sam-dynamodb-scanamo-seed.g8

A template project for quickly creating stateful serverless applications using the scanamo Dynamodb database access abstraction.

Introduction

The seed shows how to:

• setup a AWS::DynamoDB::Table resources
• how to handle get and post requests
• how to save and load using dynamodb tables

Usage

Create a new template project by typing:

sbt new dnvriend/sam-dynamodb-scanamo-seed.g8

Deploying

To deploy, you need to have your AWS account setup correctly:

• To deploy type: samDeploy,
• To remove type: samRemove,
• To get more info type: samInfo,
• To view the CloudFormation template type: samValidate.

For more information, please see the dnvriend/sbt-sam repository.

Using the example

Posting a person:

$ http post https://q2irbnzbj5.execute-api.eu-west-1.amazonaws.com/dnvriend/person name=dennis age:=42 lucky_number:=10
HTTP/1.1 200 OK
{
    "age": 42,
    "id": "410a5de5-cfbf-45c6-9fb5-4571c3899568",
    "lucky_number": 10,
    "name": "dennis"
}

Getting a person:

$ http https://q2irbnzbj5.execute-api.eu-west-1.amazonaws.com/dnvriend/person/1
HTTP/1.1 404 Not Found
{
    "msg": "person with id '1' not found"
}

$ http post https://q2irbnzbj5.execute-api.eu-west-1.amazonaws.com/dnvriend/person name=dennis age:=42 lucky_number:=10
HTTP/1.1 200 OK
{
    "age": 42,
    "id": "1977c477-3dc0-4af0-b072-30b7454fc433",
    "lucky_number": 10,
    "name": "dennis"
}

AWS::DynamoDB::Table resource configuration

The seed sam-component uses Typesafe configuration to setup resources. This seed focuses on the combination of AWS::Serverless::Function that uses an AWS::DynamoDB::Table resource for its persistence requirements.

In the base directory, a folder conf is created, containing a file sam.conf that defines AWS resources. sbt-sam plugin interprets the sam.conf and indexes the resources, in effect it 'knows' about AWS resources can in future releases will make use of this knowledge, over 'just' creating the CloudFormation template.

An example sam.conf that creates is:

dynamodb {
   // a simple table that has only a hash key 'id'
   People {
    name = people
    hash-key = {
      name = id
      type = S
    }
    rcu = 1
    wcu = 1
  }
  
   // a simple table that has only a hash key 'id', and 
   // has DynamoDB streams enabled with KEYS_ONLY
   People {
    name = people
    hash-key = {
      name = id
      type = S
    }
    stream = KEYS_ONLY
    rcu = 1
    wcu = 1
  }
  
    // a table that has has two keys, a hash 'name' and range key 'age' 
     People {
      name = people
      hash-key = {
        name = id
        type = S
      }
      range-key = {
        name = name
        type = N
      }
      rcu = 1
      wcu = 1
    }
    
  // a table that has has two keys, a hash 'name' and range key 'age' and
  // a global secondary index 
  People {
    name = people
    hash-key = {
      name = name
      type = S
    }
    range-key = {
      name = name
      type = N
    }

    global-secondary-indexes {
      people_id {
        hash-key = {
          name = id
          type = S
        }

        projection-type = ALL
        rcu = 1
        wcu = 1
      }
    }
    rcu = 1
    wcu = 1
  }

Defining multiple tables

To define multiple tables, just put them in the same dynamodb block, and give them a new name:

dynamodb {
   People {
    name = people
    hash-key = {
      name = id
      type = S
    }
    rcu = 1
    wcu = 1
  }
  Addresses {
      name = addresses
      hash-key = {
        name = id
        type = S
      }
      rcu = 1
      wcu = 1
  }
  Orders {
    name = orders
     hash-key = {
        name = id
        type = S
     }
     rcu = 1
     wcu = 1
  }

Table names

Because resources in a single AWS account has a flat namespace, a design decision has been made, in order to easily develop and deploy resources without having resource name clash. This means that sbt-sam knows about the name of a resource in the project, like eg. the person table, but will deploy the resource with a managed logical name with the following pattern:

[project-name][stage][resource-name]

For the table person, this means that the resource name will become:

sam-seed-dnvriend-people