Installation_Prerequisites

Connector Installation Requisites

If your connector's Lambda function(s) will run in a VPC, you need to ensure that the Lambda functions will have network connectivity to its dependencies, which are connector specific but can include:

1. Amazon S3 - for spilling large responses.
2. AWS Glue DataCatalog - for metadata (if your connector is Glue enabled).
3. Amazon Athena - for checking query status and preventing overscan.
4. AWS Secrets Manager - for resolving any secrets (e.g. password) that you need for your connector.
5. AWS KMS - for generating data keys to encrypt large responses that spill to S3. (you can optionally use a local key source)
6. Ability to contact your source system (e.g. MySQL, Hbase master, etc...) - for federating to the source system.

Most of the time you can resolve the above issues by adding VPC endpoints for the required services or running your connector in a VPC that has an internet gateway. The way you solve this problem is dependent on your network topology. The safest option is usually adding a VPC endpoint since that doesn't require internet access.

In addition to network connectivity requirements, your connector will require a S3 Bucket and Prefix to spill response data too. This data is encrypted by default and can optionally use KMS for data keys. For more information on what may be spilled to S3, please check our FAQ.

Connector Development Requisites

In order to develop, build, or publish a connector for later deployment to Lambda you will need:

1. maven 2.5.x installed
2. aws cli
3. aws sam build tool
4. Java 8+

If you are using Cloud9, AmazonLinux, CentOs, or Redhat you can either use these instructions to develop inside a container_env here or start with our environment prep script to see the specific commands you may need to run to setup your development environment. This script wont work for MacOS or Windows but many of the commands have analogs on those operating systems.