Exactly as described, for a reflector entity this is bandwidth still available for incoming test traffic, while for a generator entity it's the bandwidth remaining that could still be allocated to additional test instances before starting them. On a generator port already running several test instances at configured bandwidths, watching this value shrink toward zero is what tells an operator they're approaching the point where starting one more test instance on that same port would be rejected or would oversubscribe the link. For example, a generator entity on a 1Gbps port already running tests at 700Mbps combined would show roughly 300Mbps available, signaling only a modest additional test instance could still be started there.