- Make an environment with apptainer, it is the only program needed.
- Create a directory
$CONDA_PREFIX/rstudio-images - Download correct images:
apptainer pull docker://bioconductor/bioconductor_docker:devel - Copy the
run.shscript to$CONDA_PREFIX/binasrstudio-web, make sure it is executable. - Copy the
message.shscript to$cONDA_PREFIX/etc/conda/activate.d/, make the directory if not exist.
Note: Make sure your image have development tools for building some R packages.
- The RStudio server runs within an Apptainer container
- When launched, you'll receive:
- A randomly generated port number
- A randomly generated password for your session
- SSH tunneling instructions for remote access
- Direct URL for local access
- You can specify CPU cores and custom port numbers with options like
--cpusand--port
- Two temporary files will be created in your home directory:
custom_rsession.conf: Configuration for your R session- A symlink to
/vol/localat~/vol-local-entry-RSTUDIO
- These files are automatically removed when your session ends
- R packages are installed to
/vol/local/$USER/Rstudio/$R_VERSION - This location persists between sessions
- The R version is automatically detected from the container
- Using this location keeps your packages organized by R version
- Your R workspace defaults to
/vol/local/$USER/RStudio/workspace - This prevents clutter in your home directory
- The
.RDataand history files are stored in this location - This directory is created automatically if it doesn't exist
- By default, your session uses all available CPU cores
- You can limit CPU usage with
--cpus Nto use N cores - You can specify exact cores with
--cpus 0,2,4syntax - The system manages core allocations to prevent conflicts between users
- Allocated cores are automatically released when your session ends