Learning about DynamoDB data limitations

Predictable performance at any scale is a huge data point in favor of using DynamoDB. However, there are certain limitations that allow for such predictable performance. Some limitations are soft limits, meaning that they can be increased by requesting AWS, whereas other limits are hard limits that cannot be modified and may need working around using different design patterns and practices. Most of the limitations are well documented in the AWS docs (10) but I will attempt to add more value to this section by including practical information and best practices based on experience.

Throughput

DynamoDB supports two capacity modes for tables – provisioned mode and on-demand mode. While both modes have differences in the ways of metering, throttling, and scaling tables, both deal with units of throughput in similar ways. We will learn more about these capacity modes in Chapter 9, Capacity Modes and Table Classes. However, for this section...