I am using container-native load balancing (NEG on GKE) and I would like to ask, if is there is any option that same single request to load balancer can be processed twice (or more times) - by giving to more pods.

I have this suspicion, that is happening only because of NEG (there is no way to send more requests from my App) and according to logged time I found 3x same timestamp (seconds) of this LB request (this log is collected from all pods of course).

k8s final configs: https://gist.github.com/Barbarossa-42/86aac72da68c9e95a726e0845334d68a

UPDATE: its happening when LB timeout is reached or when pod crashed. For me there is a strange behaviour. Why its sending this request again to another (working) pod?

Thanks for any help