice in program organization.
  • Day 3 explores several paths toward parallel OOP, including parallel numerical libraries and Fortran 2008 coarray programming.

Examples will utilize introductory-level numerical algorithms from linear algebra and differential equations inspired by multiphysics modeling that is coupled field problems common to many interdisciplinary, engineering, and physical science simulations.
   
Additional information, including a syllabus and schedule from the NERSC URL for previous offerings of the course can be found here:
http://www.nersc.gov/users/training/nersc-training-events/oop-fortran/

Reference:

If you are interested in purchasing the reference book Scientific Software Design - The Object-Oriented Way written by Damian Rouson(Sandia National Laboratories), Jim Xia(IBM Canada Lab, Markham), and Xiaofeng Xu(General Motors Corp.), here is a discount coupon.

Registration:

Please register for the workshop from the link here.

Speaker Description: 

Damian is the manager of the Reacting Flow Research Department and a Visiting Assistant Professor at the University of Cyprus. His research interests focus on scalable scientific software design. He has recently written a book on the subject, Scientific Software Design: The Object Oriented Way . He is also the software architect of the object oriented Fortran interface to Trilinos, ForTrilinos. ForTrilinos provides Fortran applications direct access to capabilities in the Trilinos project.

Karla is a senior member of technical staff in the Reacting Flow Research Department at Sandia. She is the lead developer of ForTrilinos. Her research interests include computational fluid dynamics to multiphysics flows and scientific software architecture.

Salvatore Filippone graduated from the University of Rome "Tor Vergata" where he is currently with the Dep. of Industrial Engineering. His main research interests are in algorithms for numerical linear algebra, their implementation on high performance computers and their application to engineering domains such as fluid dynamics and electromagnetism. From 1990 to 2001 he has been with IBM Corp. where he was one of the lead developers of the numerical libraries ESSL and PESSL. He evaluates scientific projects in HPC for both the European Commission and the National Science Foundation.

Aug. 8, 2012

"ISTeC at CSU is hosting the Front Range Consortium for Research Computing's (FRCRC's) Second Annual Front Range High Performance Computing Symposium, August 13-14, 2012, at Colorado State University, Fort Collins, CO 

www.frcrc.org/events/hpc-2012

The program is available on the above website. 

The FRCRC membership includes CSU, CU, Boulder, University of Wyoming, the Colorado School of Mines, NCAR, NREL, and NOAA. It includes HPC tutorials, birds of a feather, student presentations and more! This is a great opportunity to network with your colleagues in computational science on the Front Range and learn new skills.

Advance Registration is due by August 6th, 2012 at 12:00 noon. Registration at the door is not accepted. 

We hope you will attend. Please feel free to contact Rich Loft at loft@ucar.edu for more information."


Aug. 6, 2012

The CISL User Services Section has published additional documentation to help users prepare for computing with the Yellowstone, Geyser, and Caldera resources that are being installed and tested at the NCAR-Wyoming Supercomputing Center.

The information deals with scheduling, queue selection, and running jobs in the new environment. It includes a sample batch script and points out notable differences from scripts used with the Bluefire system.

More Yellowstone documentation will be provided soon.

Feel free to use the Feedback link on our Support & Training menu to let us know what you think.

Jul. 31, 2012

With the imminent arrival of Yellowstone and the expected large-scale data production, CISL is making changes to the HPSS environment. With these changes, HPSS will discontinue support for the dual-copy class of service, while providing users with a second-copy backup area for enhanced preservation of critical data.

These changes will happen in several steps. On August 1, 2012, a new user backup area (see below) will be made available to users. Then, on September 1, 2012, HPSS will discontinue support for the dual-copy class of service in favor of a user-controlled, second-copy strategy. Finally, and also starting September 1, the HPSS team will begin converting all data files stored in the dual-copy class of service to single-copy files.

Most existing dual-copy files are over one year old, dating from the prior Mass Storage System (MSS), which used dual-copy as the default class of service. HPSS has always used single-copy as its default, and since April 2011, very little new data has been stored as dual-copy. However, converting existing dual-copy files will recover 3 PB of space in support of Yellowstone-related data storage needs.

To replace the dual-copy class of service, HPSS will provide a backup area into which users can copy critical files for which they wish to have the extra level of protection. Files copied to the backup area will be guaranteed to be on separate tape cartridges. In addition, the backup area provides protection from user errors over and above the dual-copy class of service. Specifically, the backup copies are protected from user-initiated deletions or overwrites of the primary copies. In contrast, deleting or overwriting a dual-copy file deletes or overwrites both copies.

Interested users should submit a CISL Help request for their own directory in the backup area. Users can then use the cp command to copy data to the backup area. Files in the backup area will be charged against project allocations at the same rate as primary copies. Documentation for this new feature is available at http://www2.cisl.ucar.edu/docs/hpss/copies.

All users are welcome to take advantage of the backup area, but are encouraged to carefully consider which files need the extra protection. While data loss can happen, it is very rare. CISL's tape archive has had several media failures in the past year, but all files were recovered by the vendor. Over the tape archive's lifetime, there have been only a few instances of data being lost from media failure.

If you have questions about this change or whether your files may be affected, please contact cislhelp@ucar.edu.

Jul. 30, 2012

The Bluefire system will be supporting two significant training activities in late July and early August, and other users may experience slightly slower turnaround as a result. Students participating in these activities will have access to a higher-priority queue to ensure jobs can be completed in a timely fashion for the training exercises.

The heaviest student load will occur in the afternoons from July 30 to August 3 when the two-week Dynamical Core Model Intercomparison Project (DCMIP, see http://esse.engin.umich.edu/groups/admg/dcmip/) summer school and the weeklong 2012 Community Earth System Modeling tutorial (see http://www.cesm.ucar.edu/events/tutorials/073012/) overlap. Student usage will decrease accordingly during the second week.

We apologize for the inconvenience, and ask users to plan their work accordingly.

Jul. 27, 2012

NCAR's Computational and Information Systems Laboratory (CISL) invites NSF-supported university researchers in the atmospheric, oceanic, and related sciences to submit large allocation requests for the forthcoming petascale Yellowstone system by September 17, 2012. Revised instructions have been posted for the next round of Large University Allocations, and all requesters are strongly encouraged to review the instructions before preparing their submissions.

These requests will be reviewed by the CISL High-performance computing Advisory Panel (CHAP), and there must be a direct linkage between the NSF award and the computational research being proposed. Please visit http://www2.cisl.ucar.edu/docs/allocations for more university allocation instructions and opportunities.

Allocations will be made on Yellowstone, NCAR's new 1.5-petaflops IBM iDataPlex system, the new data analysis and visualization clusters (Geyser and Caldera), the 11-PB GLADE disk resource, and the HPSS archive. Please see https://www2.cisl.ucar.edu/resources/yellowstone for more system details.

For the much larger Yellowstone resource, the threshold for Small University Allocations has been increased to 200,000 core-hours; researchers with smaller-scale needs should wait until CISL begins accepting small allocation requests in September. However, university researchers with NSF awards can apply for a Small University Allocation on Bluefire or Janus to obtain information on the resource requirements for their research and to test the computational efficiency of their code before applying for a Yellowstone allocation.

Questions may be addressed to: David Hart, User Services Manager, 303-497-1234, dhart@ucar.edu

Jul. 22, 2012
CISL and IBM have made enough progress on the installation of Yellowstone that we can now provide more accurate information about the likely schedule for acceptance and availability of NCAR's new petascale system.

Nearly all Yellowstone hardware has been delivered and is being assembled in the NWSC machine room. In a major milestone, on July 5, Yellowstone was completely powered on. While no significant problems arose, a number of hardware issues have been noted and are being addressed. The number of hardware issues are well within expectations for a system the size of Yellowstone. IBM and CISL are continuing to install software and configure the administrative and user environments. We are now estimating that the installation activities will be completed by the end of July.

At that point, IBM and CISL will jointly conduct benchmarks and run the system with a simulated workload around the clock for three weeks, a period called the Acceptance Test Period (ATP). If Yellowstone successfully completes the acceptance tests, CISL will take over the system from IBM, and we expect at least two additional weeks for NCAR-specific configuration and customization, which may involve clearing all data from the new disk system (several times) to optimize file system performance.

Our best estimate for completing these activities pushes the start of Accelerated Scientific Discovery (ASD) projects to early September. The rest of the user community will have access as soon as practical after the ASD projects begin.

We will keep you posted as we proceed, particularly about significant changes in this schedule. Please contact cislhelp@ucar.edu with any questions.
 
 
 

Pages