Skip to content

Latest commit

 

History

History
55 lines (43 loc) · 2.09 KB

Deployment.md

File metadata and controls

55 lines (43 loc) · 2.09 KB
Raw

Automatic deployments

Servers can be deployed to automatically using GitHub deploy hooks.

Config

Before a server can be deployed to, you must configure it in server/config/<env>.json with a 'deploy' attribute, containing:

tags: An array of strings that allow your server to be identified as a deploy target, e.g. ['ad_staging', 'small_instance']. Tags can be useful to deploy to various servers at the same time, along with the more standard server name, which is set in the top-level configuration attribute 'server.name'. NB: hyphens (-) are not allowed in tags.

github authentication: To notify about deployment, your server must authenticate with github. To do this, create an auth token on GitHub: http://developer.github.com/v3/#authentication Then, modify your config to include these credentials:

  "deploy": {
    "tags": ["lovely-staging-server"],
    "github": {
      "password": "x-oauth-basic",
      "username": "<your-auth-token>"
    }
  }

GitHub Deploy hooks

Servers must be configured to listen to the creation of tags on Github:

  1. Add a WebHook service hook that points at your server's deploy route (http://youdomain.com/deploy).

Github will notify the server when a new tag is created. At this point, the server will inspect the tag name to see if it matches:

<server-name>-<new-feature-name>-<id>

If is the same as 'server_name' specified in config/<env>.json, the application will automatically pull the new code and update the server's local repository. Make sure you are running your application with forever or it will not restart after a deploy.

The deploy command

To deploy, run the deploy command on your local machine:

cd server && npm run-script deploy

This will ask for the tags (or server name) you want to deploy to, and what the feature introduces, and then create a tag. Once this tag has been pushed, the Github deploy hooks will take care of streaming the status of the spawn deployments back to the deploy command, ending with a summary of the deployment(s).