Search Google Appliance


ARC Frequently Asked Questions

Use the following frequently asked questions to find answers to your questions and troubleshoot common problems.

General FAQ

How do I get a Research account?

Contact ARC to request a research account.

How do I request/create/access my lab's research share?

To request access to a research share, please fill out the I Drive Access Form.

What do I do if I forgot my password?

Access to ARC's webservers and compute servers requires a research shell account. This shell account uses the same password and username as your normal PSU Odin credentials. If you need to reset your Odin password, please visit the Odin Computer Account page.

Access to the Linux clusters (parallel compute) requires a research shell account and a cluster shell account. Both of these accounts use your normal PSU Odin credentials. If you need to reset your Odin password, please visit the Odin Computer Account page.

Access to the database server requires a database account. This account has its own password. If you need to reset your database password, please contact ARC.

How I do access to my lab's research share from off campus?

Directions for various operating systems can be found the Network File Shares.

What is my research share quota and how much am I using?

To check your Research Share quota and usage ssh to research.pdx.edu and run the following command: "df -h /vol/share/{your_share_name}".

Can I increase my research share quota?

Quota increases can be requested by contacting ARC. Please include the reason for needing the increase.

What is the policy for Long Running Processes (LRP) on the research server?

We ask that users nice their processes and be mindful that the research compute servers are a shared resource. ARC reserves the right to kill or renice processes that become a detriment to other users. Find more information at Long Running Processes.

Cluster FAQ

What do I need to run Parallel Jobs on the Rocks and Gravel clusters?

A PSU Odin ID and a research account. To request a research account, send an email to consultants@pdx.edu.

Is it possible to run serial jobs on the clusters?

No, we do not allow serial jobs to run on the clusters. Use a research machine instead.

What are the space limitations of the clusters?

Each node on Gravel has 8GB of RAM (1GB per core) and each cluster has several hundred gigabytes of shared disk space. The clusters are not a storage space, and the accounts are occasionally cleaned as needed.

How do I run a job on the clusters?

See our page on Using Linux Clusters.

Why doesn't my job run?

Submit a bug report to consultants@pdx.edu. For this, you must precisely describe the cause of failure. Attach the output and the error file as additional information, as well as any relevant output from the showq and qstat commands.

How long can I keep my result data on the clusters?

This is mainly size dependent. If the amount of free space goes under 10%, we will clean the home directories. It is recommended that any information you wish to save be moved to another location, such as a research share. If you need help doing so, send an email to consultants@pdx.edu.

Can I develop my own MPI applications?

Yes, we support the MPICH, OpenMPI, and PGI compilers. However, ARC does not support MPI based application development.

Further Resources

For more information on research computing, contact Academic & Research Computing.

You can also contact the Helpdesk for additional assistance.