Fault domains

OCI has achieved high availability by distributing resources to regions and ADs. A region is a geographically distributed area where one or more ADs are placed. During the initial process of OCI deployment, Oracle created multiple ADs inside a single region, such as Ashburn, Phoenix, Frankfurt, and the UK. These ADs are simply physically separated data centers in a single region.

To further segregate one single AD into more physically isolated areas, OCI created fault domains. A fault domain is a group of rack hardware that has been physically isolated within an AD. Each AD contains three fault domains. You can further choose which fault domain to put your cloud resources into, creating a high-availability structure even when you have just one AD within a region. Fault domains provide anti-affinity rules for your cloud resources. The physical hardware in a fault domain also has its own power supplies, which are redundant, to provide a further layer of availability. You can view a high-level logical diagram of the physically separated fault domain structure within a single AD here:

Figure 1.11 – OCI's fault domain

Fault domains are based on the compute racks within an AD. All of the resources that share a rack will also share a fault domain, and resources in different fault domains cannot exist on the same rack. Customers can choose which fault domain they want to create resources in. This selection, similarly to ADs, is randomly mapped to a physical fault domain per tenancy in order to prevent the uneven usage of fault domains.

In Chapter 4, Compute Choices on Oracle Cloud Infrastructure, we will show you how to choose a fault domain while creating an instance to distribute your workload across physical racks.