HPC Documentation

Guides, references, and tutorials for the WCM cluster

Rstudio

Interactive RStudio Server Deployment

This guide walks through spinning up an independent RStudio Server instance on an assigned compute node utilizing secure shell (SSH) local port forwarding. This workflow bypasses OpenOnDemand entirely, ensuring dedicated compute allocation access.

1. Initialize RStudio on the Cluster Node

Log into your cluster's edge entry point and invoke the automated deployment script directly on the login shell:

# Query and load your desired target RStudio environment module
module load rstudio/<version>

# Execute 
rstudio_run
Expected Terminal Confirmation Output:
RStudio Server has been successfully started on c0##

2. Establish the Cryptographic SSH Tunnel

Open a secondary terminal instance running completely on your local host laptop machine. Execute one of the following variations to map the cluster socket back to your loopback routing adapter:

Option A — VPN Proxy Gateways / Jump Host Routings

# BRB 
ssh -J [your_cwid]@scu-login02.med.cornell.edu -NL [port#]:localhost:[port#] [your_cwid]@c0##

# Cayuga 
ssh -J [your_cwid]@cayuga-login1 -NL [port#]:localhost:[port#] [your_cwid]@c00##

Option B — Preconfigured Local Config Aliases

If you maintain complex multi-proxy routing statements configured straight inside your workstation's local ~/.ssh/config matrix, call the shorthand configuration targets directly:

# BRB 
ssh -NL [port#]:c0##:[port#] c0##

# Cayuga
ssh -NL [port#]:c00##:[port#] c00##

3. Connect via Web Browser Workspace

  1. Open a standard web browser on your personal computer workstation.
  2. Navigate to the forwarded loopback web address: http://localhost:[port#]
Field EntryValue Token Identification
Username[your_cwid]
Password(Input the explicit alphanumeric secret string returned inside step 1 by rstudio_run)

4. RStudio Scripts

When tasks demand high core counts or long runtimes, bypass the logging head entirely by executing pre-written automation file parameters inside your custom storage directory structure:

# Clone the system templates out to your persistent scratch laboratory space
cp /path/to/rstudio_script.sh /athena/labname/scratch/[your_cwid]/my_rstudio.sh

System Template Paths for Slurm Customization:

  • BRB Cluster Blueprints:
    /software/apps/rstudio-server/sbatch/rstudio_script.sh
    /software/apps/rstudio-server/sbatch/rstudio_script-4.4.3-Seurat.sh
  • Cayuga Cluster Blueprints:
    /opt/ohpc/pub/software/rstudio/sbatch/4.4.3/rstudio_script.sh
    /opt/ohpc/pub/software/rstudio/sbatch/4.4.3/rstudio_script-4.4.3-Seurat.sh

SCU High-Performance Computing Technical Documentation — 2026.