An engineer is designing a QoS solution for a campus. The design must guarantee real-time traffic delivery during congestion, minimize the bandwidth consumption for possible virus or worm attacks…

While a strict queue for real-time traffic is good, using a policing policy to drop excessive traffic might be too aggressive, leading to TCP retransmissions and potentially worsening congestion or making applications unstable; a scavenger queue is generally preferred over aggressive dropping for non-critical traffic during congestion.

While queuing can be applied anywhere, the most effective place to manage traffic and prevent congestion from propagating is at the network's edge or where oversubscription occurs, making queuing on access to distribution links a more critical initial placement than distribution to core links.

Creating a strict priority queue for real-time traffic guarantees its delivery during congestion by ensuring it is always serviced first. Implementing a scavenger queue for excessive or low-priority traffic demotes it to the lowest possible priority, ensuring it is only transmitted when no other traffic is present, effectively minimizing its bandwidth consumption and impact on critical services during congestion.

Applying queuing on access to distribution links is crucial because this is often the ingress point for user traffic and where oversubscription can lead to congestion. Implementing QoS here ensures that traffic is classified, marked, and queued as close to the source as possible, allowing for early prioritization and management of congestion before it impacts the network backbone or critical applications.

Creating a shaping policy to drop excessive traffic is not the primary function of shaping; shaping buffers excessive traffic and smooths its transmission to conform to a rate, rather than dropping it directly for congestion management in the same way a scavenger queue would deprioritize it.