Azure Application Gateway is configured to send traffic to Backend Pools. When those pools contain fewer healthy VMs than expected, the application gateway is considered unhealthy. When this happens, the remaining VMs will handle the full amount of requests, which can quickly overwhelm your application. Blue Matador watches the UnhealthyHostCount metric and creates events when there are unhealthy hosts in your application gateway.
Effects
- Increased load on other VMs backing the application gateway
- Cascading failures as other VMs are unable to keep up with the load to the application gateway
Fixes
To troubleshoot unhealthy hosts try the following:
- Check the health probe endpoint for errors
- Check general host health. Look at application logs, server metrics, or APM metrics to determine what is making your VM unhealthy.
- Use autoscaling with your application gateway to ensure that you always have enough servers to handle load
