Reliability

The reliability pillar includes the ability of a workload to perform its intended function correctly and consistently when it's expected to. This includes the ability to operate and test the workload through its total lifecycle. This paper provides in-depth, best practice guidance for implementing reliable workloads on AWS. The reliability pillar provides an overview of design principles, best practices, and questions.

Definitions

This whitepaper covers reliability in the cloud, describing best practices in the following areas:

• Foundations

• Workload architecture

• Change management

• Failure management

To achieve reliability, you must start with the foundations: an environment where service quotas and network topology accommodate the workload. The workload architecture of the distributed system must be designed to prevent and mitigate failures. The workload must handle changes in demand or requirements, and it must be designed to detect failure and automatically heal itself.

Design principles

While meeting data residency requirements in AWS Regions, Local Zones, and Outposts, there are a number of principles that can help you increase reliability. Keep these in mind as we discuss best practices:

• Automatically recover from failure while maintaining data residency requirements: By monitoring a workload for key performance indicators (KPIs), you can run automation when a threshold is breached. These KPIs should be a measure of business value, not of the technical aspects of the operation of the service. This allows for automatic notification and tracking of failures, as well as automated recovery processes that work around or repair the failure. The failure recovery must be designed to comply with your data residency requirements. With prediction and testing, it is possible to anticipate and remediate most of the failures before they occur.

• Plan capacity requirements: In an AWS Region, you can monitor demand and workload utilization and automate the addition or removal of resources to maintain the optimal level that satisfies demand without over- or under-provisioning. There are still limits, but some quotas can be controlled and others can be managed (for more detail, see Manage service quotas and constraints). With Outposts, capacity is finite and should be planned ahead of time using prediction and testing to forecast capacity correctly ahead of time.