Quick start
Requirements
Exoframe is not particularly demanding and consumes a maximum of ~50MB of RAM. The most intensive task from the CLI side is packaging the project and streaming it to the server, which doesn't significantly affect RAM usage but relies more on CPU and network resources.
Running the Exoframe Server on its own is also resource-friendly:
- Exoframe Server consumes approximately ~50MB of RAM.
- Traefik started along with the server consumes around ~60MB of RAM.
However, be aware that the execution of deployments results in (1) a new Docker image being built and (2) a new Docker container being started. Depending on your project's complexity, this may require a significant amount of resources during both steps, potentially leading to failed deployments. Note: If Docker goes out of memory during the build, you will not receive any specific error message; the deployment will simply fail. Therefore, it is recommended to run Exoframe on a server with at least 1GB of RAM.
Installation and Basic Usage
To use Exoframe, you'll need two components: the Exoframe CLI and the Exoframe Server.
Installing Exoframe Server
To install Exoframe Server, run our installation shell script on your server using the following command:
bash -c "$(curl -fsSL https://raw.githubusercontent.com/exoframejs/exoframe/main/packages/exoframe-server/tools/install.sh)"
The script will guide you through the basic setup process.
For detailed manual server installation instructions, please consult the Installing Exoframe Server section.
Installing Exoframe CLI
To install Exoframe CLI, you have two options:
-
Download one of the pre-packaged binaries from the releases page.
-
Install it using npm (requires at least Node 18.x):
npm install exoframe -g
Alternatively, you can run Exoframe via npx
:
npx exoframe --help
Once you have Exoframe CLI installed, ensure that Exoframe Server is deployed and running. Confirm its proper functioning by opening http://your.exoframe.url
in the browser; if you see the Exoframe logo, it is working as expected.
After installation, set the server endpoint in the CLI:
exoframe endpoint http://your.server.url
Then, log in with:
exoframe login
Deploying Your First Project
After logging in, you're ready to deploy your first project.
First, create a new Exoframe project configuration by executing the following command in the project folder:
exoframe init
Then, deploy your project by running the following command:
exoframe
Updating Deployed Projects
Exoframe provides an easy way to update deployed projects.
This can be accomplished by adding the --update
(or -u
) flag to the deploy command:
exoframe --update
Here's how it works:
- Exoframe deploys the new version of the specified project.
- Exoframe then waits for the new version to start up.
- Finally, Exoframe removes the old running deployments for the given project.
This feature can be combined with deployment tokens to achieve simple continuous deployment for your projects.
Next Steps
Discover more about Exoframe basics in the Basics section.