I found the solution. The load balancer was configured with a default timeout of 30 seconds.
Based on further testing, while the number of responses the client received before receiving the RST_STREAM frame still varied, I saw that this consistently occurred after approximately 30 seconds. Searching explicitly for timeouts, I found a StackOverflow question (see https://stackoverflow.com/questions/44601191/kubernetes-on-gce-ingress-timeout-configuration) pointing me into the right direction. Google's documentation does not talk about that default timeout. At least I did not find it. -- You received this message because you are subscribed to the Google Groups "grpc.io" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/grpc-io/e4bf2b4b-9990-4842-bc8e-f3531aa12800%40googlegroups.com.
