Advanced: Personal renv cache on RStudio Server (BlueCloud 2026 infrastructure)

Author

VLab Course Team

This independent note explains how any RStudio Server user on the infrastructure can use a shared/global renv cache without modifying Docker images or course repositories.

Why use a personal renv cache setting?

  • Avoid reinstalling the same packages for every project.
  • Speed up renv::restore() across projects by reusing binaries.
  • Keep full reproducibility (each project still uses its own renv.lock).

Option A - Set a user-level cache in ~/.Renviron

Create (or edit) a file named ~/.Renviron and add:

RENV_PATHS_CACHE=~/blue-cloud-dataspace/GlobalFisheriesAtlas/cacheRenv

Then restart R (Session → Restart R). From now on, all your projects will use this cache automatically.

Notes:

  • The path can be any directory you can read/write.
  • Using a location under ~/blue-cloud-dataspace/ makes the cache accessible across workspaces on the infrastructure (shared but possibly a bit slower than other folders from the workspace).
  • To avoid issues caused by cache corruption or accidental deletions (since the cache folders are shared), you can isolate your project libraries from the mutual cache using:
renv::isolate()

This command copies all the required packages directly into the renv folder of your project. Be cautious when running this locally, as it can significantly increase disk usage.

Option B - Set a project-level cache in <project>/.Renviron

If you prefer to scope the cache setting to a single project, create a file named .Renviron in the project root (next to .Rproj) with:

RENV_PATHS_CACHE=~/blue-cloud-dataspace/GlobalFisheriesAtlas/cacheRenv

Restart R inside the project. Only this project will use that cache.

Verify your configuration

Run these checks in the R console:

# What cache path is currently active?
Sys.getenv("RENV_PATHS_CACHE")

# Where does renv think the cache lives?
renv::paths$cache()

# Project environment status (optional)
renv::status()

If you change ~/.Renviron or <project>/.Renviron, restart R to apply.

Using the cache with projects

Inside any project (after opening the .Rproj):

renv::restore()   # will reuse packages from the shared cache when versions match

Packages required by renv.lock will be linked or installed once into the cache and reused next time.

Good practices & caveats

  • Keep the cache path stable; moving it breaks existing links.
  • If the cache becomes very large, you can clean unused entries with renv::cache_clean() (advanced).
  • The cache is per R version; upgrading R may create a new cache tree.
  • Shared locations improve reuse across projects and users, but network storage can be slower than local disk.

Notes on group-specific caches

In this course, we configured our cache in:

~/blue-cloud-dataspace/GlobalFisheriesAtlas/cacheRenv

because it is the one we use in the Vlab 5.

But you are free to define a different cache path that matches your own setup:

  • A cache specific to your group of projects
  • A cache specific to your VLab instance

This flexibility lets different groups share a cache internally while keeping separation from other workspaces.