We've deployed an Azure Application Gateway to load balance requests between two servers. The service being served uses SAML for authentication, and we use cookie-based affinity.

How/where can I monitor the distribution of load between the servers in the backend pool? The default logs and metrics only include things like "Total number of requests", or individual servers' uptime.

CodePudding user response：

How/where can I monitor the distribution of load between the servers in the backend pool? The default logs and metrics only include things like "Total number of requests", or individual servers' uptime.

Unfortunately, we don't have any built-in signals to capture the load distribution between the backend instances.

You need to create a custom solution by enabling the diagnostics settings on the gateway and by writing custom KQL queries. you will be able to get the number of requests that were handled by each backend instances.

Here is the sample query:

AzureDiagnostics| where OperationName == "ApplicationGatewayAccess"
| where httpStatus_d == "200" and serverRouted_s == "<GatewayBackendVMIP"