Partitions

DynamoDB removes the need for companies to manage infrastructure. Nonetheless, this infrastructure and DynamoDB’s own architecture are quite complex. This architecture supports all of our requirements related to availability and scalability. It’s important to know how this architecture functions, because it can come in handy when designing systems. In this lesson, we’re going to discuss an integral part of DynamoDB’s architecture—partitions.

Important concepts

Before understanding how partitions work in DynamoDB, it’s worth clarifying some basic concepts about AWS itself:

• Region: This is the physical location where AWS has its data centers. Regions are distributed worldwide.
• Availability zone: These are physical data centers whose infrastructure is independent of each other. They’re important because they give our applications fault tolerance and resilience.
• SSD: This stands for solid state drive. It’s a type of storage device where I/O operations are faster than conventional hard disks.

What are partitions?

When information is stored in DynamoDB, it is distributed across different availability zones. Each availability zone contains thousands of SSDs. What this means is that data is split into different subparts. These subparts are called partitions and each partition is stored in an SSD.