High Performance Computing

Information on using the
LSUHSC-NO cluster

Getting Started:

Tigerfish is a Linux based cluster running the CentOS operating system. It is accessed via Secure Shell (SSH) and is behind our institutional firewall.  In order to access Tigerfish, you must first have your account created. This cluster is available to all LSUHSC-NO researchers and our research partners. To gain access, please use the "Forms" tab on this page to request your account be created. Once you have your account, you will be able to log into the cluster.

If you are on campus, you must be connected to our LSU-Secure network on Wi-Fi, or an on-campus Ethernet port.  If you are off campus, then you need to initiate a remote connection to our VPN using Pulse Secure.  Once you are successfully connected to our network, you will be able to reach Tigerfish.  Simply open your Terminal (on Mac or Linux machines) or cmd (on Windows) and enter the command:

ssh user@tigerfish.lsuhsc.edu    ðŸ¡¨ Replace user with your own username

 

You will be prompted to enter your password to authenticate.  The first time you log in, you will use the password provided to you by IT Security when your account was created.  Once you have successfully logged in to Tigerfish for the first time, you should change your password. You can do this by issuing the command "passwd" (no quotes) into the Tigerfish terminal. You will be prompted to enter your current password, and then to enter your new password and confirm your new password.  Please ensure that you use a sufficiently long and complex password to keep the system secure. Use the command "exit" (no quotes) to disconnect from the Tigerfish cluster. 
 

Running Jobs:

Your initial login to Tigerfish places you in a shell prompt on the head node (tigerfish.lsuhsc.edu).  This node of the cluster is responsible for allocating cluster resources.  You should not be running any compute intensive commands or jobs on the head node itself as this will overburden the head node and degrade cluster performance for everyone.  The SLURM scheduler is used to request cluster resources to run compute intensive jobs and provides a queue for allocating resource requests.  All computing jobs should be issued through the srun (for interactive requests) or sbatch (for batch requests) commands.  These commands are issued from the head node and place the requested job in a queue until resources are available and allocated to run the job.  Resources for learning to use the SLURM scheduler are available here.

 

   


Appropriate Cluster Behavior:  

Usage: Tigerfish is a shared institutional resource and should only be utilized for work-related activities.  Protected health information is prohibited from being transferred to or stored on the cluster as this is not a system with enterprise security.  You should only be using de-identified research data and never place any protected health information on the cluster.  In general, you should consider any data stored on the cluster to be potentially readable by others and let this guide your decision of whether a dataset is appropriate for use on Tigerfish.

Data Storage:  Tigerfish is a computational cluster, not a data storage server.  While the cluster does have a substantial amount of secondary storage available to users, this should not be used as a place to park large data sets that are not being computed on.  Please be courteous and delete large data sets from Tigerfish after computing activities have been completed.  Users who are found to be consuming large amounts of storage will be warned and subject to having data deleted or disk quotas enforced.

Data Persistence and Backup:  Tigerfish is not a data storage server, and any data you have stored on Tigerfish could be lost at any time due to unexpected system failures.  The data stored on Tigerfish is not backed up.  You are responsible for your own data storage and backup of your data.  It is highly recommended that you transfer your analysis results back to your own lab's data storage and maintain your own backups of that data once analyses have been completed on a given data set.  Please do not rely on the data stored in your Tigerfish account to be the only copy of your data as this could disappear without warning at any time due to hardware failures.

Be a Team Player:  Tigerfish is a shared institutional resource and in order to promote ease of use we have not put in place disk quotas, limitations on resource utilization, password complexity requirements, or other usage restrictions.  We are relying on our user community to be good citizens and remain cognizant that this is a shared institutional resource.  Please take care to choose sufficiently complex passwords to ensure the security of your login credentials and don’t become a resource hog.  Account usage will be monitored, and any misbehaving users will be warned about excessive resource usage prior to having restrictions placed on their account or having credentials revoked if repeated requests are unsuccessful in correcting these issues.