Awesome

serverless-plugin-tables

This Serverless plugin makes adding tables to your service file easy.

Benefits

• Less boilerplate
• Common defaults
• Integrates with existing resources
• Handles deployment limits
  • Dynamo

Contents

• Installation
• Usage
• Providers
  • AWS
    • DynamoDB
  • Others

Installation

Install the dependency:

yarn add -D serverless-plugin-tables

Add serverless-plugin-tables to your serverless.yml file:

plugins:
  - serverless-plugin-tables

Usage

Add a tables property the resources in your serverless.yml file and define tables by name according to the provider specs.

service: my-service

plugins:
  - serverless-plugin-tables

provider:
  ...

resources:
  tables:
    MyTable1:
      ...
    MyTable2:
      ...

Plugin options

The plugin can be configured by defining a custom tables object in your serverless.yml file. Database specific options should be defined as properties under their database type, like dynamo. See database specs for related options. Example options serverless.yml:

custom:
  tables:
    dynamo:
      deploymentBatchSize: 5

Providers

Common properties:

AWS

Common properties:

<a name="aws-example"></a> Example

resources:
  tables:
    # Simple DynamoDB Table
    Music:
      partitionKey: Artist
      sortKey: SongTitle
      indexes:
        - name: GenreByTitleIndex
          partitionKey: Genre
          sortKey: AlbumTitle

    # Complex DynamoDB Table
    People:
      name: ${env:PEOPLE_TABLE_NAME}
      resourceName: FavoritePeopleDynamoTable
      type: dynamo
      partitionKey: personID
      sortKey: state
      readUnits: 5
      writeUnits: 5
      indexes:
        # Global Secondary Index
        - name: EmailIndex
          partitionKey: email
          projection: all
          readUnits: 2
          writeUnits: 2
        # Local Secondary Index
        - name: PersonByCreatedTimeIndex
          sortKey: 
            name: createdTime
            type: number
          projection: keys
          readUnits: 2
          writeUnits: 2
        # Local Secondary Index with projection
        - name: PersonByAgeIndex
          sortKey: 
            name: age
            type: number
          projection:
            - dob
            - firstName
            - lastName
          readUnits: 2
          writeUnits: 2
      streamType: newItem
      ttlKey: expirationTime
      encrypted: true
      pointInTimeRecovery: true
      tags:
        STAGE: test
        TEAM: backend      
      template: 
        # Override the computed CF template
        Properties: 
          ProvisionedThroughput:
            ReadCapacityUnits : 1

DynamoDB

Type: dynamo

Note that DynamoDB tables default to using on-demand billing mode.

Options

Properties:

<a name="footnote-dynamo-provisioned-units">[1]</a>: Both read and write units are required if one is defined

<a name="dynamo-keys"></a> Keys:

Keys can be a string or an object. If a string is provided, then that will be the key name and it will be of data type string.

<a name="dynamo-data-types"></a> Data Types:

Corresponds to DynamoDB Data Types

<a name="dynamo-indexes"></a>Indexes:

Indexes can be Global or Local indexes. The difference being that Local indexes share the same partition key as the table. Therefore, to create a Local index, just omit the partitionKey field.

<a name="footnote-dynamo-index-key">[2]</a>: At least one key is required

<a name="footnote-dynamo-index-units">[3]</a>: Required if defined for the table

<a name="dynamo-stream-types"></a> Stream Types:

Corresponds to DynamoDB Stream Types

Others

If your provider or database isn't supported, open an issue to request it!