300-320 · Question #404
A network design team is experiencing sustained congestion on access and distribution uplinks. QoS has already been implemented and optimized, and it is no longer effective in ensuring optimal…
The correct answer is C. Implement higher-speed uplink interfaces. D. Bundle additional uplinks into logical EtherChannels. When QoS is already fully optimized and sustained congestion persists, the root cause is insufficient bandwidth, so the only lasting remedies are adding physical capacity through faster interfaces or link aggregation.
Question
A network design team is experiencing sustained congestion on access and distribution uplinks. QoS has already been implemented and optimized, and it is no longer effective in ensuring optimal network performance. Which two actions can improve network performance? (Choose two.)
Options
- AReconfigure QoS based on the IntServ model.
- BConfigure selective packet discard to drop noncritical network traffic.
- CImplement higher-speed uplink interfaces.
- DBundle additional uplinks into logical EtherChannels.
- EUtilize random early detection to manage queues.
How the community answered
(40 responses)- A8% (3)
- B10% (4)
- C80% (32)
- E3% (1)
Why each option
When QoS is already fully optimized and sustained congestion persists, the root cause is insufficient bandwidth, so the only lasting remedies are adding physical capacity through faster interfaces or link aggregation.
Switching from DiffServ to IntServ changes the QoS signaling model but does not add bandwidth; if current QoS is already ineffective, a different model will not resolve physical saturation.
Selective packet discard is a form of congestion management already covered under QoS; since QoS is stated as no longer effective, additional discard mechanisms will not resolve underlying bandwidth exhaustion.
Upgrading to higher-speed uplink interfaces directly increases the available bandwidth on the congested segment, addressing the root cause rather than just managing how limited bandwidth is allocated.
Bundling additional uplinks into EtherChannels (LAG/LACP) logically aggregates multiple physical links, multiplying available throughput between access and distribution layers without replacing existing hardware.
Random Early Detection (RED) helps prevent queue tail-drop and manage average queue depth, but it is a queue management tool that does not add bandwidth and is ineffective once links are fully saturated.
Concept tested: Bandwidth augmentation when QoS is insufficient
Source: https://www.cisco.com/c/en/us/td/docs/solutions/Enterprise/Campus/caDesign.html
Topics
Community Discussion
No community discussion yet for this question.