I entered this DynamoDB feedback in an AWS survey, and figured I'd repeat it here:

We are building with DynamoDB using single-table design as a best practice.

Doing the design manually is hard. Really hard.

There has to be a way AWS can help with this through automation.

/1

#awswishist

We specify the data and access patterns, and AWS software creates the DynamoDB single-table design.

Or we add the data and run for a while with poor performance, while AWS monitors the access patterns and incrementally optimizes the single-table table design.

/2

It feels like we are in the mainframe days when programmers had to design the layout of files on the disk platter cylinders.

New system software automated that menial work for us and we never looked back.

We need similar system automation in DynamoDB single-table design.

/3