Hierarchical QoS Shaping to Protect Voice at a 50 Mbps CIR

300-420 By islever Editorial Last updated

Select QoS strategies to meet customer requirements (DiffServ, IntServ) Design end-to-end QoS policies

Short answer

C Apply a hierarchical QoS policy on the router's provider-facing egress interface with a parent shaping policy capped at the 50 Mbps CIR and a child policy prioritizing voice.

Question

An engineer must design a solution to connect a customer to the Internet. The solution will include a Layer 3 circuit with a CIR of 50 Mbps from the service
provider. The hand-off from the provider's switch to the customer's router is 1Gbps. Which solution should the engineer include to prevent potential issues with
choppy voice traffic?
  1. Reduce the bandwidth of the connection to the router.
  2. Implement hierarchical QoS with a parent policing policy.
  3. Implement hierarchical QoS with a parent shaping policy. Correct Answer
  4. Add a bandwidth statement to the router interface.

Key Takeaways

The question tests direction and action of rate control: you shape on egress toward the provider to stay under the CIR, while the trap is choosing a parent policing policy that would drop voice packets instead of queuing them.

This 300-420 design question tests how to keep customer egress traffic under a 50 Mbps CIR on a 1 Gbps hand-off without damaging voice. The answer is C: hierarchical QoS with a parent shaping policy, plus a child policy that prioritizes voice traffic.

Common mistake: Option B, a parent policing policy, is the most tempting wrong choice because policing also enforces the 50 Mbps limit — but it discards excess packets outright, including voice marked priority in the child policy, causing the very choppy audio the design must avoid.

Expert Analysis

Trace the traffic path The customer router's hand-off runs at 1 Gbps, but the provider contract is a 50 Mbps CIR. Any traffic the router sends above 50 Mbps reaches the provider's edge, where the SP policer drops the excess. Because voice is a small, bursty fraction of the mix, those random drops land on voice packets and the listener hears choppy audio. The fix must therefore act on the customer router's egress interface toward the provider, before packets leave the customer network.

Where hierarchical QoS fits A hierarchical policy nests two policy-maps: the parent defines the overall rate for the interface, and the child classifies traffic — voice into a priority queue, best effort into a lower queue. The parent sets the aggregate ceiling; the child decides how that ceiling is shared. This is exactly the structure the question asks for in options B and C, so A and D are eliminated immediately: reducing physical bandwidth is not a QoS mechanism, and a bandwidth statement only influences OSPF cost and LLQ percentage calculations — it does not rate-limit anything.

Why the parent must shape (C) Shaping queues excess packets and releases them as bandwidth becomes available, holding the average rate under the 50 Mbps CIR so the provider's policer never triggers. A child priority queue still sends voice first, and the small, bounded shaping delay is imperceptible. Policing, by contrast, discards anything over the rate. If the parent polices, excess is dropped regardless of the child's priority classification — as one comment puts it, a parent police "will override priority of voice service in the child policy-map." Dropping voice packets is precisely the choppy audio the design must prevent. Option B is therefore wrong; option C — hierarchical QoS with a parent shaping policy at the CIR, plus a voice-priority child policy — is correct.

Official Reference

https://www.cisco.com/c/en/us/support/docs/quality-of-service-qos/qos-policing/19645-policevsshape.html

This link opens the official vendor documentation in a new tab.

Exam Strategy

First identify the direction of the policy: QoS applied on the customer router egresses toward the provider, so shaping is the mechanism that avoids triggering the SP's policer. Then eliminate anything that does not actually rate-limit — a bandwidth statement or a slower physical link. Finally, check whether the parent action preserves the child's voice priority: a parent police discards excess regardless of class, while a parent shape queues it, making the shaped hierarchical policy the safe design.

Frequently Asked Questions

Why shape toward the provider instead of policing the customer's own egress traffic?

Shaping queues excess packets and releases them gradually, keeping the average rate under the 50 Mbps CIR so the provider's policer never drops traffic. Policing on the customer router would discard excess immediately, cutting voice packets and causing choppy audio.

What does the parent policy do in a hierarchical QoS policy-map?

The parent policy-map defines the aggregate rate ceiling for the interface — here, shaping to the 50 Mbps CIR. The nested child policy-map classifies traffic and allocates that ceiling among classes, typically sending voice in a priority queue.

Does a bandwidth statement on the interface prevent exceeding the CIR?

No. The bandwidth command only influences OSPF cost and the percentage calculations used when deriving LLQ queue limits; it does not constrain the actual transmission rate. An explicit shaping or policing policy-map is required to limit egress rate.

More 300-420 FAQ →

Community Votes

C
75%
B
25%

75% of anonymous learners picked answer C. Votes are pick records left by other test-takers — they are not the verified answer.

Discussion 20 contributions

Compiled from public community threads and reader submissions, lightly edited for clarity. Handles and vote tallies are shown as recorded, and comments reflect the views of individual contributors.

From public community threads 20

  1. andrewChan 👍 9 Answered B
    to correctly understand this question, firstly eliminate answer which are definitely no correct, A & D; then look at answer B &C, hierarchical QoS, which is a method to put 2 policy-map nested together, the parent policy-map control the overall channel bandwidth granted, the child policy-map control actual bandwidth allocate to each type of service. in the question, the service provicer grant CIR of 50Mbps which is allocated to parent policy-map and the voice traffic is allocated in child policy-map. As 50Mbps CIR is granted by ISP so (B) policing on parent policy is correct. https://www.cisco.com/c/en/us/td/docs/ios-xml/ios/qos_plcshp/configuration/xe-16/qos-plcshp-xe-16-book/qos-plcshp-hier-clr-plc.html https://www.cisco.com/c/en/us/td/docs/routers/asr9000/software/asr9k_r4-2/qos/configuration/guide/qc42hqos.html
  2. jzzmth 👍 9 Answered C
    The answer is C. Always SHAPE on egress towards a service provider in the parent policy, then use child policies to set priorities as you see fit (in this case VoiP). Generally, shaping = egress (i.e. towards ISP), policing = ingress (this is how ISPs drop traffic that exceed the CIR you are paying them for). Sources: https://www.cisco.com/c/en/us/support/docs/quality-of-service-qos/qos-policing/19645-policevsshape.html#anc17
  3. 12504a3 👍 4 Answered C
    From OCG : - Shaping : commonly used to smooth the traffic going out to the provider (useful to not exceed the contracted rate). Slows down the rate at which packets are sent out an interface (egress). - Policing : Commonly used to give preferential treatment to critical apps (higher class) and reducing best-effort traffic (lower class). Sets the limit of traffic coming into an interface (ingress). At first I wanted to put B as the answer, but the goal here is to not exceed CIR 50 Mbps and this is something we need to deal with because our egress interface is 1 Gbps. So I'm going with C. B is one right answer, but not the best, since we are given the CIR and this is definitely related to Shaping. You are tested on this specific point on this question. Even if, from my point of view, doing both Shaping and Policing should be the best answer.
  4. Clauster 👍 4 Answered C
    Oh men after analyzing this question i just realized something, they want us to set this QoS on our Router, not the ISP router, and then it clicked, Shaping is using on Egress Interfaces and Policing is configured on Ingress Interfaces, in this case Shaping will be setup on our Router, but we can't do policing because we are not managing traffic into our network, we are managing packets out of our router (Egress) for the voice going out. This is Shaping. Makes sense
  5. andrewChan 👍 4
    sorry a bit misunderstand the feature. https://www.cisco.com/c/en/us/support/docs/quality-of-service-qos/qos-policing/19645-policevsshape.html#anc17 in this case the answer should be C, not related to the CIR but if parent policy-map use police, it will override priority of voice service in the child policy-map
Show 15 more comments
  1. 5c725f5 👍 4 Answered C
    It's a sub line rate circuit. You need to shape that or it'll be messy. Shaping is a hard requirement, not policing. You'll attach a child policy with a strict priority queue. That queue technically is policed, but you need a shaper as the parent at the very minimum.
  2. Audie 👍 4
    C is correct. Never use Policing on Voice
  3. cerifyme85 👍 3 Answered C
    https://www.cisco.com/c/en/us/td/docs/ios-xml/ios/qos_plcshp/configuration/xe-16/qos-plcshp-xe-16-book/qos-plcshp-oview.html#:~:text=in%20the%20bucket.-,Traffic%20Policing,the%20traffic%20to%20the%20speed%20of%20the%20interface%20receiving%20the%20packet.,-Was%20this%20Document
  4. Michellangelo 👍 3 Answered C
    Egress = Shaping , ingress Policing is the basic principle. To add, following this discussion, https://community.cisco.com/t5/routing/hierarchical-qos/td-p/456350 , I'm even more convinced the answer is C
  5. salmarin 👍 3 Answered C
    we will use shaping not policing
  6. Sickcnt 👍 2 Answered B
    B is correct. Whats the difference between Policing and Shaping? -If there is an exceeding burst of traffic Policing will cut off the exceeding traffic -If there is an exceeding burst of traffic Shaping will put the traffic into a que and try to send out the packets once the congestion is gone. It is silly to start Shaping (aka.: "Queuing") Voice traffic since it would be all over the place Also for QoS we are usually doing "LLQ" to prioritize Voice traffic, but that would mean voice traffic could totally overtake the whole bandwitdh > Thatswhy we put a "hard policing" at a certain bandwith on policing (not to take over the whole bandwitdth) Configuration snippet for example: https://networklessons.com/quality-of-service/policing-configuration-example
  7. andrewChan 👍 2
    BTW, for Voice traffic is always assign to policing or priority queue in LLQ Traffic Policing is recommended for Voice, Video and Rich media traffic where generally UDP based communication takes place, and delay for packets will cause poor in quality. On the other hand Traffic Shaping is recommended for TCP based applications which can bear delay in traffic but need high data transfer rate like SAP etc. https://ipwithease.com/traffic-policing-vs-shaping/
  8. beskar 👍 2 Answered C
    Pull up any hierarchal QOS example from Cisco on an ethernet interface an you will see it is the shape command for the parent policy.
  9. XalaGyan 👍 1
    Always use police on voice. voice packets are small and contain 20ms of payload. human ears autofill those 20ms and are too slow to check if a packet or two are missing. never shape to no send packets delayed and out of order as the human ear will immediately notice the delay and the out of order of words or tones. use a PQ , Fair Queue . put voice on PQ and prevent it from taking over the world by policing it down to how much BW you need to give for example 5% or fixed Mbps of total. that way everyone can live happily ever after
  10. G_Singh12 👍 1 Answered C
    AS CIR is 50Mbps, implementing policing will drop all the packets over 50Mbps thus imapcting the voice quality. Where as Shaping throttles the router's overall outbound transmission rate to match the provider's contracted limit. This keeps traffic under the carrier's radar so they don't drop voice packets.
  11. 26d13e9 👍 1
    For this specific scenario, the only way to guarantee the ISP will not drop the voice traffic is to make sure not to exceed the CIR from the customer side. This can only be guaranteed by policing out. All comments about shaping is correct but for this particular question.....policing.
  12. LSLS55 👍 1
    I think that C is the correct answer. The engineer can configure Shaping on Egress interface towards ISP. ISP is Policing on Ingress interface, the engineer has no access nor can configure ISP devices. Anyone agrees?
  13. stee_hello 👍 1
    Yep. Create a policy map with just the default class which contains the 50M CIR then a service-policy containing a priority queue for voice is how I'd do it.
  14. Emily23 👍 1
    Should I trust someone on the internet telling that payload is measured in ms ?... It is C
  15. salmarin 👍 1 Answered B
    because it's voice.

Reader comments & corrections

No reader corrections yet — spotted an error or have a note? Share it below.

Log in to comment, report an error, or add a note about this question.

POV Report Builds the Customer Platform During Discovery

S-grade

VPLS design for sites sharing one IP subnet

S-grade

Inter-VN Communication in Cisco SD-WAN Architecture

S-grade

Advertising a Default Route from Core to Aggregation to Hide Instability

S-grade

Practice All 300-420 Questions

Access 150 questions with complete answers and detailed explanations.

View Full 300-420 Practice Test →

← Back to 300-420 Study Guide