Skip to main content

Announcements

Read announcements for NREL’s high-performance computing (HPC) system users. 

Requesting Debug Partition Nodes on Eagle

July 10, 2019

In order to request a node on debug partition, a job needs to explicitly request for that partition using either -p debug or --partition=debug in your job submissions. The limits, type and number of nodes in the debug partition are as follows. More information on the debug partition can be found on the HPC site. Continue reading

Peregrine End-Of-Life

July 10, 2019

Peregrine will be decommissioned July 31, 2019. Start moving your data from Peregrine.

Peregrine decommissioning includes the Peregrine /scratch, /projects, /nopt and /home filesystems.  All data (/home, /projects, /scratch) for Peregrine will need to be moved prior to July 31st or data will be lost.

Continue reading

HPC Intro Course for interns and mentors

June 04, 2019

This interactive course will cover basics for using NREL HPC resources. An active NREL HPC user account and a computer will be required.

If you do not already have an account, you can request one here: https://www.nrel.gov/hpc/user-accounts.html

This course will be held in ESIF B308 - Insight Center Visualization Room

The two hour course will consist of two Sessions (you are welcome to attend one or both of the sessions):

Continue reading

Eagle and Mass Storage System Outage Starting Monday, July 8th

June 04, 2019

Eagle and the Mass Storage System will be unavailable starting Monday, July 8th so the HPC Operations team can complete required maintenance.  During the multi-day system time improvements will be made to Eagle's workload management (Slurm) and Lustre to fix certain types of job failures, fixes will be made for Jupyterhub to allow better access for projects, and we will replace the front end DAV nodes with nodes that have better visualization capabilities. Continue reading

Eagle Utilization is Increasing!

June 04, 2019

Users may have become accustomed to queue wait times being nearly nonexistent on Eagle over the past several months as it has been in use concurrently alongside Peregrine. As projects have migrated their software stacks to Eagle and Peregrine is being phased out, the majority of HPC usage is now taking place on Eagle. As a result, your jobs may now be subject to higher queue wait times as resources get managed by the job scheduler due to more frequent resource occupancy in overlapping job submissions. Here are some command-line tools to see information about Eagle occupancy and the queue status of your job(s): Continue reading

Why am I being "taxed" 3× to use Eagle instead of Peregrine

June 04, 2019

Eagle represents the state-of-the-art architecture as of 2018, whereas Peregrine does the same for ~2013. All aspects of computational hardware have advanced substantially in the interim, so we'll look at a few below, and how those might impact the amount of work you can get done given an hour on an Eagle node, vs. an hour on a Peregrine node. Continue reading

Transferring files between filesystems using Globus

June 04, 2019

Globus is the recommended method to transfer files between clusters.  

For example, to copy/transfer files from Peregrine /scratch to Eagle /projects use nrel#globus as the source and nrel#eglobus1 as the destination.  Then for the source path specify /scratch/<your_path> and for the destination path specify /projects/<your_path>.

If you need to move data (rather than copy) between /scratch and /projects on Eagle, using the "mv" command is the most efficient way to move files between those two filesystems.

Below is a table that shows a breakdown of which Globus endpoints are associated with which cluster.

Continue reading

The Fiscal Year (FY) 2020 HPC User Call

May 03, 2019

The Fiscal Year (FY) 2020 HPC User Call will go out on May 13, 2019. Requests must be submitted by June 14, 2019. Continue reading

Home File System Quotas

May 03, 2019

A gentle reminder that the Eagle /home directory of all HPC user accounts is limited to a size of 50GB. We are still implementing advanced space usage reporting features, so for the time being you can quantify your /home size by running du -hs ~ (which will take a minute or two to complete). The most common space-usage observed is due to large Anaconda virtual environments. Environments can be installed to /scratch and /project directories to avoid /home directory consumption. The Anaconda cache also stores many zipped python packages which can grow to be several GB–you can empty the cache by running conda clean --all and following the terminal prompts. Continue reading

HPC Operations Phone Support

April 03, 2019

Starting Monday, April 8th, HPC support will be available by phone. This will give you the ability to have quick support for basic functionality issues and questions.

HPC Contact Methods:
HPC Operations Support Phone Number – 303-384-6100
HPC Operations Support Email – HPC-Help@nrel.gov
Contact Us Form - https://www.nrel.gov/hpc/contact-us.html Continue reading