I've been having network issues with an ALB ingress in EKS, with kubectl plugin managed to Wireshark into the pod and got TCP window full, followed by TCP window Zero packet, the thing is this is from the alb to the pod and not the client to the alb.
is it possible that the alb sends a TCP window full to the pod and not the client? ive been reading around and couldn't find information that says its possible or not, also if I wanted to double-check a tcpdump should be run from the client network side, is there a way to corroborate is their network issue and not ours from our side?