High-availability patterns
High Availability (HA) is the ability of an application to continue running in a healthy state without significant downtime, absorbing temporary failures in dependent services and hardware. Most application strategies for high availability involve either redundancy or the removal of hard dependencies between application components. It is defined using the nines approach (for example, two nines = 99%, three nines = 99.9%, four nines = 99.99%, and so on).
For API platforms, the following strategies are typically adopted to offer high availability:
- Load balancing across instances: APIs will be designed to scale out by adding more instances. Load balancing strategies will improve resiliency by removing unhealthy instances out of rotation or service.
- Geo-redundancy: API platforms will be designed in a way that the Azure resources will be deployed in two geographically distributed locations so that an outage in one of the regions will not impact...