Running Nodejs on OpenShift picture
OpenShift is an excellent platform for hosting Nodejs applications. And, taking advantage of OpenShift's free hosting, or automated build and scaling features, usually only requires one or two small changes to your code.

This post answers the question, "What makes a Nodejs application OpenShift-compatible?", in two simple steps:

Step 1: Include a package.json file

All nodejs applications should include a package.json file in the root of their project. OpenShift and npm are both configured to check your (scripts.start and main) fields for information on how to launch your server:

"scripts": {
"start": "node server.js"
},
"main": "server.js"

Also, make sure to document your project's npm module dependencies in the dependencies section to ensure that any missing deps can be automatically resolved during the application's automated build phase.

Using npm install with the --save flag will help keep your project's package.json file up to date as you build:

npm install --save PACKAGE_NAME

As usual, local development environments should be initialized by running npm install followed by npm start.

Step 2: Read Configuration Details from the System Environment

OpenShift's Node.js cartridge automatically publishes server connection information to your application's environment via the following environment variables: OPENSHIFT_NODEJS_PORT and OPENSHIFT_NODEJS_IP

Include a small check to test for the presence of these configuration strings to ensure your project's portability:

var server_port = process.env.OPENSHIFT_NODEJS_PORT || 8080
var server_ip_address = process.env.OPENSHIFT_NODEJS_IP || '127.0.0.1'
 
server.listen(server_port, server_ip_address, function () {
console.log( "Listening on " + server_ip_address + ", port " + server_port )
});

This example shows how to quickly check for OpenShift-provided configuration strings while providing sensible default values for local development purposes.

MongoDB users should include some additional checks for environment-supplied credentials:

//provide a sensible default for local development
mongodb_connection_string = 'mongodb://127.0.0.1:27017/' + db_name;
//take advantage of openshift env vars when available:
if(process.env.OPENSHIFT_MONGODB_DB_URL){
mongodb_connection_string = process.env.OPENSHIFT_MONGODB_DB_URL + db_name;
}

Modules like config-chain can also be useful for normalizing this information down into a single variable, helping to provide a consistent interface for all remaining code references.

Additional application keys and secrets can be abstracted from your code using the same technique:

rhc env set GA_TRACKER=UA-579081
rhc env set SECRET_KEY=0P3N_S0URC3
rhc env list

More notes on using environment variables are available in our post on using external services like RedisCloud

That's it! -- If you have committed the changes from our first two steps, your application should now be ready to run on OpenShift.

Now, Document Your Process!

You can spin up a fresh copy of your publicly-hosted nodejs source from the terminal, using our rhc command-line tool:

rhc app create YOUR_APP_NAME nodejs-0.10 --from-code=http://github.com/USER_NAME/REPO_NAME.git

Or, launch a new instance of your app right from your browser by providing your project's source repository url during the configuration step of OpenShift's web-based app creation process.

The source url and list of service dependencies can also be encoded in a link, allowing you to offer a guided application launching workflow to new users, coworkers, or anyone else who would like to quickly spin up fresh hosted copies of your app in a few simple clicks.

Our post on using GitHub-style Ribbons to Launch applications is a nice follow-up for open source project maintainers that would like to encourage end-users, and community members to spin up their own clones from an existing hosted app.

Regardless of which process you choose, make sure to document it in your project's README file for others to follow.

What's Next?

{{cta('1ba92822-e866-48f0-8a92-ade9f0c3b6ca')}}