Instance, Reader, and Custom endpoint support

[rawalpuneet]

Utility should work for instance endpoints and cluster endpoint the same way. Looks like utility only produces output with cluster endpoint. Customer may want to just use reader endpoint dedicated to graphql in production environments.

[Cole-Greer]

Thanks for opening up the issue, I agree that the ability to restrict the AppSync API to a single reader would be a useful addition.

This can currently be achieved manually by opening the data sources in AppSync, finding the data source lambda function, opening the lambda functions configuration tab, and editing the NEPTUNE_HOST environment variable.

[triggan]

Changing the request to be a broader. We also need to investigate splitting mutation/query APIs across both cluster and reader endpoints respectively. This would allow reads to scale beyond a single instance.

We also do not currently have the ability to support custom endpoints beyond a user manually modifying the NEPTUNE_HOST environment variable in the associated lambda function.