request help: Question - When I change the Upstream from Kubernetes environment to Admin API

[kworkbee]

Issue description

When an Upstream is changed by Admin API in Kubernetes environment, there is a phenomenon in which it is restored to its original state (Declared in ApisixUpstream spec) after a specific time.

I wonder if this is the intended action, and if so, please answer whether it is reasonable to change only the ApisixUpstream spec rather than changing the Upstream information through Admin API in Kubernetes environment.

Environment

• your apisix-ingress-controller version (output of apisix-ingress-controller version --long): 0.14.0
• your Kubernetes cluster version (output of kubectl version): v1.25
• if you run apisix-ingress-controller in Bare-metal environment, also show your OS version (uname -a): AWS EKS