A database administrator experiences delay when performing storage replication between primary and secondary data centers. The data centers are…

The scenario involves storage replication over a long-distance (20 km) FCIP WAN link, where performance is impacted by delay. To ensure the replication traffic receives the highest priority and lowest possible latency, strict priority queuing (SPQ) is the most suitable QoS model. SPQ services the highest-priority queue exclusively until it is empty before servicing any lower-priority queues. This guarantees that the critical, delay-sensitive replication data is transmitted with minimal delay, effectively addressing the performance issue over the constrained WAN link. Other queuing methods provide fair access but do not offer the absolute prioritization needed here.

A. Resource Reservation Protocol (RSVP) is a signaling protocol for reserving bandwidth (IntServ), which is overly complex and not typically used for simple prioritization on an FCIP link.
B. Weighted round robin (WRR) provides proportional bandwidth to queues but does not guarantee the lowest latency for critical traffic, as it continues to service all queues.
C. Differentiated services (DiffServ) is a framework for classifying traffic, but SPQ is the specific queuing mechanism within that framework that best solves the latency problem.

1. Cisco MDS 9000 Series Quality of Service Configuration Guide, Release 8.x, "Information About QoS Queuing," Page 1-10. "With strict priority queuing, the scheduler services the queues in order of their priority... This type of queuing is effective for mission-critical data applications."

2. Cisco MDS 9000 Series NX-OS FCIP Configuration Guide, Release 8.x, "Configuring QoS for FCIP," Page 3-10. The guide explains how to apply QoS policies to FCIP tunnels, where strict priority can be used to prioritize specific traffic flows like replication.