concept

Health check is a user defined way for the load balancing instance to detect the health status of the backend ECS. Eliminate the failed nodes and include the recovered nodes to ensure the efficient availability of back-end services.

be careful:

The load balance health check uses some network address segments. You need to enable firewall access to these address segments on the back-end server.

Health check network address of load balancing area: Northeast 10.226.0.0/16 North China 10.45.0.0/16, South China 10.255.0.0/16 Hong Kong 10.255.0.0/16

working process

In load balancing, when an exception occurs to a back-end server, the health check of load balancing listening will detect that the load balancing will block the request forwarded to the failed server and forward it to the normal server. Until the failed server returns to normal, the health check will also detect it, rejoin it to the load cluster, and receive the processing request again.

The health check process is transparent to the client. It can guarantee the high availability of load balancing and avoid a single point of failure of the back-end server. The premise is that there are multiple back-end servers and at least one is providing services normally.