Microsoft Azure Architect Design (AZ-301) Practice Exam

Which consistency level should be configured to ensure that any reads from a Contoso DB account never see writes, while minimizing costs?

1. Strong consistency level

2. Eventual consistency level

3. Bounded staleness

4. Consistent prefix

The correct answer is: Strong consistency level

The correct choice is the strong consistency level. Configuring strong consistency ensures that once a write is acknowledged, any subsequent read will reflect that write or any writes before it. This means that reads will never see any data that is stale or outdated, effectively guaranteeing that users interacting with the Contoso DB will always receive the most recent and accurate information after a write operation. While strong consistency offers high accuracy, it is important to note that it can incur higher costs due to increased latency and the overhead associated with maintaining this strict consistency across distributed systems. However, since the primary requirement is to ensure that reads do not reflect uncommitted writes, the strong consistency option fits this need perfectly. Other consistency levels like eventual consistency may allow reads to see outdated data, ultimately not meeting the requirement that reads should not see writes. Bounded staleness provides a way to control how stale data can be, which could still result in reading older data before a certain threshold. Similarly, consistent prefix also allows some level of staleness, which does not satisfy the requirement of ensuring reads never see writes. Therefore, strong consistency is the ideal choice for the scenario presented.