This tar ball includes RCloud 2.4.0 including all required R package dependencies
frozen at their respective versions at the time of the release (tested with R 4.4.2
and R 4.2.2). Therefore it is intended to be used with a clean R installation
(tested on Debian 12 system with r-base-dev
), or as a container (see below).
For latest development and updates, please visit https://github.com/att/rcloud
This release can be used to build a Docker image with RCloud exposed at port 8080. It is based on Debian image and the file Makefile includes targets and can be used as reference for the appropriate docker commands. For example:
Build the image:
make build
Run the image:
make run
Once running, RCloud can be accessed at http://localhost:8080/login.R
A more complex setup with authentication and multi-user support can be built and
deployed using docker compose
:
docker compose build
docker compose up
The sample user login is rcloud
with password rcloud
(see runtime-sks
target to modify user management).
On a Unix machine with R (tested with R 4.4.2) RCloud can be run locally.
Although RCloud can be run from any location (the ROOT
environment variable just must
point to the root of the installation), for simplicity we assume this release
has been unpacked in /data/rcloud
.
The necessary R package dependencies can be installed from the release with
bin/rlib.sh
(this does not require network access). Some packages require specific system libraries, but most should be already installed if R was installed by the system package manager.
A most minimal configuration file is provided in conf/rcloud.conf
and the instance can be started with
bin/start.sh
and shut down either with the INT
signal or with
bin/shutdown.sh
RCloud is highly modular and has many optional features and services which extend its capabilities. The configurations used in the Quick Start sections are very minimalistic to reduce complexity, intended only for a quick test in single-user environments and not intended for any real deployment.
RCloud supports additional features such as secure connections, authentication, document search, distributed compute back-end, scaling and load-balancing, multi-instance support, GitHub and gist-service back-ends, Shiny and htmlwidgets-support, additional languages (Python and Jupyter kernels). Some features rely on additonal services or software not included in the release tar ball, but avaiable on GitHub.
For forther reading please see the documentation directory, deployment WiKi page and rcloud.conf configuration file documentation.
This release is based on a new Zig-based system - see README for deatils.