Stampede User News

Stampede2 Maintenance 17 July 2018

Posted by Mitchell Collins-Bailey on Jul 3, 2018 4:00:53 PM

Stampede2 will be unavailable from 8 a.m. to 7:30 p.m. (CT) on Tuesday, 17 July 2018. System maintenance will be performed during this time.   If you submit a job before the maintenance, and the time you request exceeds the time remaining until the maintenance begins, your job will run when the...

Updated on Jul 17, 2018 7:35:28 PM

Stampede2 is back in full production. 

Original Posting

Stampede2 will be unavailable from 8 a.m. to 7:30 p.m. (CT) on Tuesday, 17 July 2018. System maintenance will be performed during this time.

 

If you submit a job before the maintenance, and the time you request exceeds the time remaining until the maintenance begins, your job will run when the maintenance is over. The squeue command will report "ReqNodeNotAvailable" ("Required Node Not Available"). The showq utility will list the job as "BLOCKED" and report its status as "WaitNod" ("Waiting for Nodes"). Note that the hours leading up to the maintenance are an excellent time to submit shorter, smaller jobs that can complete before the maintenance begins: as the queues drain there will be many nodes available, and your wait time may be short.

TACC Institute: Visualizing and Interacting with Data

Posted by Charlie Dey on Jul 11, 2018 2:36:39 PM

We're pleased to announce the next institute in TACC's Summer Institute Series,  Visualizing and Interacting with Data, July 30 - August 3, 2018. Every dataset tells a story, in this institute, you will spend a week with TACC's visualization experts learning how to generate rich, informative...

We're pleased to announce the next institute in TACC's Summer Institute Series,  Visualizing and Interacting with Data, July 30 - August 3, 2018.


Every dataset tells a story, in this institute, you will spend a week with TACC's visualization experts learning how to generate rich, informative visualizations for your simulation results and data analyses to find that story. You will get a a solid start with VisIt and Paraview, learning the basics of developing interesting and worthwhile geographic and information visualizations, and try your hand at using some of the latest augmented- and virtual-reality tools for manipulating and exploring you visualizations.
 

Computational Techniques in Life Sciences Summer Institute

Posted by Charlie Dey on Jun 29, 2018 3:40:52 PM

July 16th - 20th, 2018 Announcing the next institute in TACC's Summer Institute Series: Computational Techniques in Life Sciences. Learn the basic skills necessary to develop and execute scientific workflows on large-scale systems, run parallel analyses, use Bash and Python to chain common life...

July 16th - 20th, 2018
Announcing the next institute in TACC's Summer Institute Series: Computational Techniques in Life Sciences.

Learn the basic skills necessary to develop and execute scientific workflows on large-scale systems, run parallel analyses, use Bash and Python to chain common life sciences applications together to form more complex workflows, and analyze result data in order to gain useful insights. Spend time one-on-one with TACC's Life Sciences Computing group discussing how to accelerate your existing research or how to start a new project by beginning with the best computational practices. The skills developed from this institute are vast and can be applied to various other domain sciences.

TACC Institute: Data Intensive Computing and Machine Learning

Posted by Jason Allison on Jun 27, 2018 11:32:41 AM

July 16th - 20th, 2018 Want to gain more insight from your data? Is analyzing datasets too much for your favorite spreadsheet? Considering using machine learning to improve insights? Spend a week with TACC's data analysis experts learning about the latest tools and techniques for analyzing and...

July 16th - 20th, 2018

Want to gain more insight from your data? Is analyzing datasets too much for your favorite spreadsheet? Considering using machine learning to improve insights? Spend a week with TACC's data analysis experts learning about the latest tools and techniques for analyzing and processing your Big Data problems.

Try your hand at using R, Jupyter notebooks, Hadoop, and Spark with guided hands-on exercises, spend time one-on-one with our data analysis staff talking about your own projects and goals, and hear about the latest tools and research in Big Data.

For more information and access to registration, please visit the following link:
https://www.tacc.utexas.edu/education/institutes/data-intensive-computing-machine-learning

TACC Institute: Parallel Programming Foundations

Posted by Jason Allison on May 18, 2018 4:38:30 PM

June 25th - June 29th, 2018 Ready to parallelize your software? Spend a week with TACC’s high performance computing experts learning to write parallel applications using OpenMP and MPI, profiling and optimizing sequential and parallel applications, debugging parallel software, developing software...

Updated on Jun 15, 2018 1:48:42 PM

We still have spaces available for the Parallel Programming Foundations Institute - June 25th through June 29th, 2018. Please register ASAP to avoid missing out on this opportunity.

Original Posting

June 25th - June 29th, 2018

Ready to parallelize your software? Spend a week with TACC’s high performance computing experts learning to write parallel applications using OpenMP and MPI, profiling and optimizing sequential and parallel applications, debugging parallel software, developing software for many-core architectures, and more. Gain practical experience developing parallel applications with guided hands-on exercises, spend time with TACC experts one-on-one talking about your own research project goals and roadblocks, and hear from TACC staff, users, and collaborators about the latest trends, technologies, and research projects related to large-scale parallel computation.

Topics covered include:
- OpenMP SIMD Directives
- MPI Collectives
- Hybrid OpenMP/MPI Programming
- Profiling and Debugging

For more information and access to registration, please visit the following link:
https://www.tacc.utexas.edu/education/institutes/parallel-programming-foundations

Stampede2 Maintenance June 12, 2018

Posted by Matthew Edeker on Jun 4, 2018 10:52:26 AM

Stampede2 will be unavailable from 8 a.m. to 7:30 p.m. (CT) on Tuesday, 12 June 2018. System maintenance will be performed during this time.   If you submit a job before the maintenance, and the time you request exceeds the time remaining until the maintenance begins, your job will run when the...

Stampede2 will be unavailable from 8 a.m. to 7:30 p.m. (CT) on Tuesday, 12 June 2018. System maintenance will be performed during this time.

 

If you submit a job before the maintenance, and the time you request exceeds the time remaining until the maintenance begins, your job will run when the maintenance is over. The squeue command will report "ReqNodeNotAvailable" ("Required Node Not Available"). The showq utility will list the job as "BLOCKED" and report its status as "WaitNod" ("Waiting for Nodes"). Note that the hours leading up to the maintenance are an excellent time to submit shorter, smaller jobs that can complete before the maintenance begins: as the queues drain there will be many nodes available, and your wait time may be short.

Stampede2 Status June 1, 2018

Posted by Matthew Edeker on Jun 1, 2018 11:39:09 PM

The Stampede2 filesystem currently has one storage target offline and administrators are working to correct the problem. Users may experience hangs or pauses trying to access files on this storage target. An update to this notification will be sent once the filesystem target has been recovered.

Updated on Jun 2, 2018 8:02:23 AM

The previously unavailable storage target on the /scratch filesystem has recovered and is back in normal operation now.



Original Posting

The Stampede2 filesystem currently has one storage target offline and administrators are working to correct the problem. Users may experience hangs or pauses trying to access files on this storage target. An update to this notification will be sent once the filesystem target has been recovered.

TACC Institute: Computational Science in the Cloud

Posted by Jason Allison on May 18, 2018 4:25:31 PM

June 4th - June 8th, 2018 Spend a week with TACC’s web and cloud computing experts learning about tools, techniques, and best practices for doing high throughput computing, using virtual machines, containers, and recipes to ensure that your runtime environment is reproducible, even if you need to...

Updated on May 29, 2018 10:14:07 AM

We still have spaces available for the Computational Science in the Cloud Institute.  Please register ASAP to avoid missing out on this opportunity.  The registration deadline for this event has been extended to May 30.

Original Posting

June 4th - June 8th, 2018

Spend a week with TACC’s web and cloud computing experts learning about tools, techniques, and best practices for doing high throughput computing, using virtual machines, containers, and recipes to ensure that your runtime environment is reproducible, even if you need to scale out. Be prepared to take the lessons you learn at TACC back to your own research project with hands-on exercises and one-on-one time with our experts talking about the specifics of your own research project and goals. Hear about the latest tools and research in distributed and cloud computing with invited talks from fellow users, collaborators, and guests.

Topics covered include:

- Introduction to Cloud Concepts
- Virtual Machine Construction using Jetstream
- Making API Calls in Python
- Docker & Singularity Containers
- Automated Deployment Tools


For more information and access to registration, please visit the following link:
https://www.tacc.utexas.edu/education/institutes/computational-science-in-the-cloud

Stockyard WORK Status May 17 2018

Posted by Jacob Getz on May 17, 2018 2:28:47 PM

$WORK file-system is experiencing intermittent issues.  This may cause some jobs to fail and /work calls to stall.  Administrators are currently working to restore system functionality.

 An additional announcement will be posted when $WORK is fully operational.

-TACC Team

Updated on May 17, 2018 3:40:46 PM

Work is back in full production. Please report any issues you encounter to the helpdesk.


-TACC Team

Original Posting

$WORK file-system is experiencing intermittent issues.  This may cause some jobs to fail and /work calls to stall.  Administrators are currently working to restore system functionality.

 An additional announcement will be posted when $WORK is fully operational.

-TACC Team

Stampede2 Status May 15, 2018

Posted by Jacob Getz on May 15, 2018 9:19:17 AM

Stampede2 /work filesystem is currently unavailable.  TACC staff are currently working to resolve the issue.

-TACC Team

Updated on May 15, 2018 9:46:27 AM

The Stampede2 /work filesystem has been restored. 

Original Posting

Stampede2 /work filesystem is currently unavailable.  TACC staff are currently working to resolve the issue.

-TACC Team

Stampede 2 status May 8, 2018

Posted by David Littrell on May 8, 2018 12:30:04 PM

The Stampede 2 /work filesystem is down for maintenance and should be back up shortly. 

Updated on May 8, 2018 2:10:38 PM

The /work filesystem has been restored. 

Original Posting

The Stampede 2 /work filesystem is down for maintenance and should be back up shortly. 

Introduction To MPI Using The Interactive Parallelization Tool (IPT)

Posted by Jason Allison on May 7, 2018 1:05:37 PM

In parallel codes MPI (Message Passing Interface) is the predominant method for communicating data between the nodes of a compute cluster. It is widely used and performs well on a large variety of hardware platforms, and therefore a basic understanding of MPI programming is beneficial to anybody...

In parallel codes MPI (Message Passing Interface) is the predominant method for communicating data between the nodes of a compute cluster. It is widely used and performs well on a large variety of hardware platforms, and therefore a basic understanding of MPI programming is beneficial to anybody working in the field of high-performance computing (HPC). In this beginner-level training session, we will provide an overview of the basic concepts of MPI. We will introduce the trainees to the Interactive Parallelization Tool (IPT) that is designed for parallelizing serial C/C++ programs semi-automatically. The participants in the training session will be introduced to MPI and will learn to use IPT for parallelizing their C/C++ applications. 

Prerequisites: Experience working in a Linux environment, and familiarity with C/C++/Fortran or other programming language. 

To register and for more information please visit: https://learn.tacc.utexas.edu/ 

If you have any questions please contact me at jasona@tacc.utexas.edu

Stampede Status 2 May 2018

Posted by Sergio Leal on May 2, 2018 2:58:31 PM

Stampede2 /work filesystem experienced an interruption in service an hour ago.  It is back in production.  

Stampede2 /work filesystem experienced an interruption in service an hour ago.  It is back in production.  

Stampede Status 1 May 2018

Posted by Sergio Leal on May 1, 2018 10:35:41 AM

Stampede scratch filesystem is currently unavailable.  TACC staff are currently working on resolving the issue. 

Updated on May 1, 2018 1:26:01 PM

Stampede2 scratch filesystem is back in full production.

Original Posting

Stampede scratch filesystem is currently unavailable.  TACC staff are currently working on resolving the issue. 

Stampede2 is now back in production 4 April 2018

Posted by Mitchell Collins-Bailey on Apr 27, 2018 5:48:01 PM

Stampede2 is now back in production. Thank you for your patience. To better serve the Stampede2 user community, we’ve made several important changes to the queues: A new “long” queue allows you to run KNL jobs for up to 96 hrs. We’ve slightly relaxed the development queue limits. There are now...

Stampede2 is now back in production. Thank you for your patience.


To better serve the Stampede2 user community, we’ve made several important changes to the queues:

  • A new “long” queue allows you to run KNL jobs for up to 96 hrs.
  • We’ve slightly relaxed the development queue limits.
  • There are now fewer nodes in the flat-quadrant queue; we’ve made a corresponding change to the associated queue limits.

See the Stampede2 User Guide (portal.tacc.utexas.edu/user-guides/stampede2) especially Table 5 (portal.tacc.utexas.edu/user-guides/stampede2#table5) for more information.

Stampede2 Log in issues 16 april 2018

Posted by Mitchell Collins-Bailey on Apr 16, 2018 10:09:17 PM

Stampede 2 is currently experiencing log in issues.  This is currently being addressed.  This should not impact any jobs currently being run or in queue.  Notifications will be sent when solution is found.

Updated on Apr 17, 2018 2:56:13 AM

Stampede2 home filesystem is back in full production.

Original Posting

Stampede 2 is currently experiencing log in issues.  This is currently being addressed.  This should not impact any jobs currently being run or in queue.  Notifications will be sent when solution is found.

Stampede2 Extended Maintenance 23 April 2018

Posted by Matthew Edeker on Apr 10, 2018 4:18:40 PM

Stampede2 will be unavailable for a five-day period from Monday, 23 April 2018 at 8 am CDT to Friday, 27 April 2018 at 7:30 pm CDT. During this extended downtime, TACC staff will conduct system maintenance, apply file system updates, and run tests that exercise the entire system. Please submit any...

Stampede2 will be unavailable for a five-day period from Monday, 23 April 2018 at 8 am CDT to Friday, 27 April 2018 at 7:30 pm CDT. During this extended downtime, TACC staff will conduct system maintenance, apply file system updates, and run tests that exercise the entire system.

Please submit any questions you have via the TACC User Portal.

Thank you.

Stampede2 Status 31 March, 2018

Posted by Matthew Edeker on Mar 31, 2018 9:39:51 AM

Stampede2 is currently inaccessible. TACC staff are working to resolve the issue as quickly as possible.

Updated on Mar 31, 2018 10:25:27 AM

Stampede2 is back in full production.

Original Posting

Stampede2 is currently inaccessible. TACC staff are working to resolve the issue as quickly as possible.

Training: OpenMP Training Events - April 12th and 13th, 2018

Posted by Jason Allison on Mar 22, 2018 12:19:22 PM

We are pleased to announce the following OpenMP training events are being offered to both in-person and webcast participants April 12th and 13th, 2018. The courses include hands-on exercises on TACC systems. Local participants are strongly encouraged to attend in person. Instructors will be...

We are pleased to announce the following OpenMP training events are being offered to both in-person and webcast participants April 12th and 13th, 2018. The courses include hands-on exercises on TACC systems. Local participants are strongly encouraged to attend in person. Instructors will be available after class to consult on individual projects with in person participants.

4/12/18 9am-12:30pm CT - Introduction to OpenMP
This course will introduce participants to the OpenMP threading model, and describe the basic constructs necessary to parallelize loops on multi-core architectures. Topics include the fork/join threading model, using OpenMP directives, and loop parallelization. The fundamentals of hybrid computing (MPI & OpenMP) will be explained and illustrated.

4/13/18 9am-12:30pm CT - Advanced OpenMP
This course will provide an introduction to OpenMP optimization techniques for multi-core and vectorized architectures. Topics will include OpenMP SIMD directives, configuring OpenMP thread affinity, tasking, and task dependences.

To register and for more information please visit: https://learn.tacc.utexas.edu/

If you have any questions please contact me at jasona@tacc.utexas.edu

Stampede2 Maintenance 20 March 2018

Posted by Matthew Edeker on Mar 12, 2018 11:14:37 AM

Stampede2 will be unavailable from 8 a.m. to 7:30 p.m. (CT) on Tuesday, 20 March 2018. System maintenance will be performed during this time.   If you submit a job before the maintenance, and the time you request exceeds the time remaining until the maintenance begins, your job will run when the...

Updated on Mar 21, 2018 1:59:37 AM

Stampede2 is back in production after the system maintenance.  


Thanks,
TACC Administration

Updated on Mar 20, 2018 7:35:10 PM

Today's filesystem work is taking longer than expected and the Stampede2 maintenance will be extended.

Original Posting

Stampede2 will be unavailable from 8 a.m. to 7:30 p.m. (CT) on Tuesday, 20 March 2018. System maintenance will be performed during this time.

 

If you submit a job before the maintenance, and the time you request exceeds the time remaining until the maintenance begins, your job will run when the maintenance is over. The squeue command will report "ReqNodeNotAvailable" ("Required Node Not Available"). The showq utility will list the job as "BLOCKED" and report its status as "WaitNod" ("Waiting for Nodes"). Note that the hours leading up to the maintenance are an excellent time to submit shorter, smaller jobs that can complete before the maintenance begins: as the queues drain there will be many nodes available, and your wait time may be short.

Stampede2 Status, March 8, 2018

Posted by David Littrell on Mar 8, 2018 3:22:00 PM

One of the /home1 Lustre targets is currently unavailable.  TACC Staff are working to resolve the issue as quickly as possible. Please stay tuned to TACC User News for further updates.

Updated on Mar 8, 2018 4:09:48 PM

/home1 is now back in full production.

Original Posting

One of the /home1 Lustre targets is currently unavailable.  TACC Staff are working to resolve the issue as quickly as possible. Please stay tuned to TACC User News for further updates.

TACC Maintenance 11 March, 2018

Posted by Mitchell Collins-Bailey on Feb 26, 2018 5:32:29 PM

Access to all TACC systems will be unavailable from 9:00 AM CDT  until 2:00 PM CDT on March 11, 2018 to allow for upgrades to the TACC core network hardware. Jobs will continue to run, but users will have no access to TACC services and systems until the upgrade is complete.

Access to all TACC systems will be unavailable from 9:00 AM CDT  until 2:00 PM CDT on March 11, 2018 to allow for upgrades to the TACC core network hardware. Jobs will continue to run, but users will have no access to TACC services and systems until the upgrade is complete.

TACC Maintenance 11 March, 2018

Posted by Mitchell Collins-Bailey on Feb 26, 2018 5:28:56 PM

Access to all TACC systems will be unavailable from 9:00 AM CDT  until 2:00 PM CDT on March 11, 2018 to allow for upgrades to the TACC core network hardware. Jobs will continue to run, but users will have no access to TACC services and systems until the upgrade is complete.

Access to all TACC systems will be unavailable from 9:00 AM CDT  until 2:00 PM CDT on March 11, 2018 to allow for upgrades to the TACC core network hardware. Jobs will continue to run, but users will have no access to TACC services and systems until the upgrade is complete.

Stampede2 will be unavailable from 8 a.m. to 7:30 p.m. (CT) on Tuesday, 20 February 2018

Posted by Mitchell Collins-Bailey on Feb 12, 2018 4:29:35 PM

Stampede2 will be unavailable from 8 a.m. to 7:30 p.m. (CT) on Tuesday, 20 February 2018. System maintenance will be performed during this time.   If you submit a job before the maintenance, and the time you request exceeds the time remaining until the maintenance begins, your job will run when the...

Stampede2 will be unavailable from 8 a.m. to 7:30 p.m. (CT) on Tuesday, 20 February 2018. System maintenance will be performed during this time.

 

If you submit a job before the maintenance, and the time you request exceeds the time remaining until the maintenance begins, your job will run when the maintenance is over. The squeue command will report "ReqNodeNotAvailable" ("Required Node Not Available"). The showq utility will list the job as "BLOCKED" and report its status as "WaitNod" ("Waiting for Nodes"). Note that the hours leading up to the maintenance are an excellent time to submit shorter, smaller jobs that can complete before the maintenance begins: as the queues drain there will be many nodes available, and your wait time may be short.