Using RStudio

First, make sure you are in your workgroup, then launch an interactive job on a compute node. For example using workgroup it_css, account traine and specifying the standard partition, 2GB of memory and 1 hour of time, it would look something like this

[traine@login00.darwin ~]$ workgroup -g it_css
[(it_css:traine)@login00.darwin ~]$ SLURM_EXPORT_ENV=NONE salloc --mem=2G --time=1:00:00 --partition=standard
salloc: Pending job allocation 5406889
salloc: job 5406889 queued and waiting for resources
salloc: job 5406889 has been allocated resources
salloc: Granted job allocation 5406889
salloc: Waiting for resource configuration
salloc: Nodes r1n17 are ready for job 
[(it_css:traine)@r1n17 ~]

Once on a compute node the prompt changes accordingly, in this case the compute node name is r1n17. Now use VALET to load the RStudio package and the necessary software dependencies.

[(it_css:traine)@r1n17 ~]$ vpkg_require r-studio-server/4.2.1
Adding dependency `squashfs-tools/4.5.1` to your environment
Adding dependency `singularity/3.11.1` to your environment
Adding package `r-studio-server/4.2.1` to your environment

Once the RStudio software is loaded, then start the rserver within the Singularity container. This will initialize RStudio, allowing you to access it remotely through a web interface. It will look something like this

[(it_css:traine)@r1n17 ~]$ Sexec rserver --server-user=$USER
TTY detected. Printing informational message about logging configuration. Logging configuration loaded from '/etc/rstudio/logging.conf'. Logging to '/var/log/rstudio/rstudio-server/rserver.log'.
TTY detected. Printing informational message about logging configuration. Logging configuration loaded from '/etc/rstudio/logging.conf'. Logging to '/var/log/rstudio/rstudio-server/rsession-traine.log'.

Make a note of the compute node name, in this example r1n17, as it will be needed for when establishing the SSH tunnel.

If you are using Linux or MacOS, open the terminal on your laptop or desktop computer, and type

ssh -L 8787:r1n17:8787 traine@darwin.hpc.udel.edu

be sure to replace traine with your actual username, and substitute r1n17 with the compute node name where your interactive job is running and where you started rserver in the previous step. You will need to log in using your DARWIN credentials in this terminal. If you are doing this on Caviness, then you would use the same command but replace darwin.hpc.udel.edu with caviness.hpc.udel.edu.

If you are using a Windows machine, open PuTTY and enter the appropriate hostname

If you are doing this on Caviness, then you would use caviness.hpc.udel.edu instead of darwin.hpc.udel.edu. In addition to your standard connection settings, you'll need to set up an SSH tunnel. To do this, navigate to Connection → SSH → Tunnels in PuTTY. Then, add the Source Port to 8787 and the Destination to r1n17:8787 (remember to replace r1n17 with the compute node name where your interactive job is running and where you started rserver in the previous step).

Click Add to confirm the tunnel settings, then click Open to connect. You will be prompted to enter your username and password for your DARWIN (or Caviness) account depending on which hostname you used in the previous step.

Once SSH tunnel connection is established, then open a web browser on your laptop or desktop computer and navigate to the following URL

localhost:8787

You should see the RStudio session start. Once you have completed your work, exit the RStudio session by closing the browser, close the SSH tunnel terminal, stop the rserver on the compute node with CTRL-C, and then type exit to end the interactive job.