PM2 Setup and Deployment with Ecosystem Configuration

PM2 Setup and Deployment with Ecosystem Configuration

We announced it at NodeJS paris meetup and we did it! For the release of PM2 0.9.xa new awesome and simple feature will make your life much easier

We announced it at NodeJS paris meetup (here are the slides) and we did it! For the release of PM2 0.9.x, a new awesome and simple feature will make your life much easier.

We embedded a modified version of deploy from the Node.js Mozart, TJ Holowaychuk and we also refactored the JSON application declaration. Internally it’s a simple bash script that doesn’t need any external dependencies to be installed.

Here are the features of this embedded deploy system:

$ pm2 deploy <configuration_file> <environment> <command>

  Commands:
    setup                run remote setup commands
    update               update deploy to the latest release
    revert [n]           revert to [n]th last deployment or 1
    curr[ent]            output current release commit
    prev[ious]           output previous release commit
    exec|run <cmd>       execute the given <cmd>
    list                 list previous deploy commits
    [ref]                deploy to [ref], the "ref" setting, or latest tag

You can display the PM2 deploy help with:

pm2 deploy help

And dont forget that this feature is available on PM2 > 0.9, so update PM2 accordingly.

The ecosystem.config.js file

Now, a new file called by default ecosystem.config.js (previously ecosystem.json) will allow you to declare your applications/services and also the different hosts you want to deploy code into.

You can generate a sample ecosystem.config.js via the command:

pm2 ecosystem

This file will look like this:

{
  "apps" : [{
    "name"      : "API",
    "script"    : "app.js",
    "env": {
      "COMMON_ENV_VAR": "true"
    }
    "env_production": {
      "NODE_ENV": "production",
    }
  },{
    "name"      : "WEB",
    "script"    : "web.js"
  }],
  "deploy" : {
    "production" : {
      "user" : "node",
      "host" : "212.83.163.1",
      "repo" : "git@github.com:repo.git",
      "ref"  : "origin/master",
      "path" : "/var/www/production",
      "post-deploy" : "pm2 startOrRestart ecosystem.config.js --env production"
    },
    "dev" : {
      "user" : "node",
      "host" : "212.83.163.1",
      "repo" : "git@github.com:repo.git",
      "ref"  : "origin/master",
      "path" : "/var/www/development",
      "post-deploy" : "pm2 startOrRestart ecosystem.config.js --env production"
    }
  }
}

There is the apps section that describes your different applications/services to be run and the deploy section that describes where your code should be deployed.

As you can notice, there is also a env_<ENVIRONMENT> attribute that specify dedicated environment variable depending on where you gonna deploy your code.

As always, locally you can launch all theses services via:

pm2 start ecosystem.config.js

Deploy declaration

On each enviroment (production, dev…) there are some attributes configurable like these ones:

key /path/to/some.pem
user deployer
host n.n.n.n
repo git@github.com:visionmedia/express.git
path /var/www/myapp.com
ref origin/master
pre-deploy ./bin/something
post-deploy /var/www/myapp.com/update.sh

Customize / add the keys depending on your needs.

Important Note : by default the post-deploy is set with some logic dedicated to PM2.
Internally if this attribute is not set, the post-deploy will call pm2 startOrRestart ecosystem.config.js (startOrRestart)

Configuring deployment

On your target servers you must already have PM2 installed via:

npm install pm2@latest -g

At the same time have PM2 be able to start on server startup via:

pm2 startup ubuntu

Disconnect from your remote server and make sure that your remote server has your ssh keys. You can generate and copy them via:

ssh-keygen -t rsa
ssh-copy-id node@myserver.com

Once you have generated the ecosystem.config.js and edited it accordingly to your needs, you’re all set and ready to go!

Setup and Deployment

Setup

Important note: Commit your node_modules folder on your project Git repository. It’s a best practice when you deploy code to keep version consistency between all the environments.

First PM2 has to initialize your remote system. To do that:

pm2 deploy <environment_key> setup

This will create folders depending on the path attribute
Note: By default PM2 will look for a local ecosystem.config.js file, if you want to set another file do $ pm2 deploy <ecosystem_file> <environment_key> setup (for instance you can use the package.json instead!)

Deploying

Now PM2 has to grab the latest version of your app and start the applications configured in the ecosystem.config.js under apps attribute. To do that:

pm2 deploy <environment_key>

Done !

Other commands

With the ecosystem.config.js on the current folder:

# Update the code
pm2 deploy <environment_key> update

# Revert to [n] th commit
pm2 deploy <environment_key> revert 1

# Execute a command on the server
pm2 deploy production exec "pm2 restart all"

Conclusion

Hope this will be helpful to you! Some other links that might interests you:

Cheers!

With PM2 Harden your Node.js Workload and be Ready to Scale

Setup takes 2 minutes with no configuration change