what cpu/memory limits have you configured for the pods?
how much memory is the Beam VM using when timeouts starts to occur?