Existing mechanisms to provide network and/or compute priority for user traffic

Is there a way today to provide prioritized service (network and/or compute) for a subset of the user traffic in the mesh? Providing low latency service for real-time users while bandwidth intensive background jobs are running is one example scenario where such mechanisms would prove useful.

If such mechanisms do not exist today, how is Istio-generated maintenance/logging/auth traffic (which isn’t latency sensitive) de-prioritized when it competes with user traffic?