DynamoDB Data Importer

!!! WARNING: This Cloudformation script will launch a DynamoDB table at 40,000 WCU this can be expensive if left for more than an hour !!!

What's in side the box

A NodeJS Lambda function streams a S3 File and then imports it into DynamoDB using the Batch API.

The full article for this code can be found here -> https://www.rehanvdm.com/serverless/dynamodb-importer/index.html

A note on the Environment variables passed to the function:

EXECUTE_LOCAL: If set uses the AWS profile specified on app.js line 6, currently set to "rehan', change to yours. This is only used when we locally test the function
CSV_BUCKET_NAME: Name of the bucket where the CSV is stored
CSV_KEY_NAME: Name of the CSV file in the bucket
DYNAMO_TABLE_NAME: DynamoDB table name where we will import into
CONCURRENT_BATCH_SUBMITS: Make reference to the article, this is the amount of concurrent batch api writes made to the DynamoDB table
READ_AHEAD_BATCHES: Make reference to the article, this is the amount of batches that will be read from the stream before they will be sent to DynamoDB at the CONCURRENT_BATCH_SUBMITS rate. This value must be greater or equal to CONCURRENT_BATCH_SUBMITS.
MAX_ROWS_SUBMIT: Used for testing, setting to 0 means now restriction else stop the import when this amount of rows has been imported.
AWS_NODEJS_CONNECTION_REUSE_ENABLED: AWS SDK NodeJS variable that allows the SDK to reuse TCP connections for faster API calls

Example:

EXECUTE_LOCAL: false
CSV_BUCKET_NAME: !Join ['', [!Ref AppName, '-source-', !Ref AppEnvironment]]
CSV_KEY_NAME: "data_s3.csv"
DYNAMO_TABLE_NAME: !Join ['', [!Ref AppName, '-table-', !Ref AppEnvironment]]
CONCURRENT_BATCH_SUBMITS: 20
READ_AHEAD_BATCHES: 40
MAX_ROWS_SUBMIT: 10000
AWS_NODEJS_CONNECTION_REUSE_ENABLED: 1

Install

Change the ENV variables in the SAM/CloudFormation template to suite your needs. You would prob only change MAX_ROWS_SUBMIT to 0 so that it imports everything after you did a small test with MAX_ROWS_SUBMIT amount of records.
In the /package.json file, use your bucket for SAM deployments so change the content in line 9 (_sam_package): --s3-bucket <YOUR EXISTING MANUAL CREATED S3 BUCKET NAME>
Run npm install
Runn the sam_deploy npm command to let SAM build, package and deploy your CloudFormation

Batteries included

Data generation tool

Using mocker-data-generator which abstracts over faker 2 CSV files are generated.

A CSV file that contains (25*41) + 1 records that will be used by the File test to test most code paths
A large CSV file containing 3,000,000 records that will be roughly around 250MB. Upload this file to your bucket before running the S3 test or the Lambda in the cloud.

Tests

Inside the /app-tests/lambda/importer/test-importer.js two tests can be found:

File: Streaming a small CSV from the file system and then doing the import
S3: Streaming a S3 CSV file and then doing the import