Awesome

Serverless Import Plugin

Split your serverless.yaml config file into smaller modules and import them.

By using this plugin you can build your serverless config from smaller parts separated by functionalities. Imported config is merged, so all keys are supported and lists are concatenated (without duplicates).

Works on importing yaml files by path or node module, especially useful in multi-package repositories.

Installation

Install with npm:

npm install --save-dev serverless-import-plugin

And then add the plugin to your serverless.yml file:

plugins:
  - serverless-import-plugin

Usage

Specify config files to import in custom.import list:

custom:
  import:
    - ./path/to/serverless.yml # path to YAML file with serverless config
    - ./path/to/dir # directory where serverless.yml can be find
    - module-name # node module where serverless.yml can be find
    - '@myproject/users-api' # monorepo package with serverless.yml config file
    - module-name/custom-serverless.yml # path to custom config file of a node module

custom.import can be also a string, when only one file needs to be imported:

custom:
  import: '@myproject/users-api'

Relative paths

All function handler paths are automatically prefixed by the imported config directory.

functions:
  postOrder:
    handler: functions/postOrder.handler # relative to the imported config

For other fields you need to use ${dirname} variable manually. ${dirname} points to a directory of imported config file.

custom:
  webpack:
    webpackConfig: ${dirname}/webpack.config.js

Customizable boilerplates

In case you want to customize imported config in more dynamic way, provide it as javascript file (serverless.js).

module.exports = ({ name, schema }) => ({
  provider: {
    iamRoleStatements: [
      // ...
    ],
  },
  // ...
})

You can pass arguments to the imported file using module and inputs fields:

custom:
  import:
    - module: '@myproject/aws-dynamodb' # can be also a path to js file
      inputs:
        name: custom-table
        schema:
          # ...