RStudio on the Cluster

1 Overview: Using RStudio on the Cluster | 2 Using | 2.1 Previous Issues

Overview: Using RStudio on the Cluster

RStudio is available on the cluster, and thus in principle can be loaded and used anywhere. To use it effectively:
- Do not load and use on the login nodes. This will affect ALL other users trying to log on and use the cluster.
- We recommend to use the SouthPass service using an Interactive Desktop session.
It is recommended to use RStudio to edit and update code in a GUI environment, and test/run simple jobs for debugging purposes. Some R packages, if they have the ability to use multiple cores, will simply use ALL the cores available on a node, so please explicitly define within your scripts to only test against one/two cores. Once you’re happy your code is working then use salloc/sbatch to use the computation nodes and to submit your much longer runs that use significantly more cores.

Using

Use the module name rstudio to discover versions available and to load the application.

You will need to load a version of R before starting RStudio. If you don’t then you’ll see the following error message regards “no R”:

[1240032:1011/144405.782798:ERROR:gl_factory.cc(120)] Requested GL implementation (gl=none,angle=none) not found in allowed implementations: [(gl=egl-angle,angle=default),(gl=egl-gles2,angle=none)].
[1240032:1011/144405.784510:ERROR:viz_main_impl.cc(186)] Exiting GPU process due to errors during initialization
/usr/bin/which: no R in (/apps/u/opt/linux/rstudio/2023.09.0/usr/lib/rstudio:/home/salexan5/bin:/home/salexan5/bin:/apps/s/projects/core_hour_usage/bin:/apps/s/arcc/1.0/bin:/apps/s/slurm/latest/bin:/apps/s/turbovnc/turbovnc-2.2.6/bin:/home/salexan5/bin:/home/salexan5/bin:/usr/local/bin:/usr/bin:/usr/local/sbin:/usr/sbin:/home/salexan5/.local/bin:/home/salexan5/bin:/apps/s/singularity/3.10.3/bin:/home/salexan5/.local/bin:/home/salexan5/bin)

Previous Issues

There was a previous issue with an earlier version of rstudio/2022.12.0 where if you had performed a module load before starting RStudio, any LD_LIBRARY_PATHS were not being picked up. This has been resolved in the later versions. Version 2022.12.0 has been removed and is no longer available.