N10-009 Question 318
Single answerCongestion/contention: Bottlenecking, Bandwidth, Throughput capacity, Latency, Packet loss, JitterA company has a 200 Mbps internet circuit shared by office users and a cloud-based call center. Every day at 2:00 p.m., users begin large backups to a remote storage service. At the same time, call center agents report choppy audio and occasional gaps during VoIP calls. Monitoring shows the WAN link reaches 95-100% utilization during backups, average latency rises sharply, and jitter increases significantly. Which action is the BEST way to reduce the call quality problem without immediately upgrading the circuit?
- A
Implement QoS to prioritize VoIP traffic over backup traffic on the WAN edge
- B
Increase the Ethernet MTU on user PCs so voice packets are delivered more efficiently
- C
Disable CRC checking on switch ports to reduce delay introduced by frame validation
- D
Lower the TTL value on backup traffic so congested packets are discarded sooner
- E
Replace all Cat 6 patch cables used by call center phones with Cat 8 cables
Show answer and explanation
Correct answer: A
Explanation
The key issue is bottlenecking on the WAN circuit. When the shared 200 Mbps link reaches 95-100% utilization, throughput capacity for all traffic is effectively constrained by the slowest or most heavily used segment. Real-time voice traffic is especially sensitive to congestion because excess queuing introduces latency and jitter, and sustained saturation can also lead to packet loss. The best operational response short of adding bandwidth is to implement QoS so voice traffic receives priority treatment over bulk transfers such as backups. This aligns with common enterprise network design practices and vendor guidance for converged networks: classify time-sensitive traffic, prioritize it in egress queues, and leave lower-priority traffic like backups as best-effort or rate-limited. By contrast, MTU changes, TTL changes, disabling CRC, or replacing cabling do not address the actual symptom pattern shown in the monitoring data.
- A. Correct.
Correct. This scenario indicates WAN congestion and contention caused by backup traffic consuming most available bandwidth. When a link is saturated, delay-sensitive applications such as VoIP are affected first because they are highly sensitive to latency, jitter, and packet loss. Applying QoS at the WAN edge to classify and prioritize voice traffic helps ensure that RTP/voice packets are queued ahead of lower-priority backup traffic. This does not increase total bandwidth, but it reduces the impact of bottlenecking on real-time traffic and is a standard best practice for mixed-use WAN links.
- B. Incorrect.
Incorrect. Increasing MTU on endpoints is not an appropriate fix for VoIP quality issues caused by a saturated WAN circuit. In fact, mismatched MTU settings can introduce fragmentation or black-holing problems. VoIP packets are typically small already, so larger MTU settings do not meaningfully improve voice quality in a congestion scenario. The root issue here is contention for bandwidth, not packet size inefficiency on the LAN.
- C. Incorrect.
Incorrect. CRC checking is a normal Layer 2 error-detection function and should not be disabled. Disabling integrity checks would not solve congestion, jitter, or latency issues, and it would risk allowing corrupted frames to pass unnoticed. The delay from CRC validation is negligible compared with the queuing delay caused by a fully utilized WAN link.
- D. Incorrect.
Incorrect. TTL controls how many Layer 3 hops a packet can traverse before being discarded to prevent routing loops. It is not a congestion-management mechanism and does not selectively improve VoIP quality on a busy link. Lowering TTL for backup traffic would create delivery failures rather than a controlled prioritization strategy.
- E. Incorrect.
Incorrect. Cable replacement is plausible when there are physical-layer errors, interface flaps, or speed/duplex problems, but the scenario specifically identifies near-total WAN utilization with increased latency and jitter during backups. That points to a bandwidth bottleneck, not a cabling deficiency. Cat 8 cabling would not improve a congested 200 Mbps internet circuit.