GKE unhealthy LB targets

Posting a question as I’m wondering if anybody else has had a similar experience.

*Description*:

If my gateway Deployment object, which manages my gateway, moves to a different node, the L4 GKE load balancer flags all instances as unhealthy. In this case, the gateway might still function, but the health check fails to detect it properly.

*Steps*:

Deploy a gateway and an HTTPRoute, then verify that the load balancer shows a healthy status (for one node).
Schedule the Deployment managed by the gateway onto another node.
Observe that the load balancer no longer has any healthy targets.

*Resolution*:
Delete the gateway resource and recreate it. This forces GCP to recreate the load balancer, resolving the issue.

[optional *Relevant Links*:]
https://github.com/envoyproxy/gateway/issues/5195