Skip to content

Other

Q: "What is 'Disk Quota Exceeded'"?

A: This message refers to one or more file quotas being reached.

  • Users are advised to check their quotas regularly with showquota.
  • SOLUTION: Clear out the problem directories of any unnecessary files.

Extra Tips:

  • Some files may be hidden or stored deep within your subdirectories.
  • Hidden files can be seen with the ls -la or tree -a commands.

  • Many users may have hidden folders such as .local, .cache, and .vscode-server which takes up a lot of files in $HOME quota. A solution is to move these folders to $SCRATCH and create a symlink to it from the $HOME directory which can be done with the following commands:

    1. cd $HOME
    2. mv .local $SCRATCH
    3. ln -s $SCRATCH/.local
    4. Repeat steps b & c for .cache and .vscode-server

  • The following commands will show you the number of files within each top directory:

% ml mpifileutils/0.9.1-intel-2019a
% dwalk mydir
[2020-06-03T17:33:49] Walking /scratch/user/netid/mydir
[2020-06-03T17:33:49] Walked 896 items in 0.015395 seconds (58200.092672 files/sec)
[2020-06-03T17:33:49] Items: 896
[2020-06-03T17:33:49]   Directories: 12
[2020-06-03T17:33:49]   Files: 884
[2020-06-03T17:33:49]   Links: 0
[2020-06-03T17:33:49] Data: 115.775 GB (134.110 MB per file)
  • The following command will show you the number of files within the current directory (first column, ignore second and third): find . | wc

Q: "What is the 'Socket timed out' sbatch error"?

A: This occurs when the Slurm scheduler is temporarily overloaded due to hundreds of jobs being submitted in a short period of time. Usually it is one or a few users who submit jobs at the rate of hundreds of jobs per minute using a script or command loop. The Slurm scheduler will become responsive again after it has had a chance to schedule the high number of jobs but for the sake of all users, avoid submitting jobs using this approach. If you are submitting more than 100 jobs, you should understand that using the TAMUlauncher or a job array will reduce the scheduling load on Slurm. Contact the HPRC helpdesk if you would like assistance in creating a single job script instead of submitting hundreds of jobs per minute.

Q: "Why does my program stop after 1 hour on the login nodes?"

A: Since the login nodes are resources which are constantly shared by many users, we must enforce limits on computing on the login nodes in order to prevent irresponsible usage. One of these limits is on CPU time. Users are limited to ONE HOUR of CPU time per login session. If you need more than one hour of CPU time, you will need to submit a job to the batch system. More information on batch processing can be found on our batch processing page. You are expected to be responsible and courteous to other users when using software on the login nodes.

Q: "How do I acquire an HPRC account for a Texas A&M credit-bearing course?"

A: Details for creating HPRC accounts for use in connection with a Texas A&M University credit-bearing course can be found on our KB: Hosting a Class with HPRC

Q: "How can I add output to my .bashrc without breaking anything?"

A: Avoid messing with your .bashrc if at all possible. However, if you must add output to your .bashrc to print every time you log into a machine, add the following boolean:

if [ "$SSH_TTY" ] then   # Put output here fi

Q: "How do I set up two-factor authentication?"

A: In order to set up two-factor authentication (needed to access the TAMU VPN and log into CAS), go to https://duo.tamu.edu/. There will be an option to "Enroll/Manage Devices". Clicking this option will bring you to the CAS login page, where you enter your NetID and password. Next, you will add a device to use for authentication. Enter the requested information. Download the Duo Mobile app (Google Play/Apple App Store) on your registered device and log in. Now, whenever you log into a device through CAS, a request will be sent to your registered device asking for approval of the log in.

A: Your account will be locked if you failed to provide the correct password after 7 attempts. It will be unlocked after 10 minutes. Please note that there are multiple login nodes, and each node has its own counter for failed attempts. Instead of the general login, you can directly log in to a specific node; for example, rather than using ssh grace.hprc.tamu.edu, you might use ssh grace1.hprc.tamu.edu. Be cautious as certain MobaXterm setups might cause problems when duo authentication is enabled. Please refer to Access page for more information.

Q: "Application xyz failed to connect to the cluster, what can I do?"

A: Starting November 4th, 2019, Duo authentication is required for every SSH login request. This has been known to cause issues with certain applications. As of this writing, we don't have a complete list of which applications will have issues when duo authentication is enabled. Please send us an email at help@hprc.tamu.edu when you encounter login-related issues with the software you are using. We will try our best to assist you.

Q: "How do I unsubscribe from the TAMU HPRC email list?"

A: As of right now, the only way to unsubscribe from our mailing list is to deactivate your HPRC account.

Back to top