Map application dependencies
Identify which services sit at headquarters, in a data centre or in the cloud. Note which branch activities must continue if a link or central service becomes unavailable.
Assess the available links
Record provider availability, normal performance, commercial limits and whether alternative links share a physical route or common upstream dependency. Different contracts do not necessarily mean independent paths.
Define the traffic policy
Agree which applications have priority, how routing decisions are made and what should happen during a degraded link. Include voice, video, administrative access and any site-specific systems.
Test the failure conditions
Validate both link failure and service recovery. Document how monitoring will identify a problem, who investigates it and which provider or technical team owns the next step.
This guide is a planning aid. The scope and design for your environment should be agreed after discovery.