You need to ensure that voice traffic sent through an ArubaOS-CX switch arrives with minimal latency What is the best scheduling technology to use for this task?

Strict queuing, also known as Strict Priority (SP) queuing, is the most effective scheduling method for minimizing latency for delay-sensitive traffic like voice. This algorithm services the highest-priority queue exclusively until it is empty before servicing any lower-priority queues. By placing voice traffic in the highest-priority queue, the switch ensures these packets are transmitted with the least possible delay, as they are not required to wait for any other traffic. This makes it the ideal choice for real-time applications where minimizing latency is the primary objective.

B. Rate limiting: This is a traffic policing mechanism that drops packets exceeding a configured rate; it does not prioritize traffic to reduce latency.
C. QoS shaping: This technique buffers and delays excess packets to smooth traffic bursts, which intentionally increases latency to avoid packet loss.
D. DWRR queuing: Deficit Weighted Round Robin (DWRR) allocates bandwidth fairly among queues based on configured weights but does not provide the absolute, immediate priority required for minimal latency voice traffic.

1. ArubaOS-CX 10.12 Quality of Service Guide, Page 21, "Queue scheduling profiles". The document states, "The strict priority scheduler services the highest-priority queue first, as long as that queue is not empty. This scheduler is recommended for delay-sensitive traffic, such as voice." This directly supports the use of strict queuing for voice traffic.

2. ArubaOS-CX 10.12 Quality of Service Guide, Page 21, "Deficit Weighted Round Robin (DWRR)". The guide explains that DWRR services queues based on weights, which provides fair bandwidth allocation rather than the absolute priority needed for minimal latency.

3. ArubaOS-CX 10.12 Quality of Service Guide, Page 25, "Rate limiting and shaping". This section differentiates shaping (which delays traffic) and rate limiting (which drops traffic) from scheduling algorithms, clarifying they serve different purposes than latency minimization through prioritization.