Skip to content

FF chevron

FAQ: Cluster & General Issues

This page provides answers to some frequently asked questions about the campus HPC clusters in general. Please review this page before asking HPC staff for help.

What are the high-performance computing clusters?

The Division of Information Technology provides high-performance computing (HPC) cluster(s) for general campus use. This beowolf cluster consists of hundreds of x86_64 based compute nodes configured for running large-scale calculations in support of the campus research community. The Zaratan cluster is the current flagship cluster, designed especially for parallel computation.

Who owns/runs the cluster?

The Zaratan cluster was purchased with funds from the Division of Information Technology at the University of Maryland along with contributions from various colleges, departments, and research groups. Contributing groups receive high priority allocations, replenished quarterly, based on the amount of CPU time their contribution added to the cluster.

The cluster is managed by the Division of Information Technology, and campus IT staff handle all the hardware and system issues, and maintain a large software collection for the users.

What is the Allocations and Advisory Committee (AAC)?

The Allocations and Advisory Committee (or AAC for short) is composed of faculty representing colleges and/or departments that have contributed hardware to the Zaratan cluster. These faculty members have extensive experience in using HPC clusters for research. This group provides oversight, sets policy, and allocates computational resources to campus researchers.

Where can I find detailed documentation on the clusters?

This website provides extensive information on the DIT maintained HPC clusters. This includes:

What are the advantages of joining one of the campus clusters as opposed to starting my own?"

There are several benefits of using campus-managed High-Performance Computing (HPC) clusters over setting up and maintaining individual ones.

  • Reduced Burden: The Division of IT handles setup, maintenance, power, and cooling, thus, freeing researchers to focus on their work.
  • Increased Flexibility: Researchers can run larger, more parallel jobs than their individual contributions might allow, especially for testing or urgent deadlines. If you build your own cluster, your jobs are necessarily restricted to the resources you purchased. If you contribute to the campus cluster, you will receive an allotment of SUs each quarter year based on the amount contributed. While this will essentially allow you can run for the entire quarter using the same amount of resources as in the case wherein you built your own cluster, if you only run jobs for half of the quarter, you can use twice the amount of resources, and so forth.
  • Risk-Free Exploration: Researchers can apply for a free developmental allocations to explore HPC techniques and whether and to what degree it can benefit their research.

How do I contribute to one of the campus clusters?

We encourage anyone interested in contributing to the Zaratan HPC cluster to begin by requesting a free allocation from the campus Allocations and Advisory Committee (AAC) if you do not already have such --- this will enable you to "test-drive" the cluster before committing funds.

There typically is some excess capacity in the cluster; purchases of resources from this excess capacity can be handled rather quickly based on our current pricing list. If you are interested in making such a purchase, open a ticket iwth HPC staff](mailto:hpcc-help@umd.edu) explaining what resources are desired and when.

Note

Purchased resources are specific to specific quarters of the year (with quarters starting 1 Jan, 1 Apr, 1 Jul, and 1 Oct). Unused resources (e.g. SUs at the end of a quarter do not rollover to the next quarter; they simply vanish. If the allocation is still unexpired, it will be replenished at the start of the quarter with the resources purchased for the new quarter.

Purchases from the excess capacity of the cluster can generally be allocated relatively quickly, but are limited to the amount of (and type of) resources in the excess capacity pool.

If you have larger needs, or needs for hardware not present on the cluster, we can work with you. Such requests will require a dialog, please open a ticket with HPC staff to start this dialog. Please include as much information as you can about your requirements. As we will need to purchase additional hardware, this will take time.

Please see the section on paid allocations in the knowledge base for more information.

How can I get help using the clusters?

Here's a summary of the HPC cluster support structure:

  • System Staff Focus: The system staff primarily handles basic access and system usage issues. They are not experts in specific research software applications or fields of study.
  • Application Support: For questions about research applications or field-specific inquiries, researchers should consult with their colleagues, either in person or on-line.
  • Support Resources:
    • Usage documentation is the primary resource for answering questions.
    • System-related issues can be addressed by opening a ticket.
    • Application-related questions may be addressed through the hpcc-discuss mailing list.

How should I acknowledge the use of one of the clusters in papers, etc.?

Maintaining a first class HPC environment is expensive, and the Division of IT requests that you acknowledge your use of our clusters in papers or other publications of research which benefited from this campus resource. Such acknowledgements assist us in convincing University management of the value of this resource to campus, and helps us to obtain funding for its continued maintenance and/or future expansion.

To acknowledge your use of the cluster, we request that you use this wording.

The authors acknowledge the University of Maryland supercomputing resources (http://hpcc.umd.edu) made available for conducting the research reported in this paper.

We would also appreciate if you would update your list of publications to which the use of HPC resources have contributed; this list is also helpful to convince University management of the need for this resource.

I just received email about my home directory being over quota. What does this mean?

Here is our storage policy for home directories on the Zaratan HPC cluster:

  • Limited Storage: Home directories are for essential files, not large datasets. They prioritize reliability over speed and have limited space compared to other storage options like Lustre.
  • Soft Quota: A 10 GB soft quota is enforced. This is not a hard limit -- users can temporarily (up to one week) exceed this soft limit by up to another 10 GB (or 20 GB total). However, if you exceed this 20 GB limit, or if you exceed the 10 GB for more than a week, you will no longer be able to store any more data in your home directory. As many applications, etc. use the home directory for storing some configuration, etc. data, a full home directory can break many things, including the use of the OnDemand portal.
  • Email Warnings:
    • "Friendly notice: Your home directory on zaratan is over quota" emails are sent daily when the quota is exceeded, starting the 7-day grace period.
    • "URGENT: You are OVER your quota on Zaratan homespace" emails are sent after the grace period, indicating a policy violation.
  • Consequences of Overuse: Failure to reduce usage after the "URGENT" email may result in contacting the user's advisor or suspending cluster access. System staff are notified of these urgent emails.
  • Recommendations: Users should promptly remove unnecessary files, transfer data off the cluster, or move active research data to Lustre or other appropriate storage.