aws-cdk-dynamodb-seeder

A simple CDK JSON seeder for DynamoDB

Why this package

Glad you asked!

Using AWS CDK for automating infrastructure deployments is an amazing way of integrating the development and operations into one process and one codebase.

However, building dev or test environments that come pre-populated with data can be tricky, especially when using Amazon DynamoDB.

How do I use it

Install using your favourite package manager:

yarn add aws-cdk-dynamodb-seeder

Example TypeScript usage

import { Seeder } from 'aws-cdk-dynamodb-seeder';
...
const myTable = new Table(stack, "MyTable", {
    tableName: "MyTable",
    partitionKey: { name: "Id", type: AttributeType.STRING },
});
...
new Seeder(stack, "MySeeder", {
    table: myTable,
    setup: require("./items-to-put.json"),
    teardown: require("./keys-to-delete.json"),
    refreshOnUpdate: true  // runs setup and teardown on every update, default false
});

For a more in-depth example, see: elegantdevelopment/aws-cdk-dynamodb-seeder-examples.

Importing seed data

Data passed into setup ("Items" to put) or teardown ("Keys" to delete) should be an array of objects (that are, in turn, representations of string to AttributeValue maps).

setup elements should use the format of params.Item from AWS.DynamoDB.DocumentClient.put()
teardown elements should use the format of params.Key from AWS.DynamoDB.DocumentClient.delete()

Versioning

We will attempt to align the major and minor version of this package with AWS CDK, but always check our release descriptions for compatibility.

We currently support

Internals

Behind the scenes we use an AwsCustomResource as a representation of the related table's seed state. The custom resource's event handlers invoke a Function to perform setup and/or teardown actions.

Deploying a stack

On deployment, we write copies of your seed data locally and use a BucketDeployment to write it to an S3 Bucket.

We then create the handler function and custom resource to field seed requests (the onCreate event will immediate fire as the stack deploys, reading the data from the bucket and seeding the table using AWS.DynamoDB.DocumentClient).

Updating a stack

On a stack update, the onUpdate handler is triggered when refreshOnUpdate is true.

This will run AWS.DynamoDB.DocumentClient.delete() on every teardown "Key" followed by AWS.DynamoDB.DocumentClient.put() on every setup "Item".

Destroying a stack

When the stack is destroyed, the event handler's onDelete function will be invoked, providing teardown is set.

This simply runs AWS.DynamoDB.DocumentClient.delete() on every teardown "Key" before destroying the Seeder's resources.

rheinardkorf/aws-cdk-dynamodb-seeder