General availability: Fast Failover and TCP Probing in Azure Traffic Manager
Thursday, June 15, 2017
Two new features in Azure Traffic Manager are now generally available.
Fast Failover: You can choose to have faster redirection of your users away from an endpoint that has become unhealthy. Specifically, you can:
- Choose a shorter interval of 10 seconds for Traffic Manager to check endpoint health status (versus the default 30 seconds).
- Configure the number of tolerated failures (0–9) that should happen consecutively before endpoint is marked as unhealthy.
- Configure the time-out interval for each probe attempt (5–9 seconds if the probe interval is 10 seconds, and 5–10 seconds if probe interval is 30 seconds).
- Set the TTL response all the way down to zero.
Choosing to use the fast interval of 10 seconds for the health checks costs an extra $1 per endpoint per month for checking Azure endpoints, and an extra $2 per endpoint per month for checking external endpoints. Please visit the Azure Traffic Manager pricing page for more details.
TCP Probing: You can choose to have Azure Traffic Manager determine your endpoints’ health by listening to responses for a TCP connection request. You can then add to Traffic Manager profiles those endpoints that do not have the ability to respond to HTTP health checks, such as back-end databases. This feature helps you reduce the resource usage in your endpoints because it typically has a lower overhead. There is no additional charge for using this feature.
You can access the settings for Fast Failover and TCP Probing by using the Azure portal, PowerShell, CLI, and REST APIs/SDKs.
To start using these features, see the endpoint monitoring documentation for Azure Traffic Manager.