What is true about a scan operation in DynamoDB?

A scan operation in DynamoDB does indeed read every item in a table or secondary index. This means that when performing a scan, the operation examines every single item in the specified table or index, retrieving the data irrespective of the item's keys or any sort of filter. It does not require any specific query criteria and can potentially return a large amount of data, depending on the size of the table.

This characteristic of a scan operation is crucial for understanding its impact on performance and cost. Since it processes every item, it tends to be less efficient than a query, which retrieves items based on specific key conditions and can narrow down the search significantly. Additionally, scans are limited to one table or one secondary index at a time, meaning they cannot pull data from multiple tables simultaneously. While scans can utilize filters to reduce the amount of data returned, they do not operate based on keys like queries do; rather, they can return items regardless of their keys unless additional conditions are applied.