14 VS Code Sessions (Preview)
RStudio Server Pro allows you to launch VS Code sessions from the home page via the Job Launcher, if configured. Users can start VS Code sessions that allow them to work with VS Code while still working within the administrative framework provided by RStudio, such as authentication, PAM session management, etc.
Currently, VS Code Sessions are a Preview feature. The feature itself is stable and usable, but you may find some bugs, and we are still working to complete some aspects of the VS Code development workflow. We highly encourage you to submit your feedback to let us know how we can improve! See the Professional Support section for more information on submitting feedback.
Note: Before VS Code sessions can be launched, the Job Launcher must be setup correctly. For more information, see the Job Launcher section.
14.2.1 General Installation
In order to add VS Code integration to RSP, you must first install the open-source
code-server wrapper, available at https://rstd.io/vs-code-server.
code-server is a wrapper which allows access to VS Code via a webserver in a browser, allowing RSP to create VS Code sessions and proxy them through the browser.
To install, simply run the command
rstudio-server install-vs-code <path to installation directory>. This will install the
code-server binary, the R and Python extensions, and automatically configure
/etc/rstudio/vscode.conf. For example, to install everything at
# likely need root privilege to install to /opt sudo rstudio-server install-vs-code /opt/code-server
Once the installation completes, you’ll need to restart the RStudio service for it to detect availability of VS Code sessions.
Note: You must also install code-server on any Launcher nodes (for Local or Slurm plugins) and within any containers used (if using Kubernetes).
If you’d like to install everything manually, simply download the
code-server Linux distributable and extract it to the desired location on the RSP host(s).
For example, to manually install
# create directory to house code-server mkdir /opt/code-server cd /opt/code-server # download the code server package wget https://rstd.io/vs-code-server -O vs-code-server.tar.gz # extract code-server binary tar zxvf vs-code-server.tar.gz --strip 1 # remove the archive rm vs-code-server.tar.gz
Note: At the time of writing, only code-server versions 3.2.0 and 2.x have been confirmed to work with RSP.
14.2.2 Installing Language Extensions
code-server does not provide access to the official Visual Studio Marketplace. There are three ways in which you can add extensions to
- Install the extension through
code-server’s hosted repository. This can be done from within a VS Code session via the UI (similar to how extensions are added in VS Code), or via command line with the following command:
/opt/code-server/code-server --extensions-dir /opt/code-server/extensions --install-extension <extension name>
- Download an extension in VSIX format from an online marketplace such as Open VSX. Then, install it using the
--install-extensioncommand like the example above, passing the VSIX file path as the extension name. Note that if the latest version of the extension is incompatible with the version of code-server, it will fail to install. You can try using an older version of the extension in this case.
Note: It is against VS Code’s Terms of Service to use the official Marketplace extensions with third party tools like
code-server. We strongly recommend using a free and open source alternative like Open VSX.
- Build the extension from source. There are several extensions freely available on GitHub that can be built into a VSIX file yourself and then installed via the
--install-extensioncommand. Building third party extensions from source is outside of the scope of this document.
184.108.40.206 Recommended Extensions
We recommend installing support for both R and Python. If you installed
code-server via the
rstudio-server install-vs-code command (noted above), the R and Python extensions were also installed. However, if you wish to install them manually, follow the instructions below.
To install the Python extension, download and extract the VSIX file into the correct format:
curl -L https://rstd.io/vs-code-python-ext -o ms-python.python-2020.5.86806.vsix.gz && gunzip ms-python.python-2020.5.86806.vsix.gz
Then, install the extension:
/opt/code-server/code-server --extensions-dir /opt/code-server/extensions --install-extension /path/to/ms-python.python-2020.5.86806.vsix
Similarly, to download and install the R extension:
curl -L https://rstd.io/vs-code-r-ext -o Ikuyadeu.r-1.1.0.vsix.gz && gunzip Ikuyadeu.r-1.1.0.vsix.gz /opt/code-server/code-server --extensions-dir /opt/code-server/extensions --install-extension /path/to/Ikuyadeu.r-1.1.0.vsix
Your extensions will now be available, but when configuring
args in /etc/rstudio/vscode.conf, make sure to add the
--extensions-dir= argument to whatever directory you installed your extensions at.
14.3.1 VS Code Configuration
Configuration of the VS Code feature is handled via the config file
/etc/rstudio/vscode.conf. Note that this file is not automatically created by RStudio, and must be created before being configured, and the VS Code feature is disabled by default. The following table lists the various configuration options that are available to be specified in the
vscode.conf configuration file:
|Config Option||Description||Default Value|
|enabled||Enables launching of VS Code sessions.||0 (disabled)|
|exe||Path to the
|args||Arguments to be passed to the code-server command. Note that this does not override the default value - if you wish to use some of the default arguments, you must contain the default arguments in the configuration value. It is strongly recommended that you do not change this unless you know what you’re doing! In almost all configurations, the
|session-clusters||Comma-delimited list of available Job Launcher clusters for launching VS Code sessions. Leave blank to specify all clusters.|
|default-session-cluster||The default Job Launcher cluster to use when launching a VS Code session.|
|default-session-container-image||The default container image to use when launching a containerized VS Code session.|
|session-container-images||Comma-delimited list of images that may be used for running VS Code sessions.|
|vscode-session-path||Path to the VS Code Session launcher executable/script. It is recommended that you do not change this unless you know what you’re doing, and you need to point to a different script.||/usr/lib/rstudio-server/bin/vscode-session-run|
|session-no-profile||Enables/disables running of bash profile scripts when starting VS Code sessions.||0 (run profile scripts)|
For example, your
vscode.conf file might look like the following:
exe=/usr/bin/code-server enabled=1 default-session-cluster=Kubernetes default-session-container-image=rstudio:vscode-session
14.3.2 Launcher Configuration
When creating containerized VS Code sessions via the Job Launcher, you will need to specify mount points as appropriate to mount the users’ home drives and any other desired paths. In order for sessions to run properly within containers, it is required to mount the home directories into the containers.
For more information, see Launcher Mounts. Note that you can specify the
VS Code to configure mount entries that should only be mounted for VS Code sessions.
14.3.3 Container Configuration
When running VS Code sessions in containers, such as by using the Kubernetes Job Launcher plugin, you will need to ensure that the image(s) used to launch VS Code sessions contain, at minimum, the following:
- code-server installation
- RSP session binaries
- If creating container users (see Server Configuration), you must have the
libuser-develpackages, depending on your platform to install the libuser library and development tools.
For ease of use, it is recommended that you use the
r-session-complete Docker image as a base for any VS Code session images you intend to create.
If you experience issues related to running VS Code sessions, you can use the Launcher verification tool which will attempt to launch VS Code sessions and provide diagnostic output about what could be going wrong. For more information, see the Troubleshooting section for the Job Launcher integration documentation.