Faculty Quickstart
Signing Up
As a Faculty:
Any CWRU faculty can get an HPC account as the Principal Investigator (PI).
Faculty may use the AMARA portal (any browser with Single Sign On + Duo authentication) to request a cluster account, as well as to manage the list of lab members that may utilize the resources associated with the HPC account. Faculty may also designate one or more Senior Personnel as managers with permission to add or remove members from their HPC account.
Increased Resources and Membership
If you are getting HPC Membership or additional storage, you can find details regarding Compute Resources and Increased Membership Storage Limits, Additional Storage Services offered and how to make payments with CWRU speedtype through iLAB Billing on our website.
As a Research Associate, a Postdoctoral Researcher or a Student:
If you are the above, you would need to get an account through a CWRU faculty. Please ask the faculty to add you once their HPC account is set up properly. The addition can be accomplished by using the online portal AMARA. Please note that only Faculty can apply, a postdoc or a student does not have the permission to access the portal for the application. After the account is created, the faculty or a designated manager can manage the account membership. Other members of the account would be able to access the portal to check their available resources.
Training Workshops, Tutorials, and Helpful References
We have a few training workshops every semester, and we always have the HPC Intro session early in the Fall/Winter semester. Please try to follow the session live, if you can. You can find more information about the workshops at Workshop & Events.
We also have HPC Tutorials that can help you accomplish some of the basic tasks on the cluster.
You can refer to Helpful references for more related and advanced topics.
Accessing the Cluster
You need the Case VPN to connect to the cluster when you are not on the campus network or when you use the CaseGuest wifi. VPN (with DUO security) can be obtained from vpnsetup.case.edu. VPN connections are set to terminate after 24 hours, or if the computer goes to sleep.
The login nodes are meant for development and for organizing jobs to run on the compute nodes. If we find jobs running on the login nodes, they will be stopped to maintain cluster accessibility. Always use "srun" or "sbatch" command to submit your jobs to the cluster. You can kill all your processes accidentally running on the login node with:
pkill -u <caseID>
kill -9 `ps -ef | grep <caseID> | grep -v grep | awk '{print $2}'`
If you submit multiple jobs by mistake, cancel all the jobs using:
scancel -u <caseID>
If you are accessing the cluster via command line: you can use Terminal (Mac/Linux) and MobaXterm (Windows).
If you need to use a graphical interface on the cluster see Graphical Access.
You can also access the cluster on the web using HPC OnDemand.
Quick Start Guide
You can find the User Quickstart Guide page to help you log on the cluster and run a simple application.
Other Helpful Tips
Use keyword search for the topics of your interest in this website.
If you are the first time HPC user, please go through HPC documentation along with HPC FAQs before sending email to hpc-support[AT]case[DOT]edu.
The default linux shell is Bash, but you can change it by following "Changing Linux Shell @HPC."
You can perform interactive work directly on a compute node using the srun command (e.g. "srun --x11 --pty bash"). An interactive session may be used for interactive job submission, running graphically-intensive packages (e.g. MATLAB), or to run scripts.
For high memory jobs, see the section "High Memory Job" at HPC Interactive and Batch Submission.
If your job is creating large number of output files or you have the case where the number of files in a directory is huge, follow the Panasas Storage Guideline for Huge Directory.
By default, Linux has permission set to provide read access to only the user and the group, and not to the other HPC users. The owner, however, can change the access as desired.