This blog entry provides answers and detailed explanations of the QoS questions posed in the previous blog entry.
In our previous blog entry, I challenged you with five QoS review questions. Hope you had fun with those. Here are the answers and explanations:
Question #1:A. WFQ is enabled by default on interface speeds greater than an E1.Answer:Explanation:A. The customer edge (CE) router must implement multiple QoS mechanisms.Answer:Explanation:A. Can starve out low priority traffic in the presence of high priority flowsAnswer:Explanation:
Identify two statements which are correct regarding Weighted Fair Queuing (WFQ). (Choose 2)
B. WFQ has 64 RSVP reservable queues by default.
C. WFQ has a default Hold Queue Out (HQO) value of 1000 packets.
D. WFQ supports a maximum of 256 dynamic queues by default.
E. WFQ has a default Congestion Discard Threshold (CDT) value of 128 packets.
C, D
Weighted Fair Queuing (WFQ) is on by default on interfaces with speeds of an E1 (i.e. 2.048 Mbps) or less. With WFQ, each flow gets its own queue, up to the maximum number of dynamic queues that WFQ is configured for, which is 256, by default. Also by default, the interface’s output queue can hold 1000 packets, which is called the Hold Queue Out (HQO). Even thought the interface’s output queue can hold the HQO number of packets, once the queue depth exceeds the Congestive Discard Threshold (CDT), newly arriving packets from the flow with the worst finish time (i.e. the most aggressive flow) will not be allowed into the queue. By default, the CDT is 64 packets. WFQ can also be configured to work with the Resource Reservation Protocol (RSVP), where RSVP traffic is placed into reservable queues, which are given extremely high priority treatment. However, by default, zero reservable queues are configured for WFQ.
Question #2:
Some service providers overprovision their bandwidth to avoid the configuration of complex QoS mechanisms. Identify a drawback of this design philosophy.
B. The provider edge (PE) router must implement multiple QoS mechanisms.
C. The unpredictability of traffic resulting from the lack QoS mechanisms in the service provider’s core results in a higher probability of network congestion.
D. The service provider’s recurring expenses are higher for an overprovisioned network as opposed to a QoS-enabled network.
D
While some service providers prefer to purchase excess bandwidth for their backbone network, to avoid the necessity of configuring QoS mechanisms, this approach does have its drawbacks. For example, the service provider is burdened with recurring monthly charges for bandwidth that goes largely unused. Also, in the absence of QoS mechanisms, anomalous traffic patterns (e.g. a broadcast storm or a denial of service (DoS) attack) could impact high-priority.
Question #3:
Cisco IOS supports a wide variety of queuing mechanisms. From the following list of queuing characteristics, select three characteristics of Weighted Fair Queuing (WFQ). (Choose 3)
B. Is enabled by default on slow speed serial interfaces (i.e. speeds of 2.048 Mbps and slower)
C. Is configured using the three-step MQC configuration process
D. Lacks the ability to provide a minimum bandwidth guarantee
E. Uses a packet’s size and a packet’s IP Precedence value when calculating the weight of a packet.
B, D, E
Weighted Fair Queuing (WFQ) is considered to be a legacy queuing mechanism, and is widely supported in multiple IOS versions. As the name suggests, WFQ attempts to treat traffic fairly, by not allowing one type of traffic (e.g. an aggressive flow, such as an FTP flow) from starving out another type of traffic (e.g. a flow not requiring much bandwidth, such as a Telnet flow). The way WFQ accomplishes this is to place each flow into its own queue (with a maximum of 256 queues by default), and then empty traffic from those queues based on the packet arrival time, the size of the packet, and the IP Precedence value assigned to the packet. Due to this automatic queue separation, no complex traffic classification configuration is necessary. Also, WFQ is the default queuing method on link speeds running at an E1 speed (i.e. 2.048 Mbps) and lower, while FIFO (First-In First-Out) queuing is used by default on link speeds running at speeds greater than 2.048 Mbps. However, unlike CB-WFQ, since WFQ is trying to treat all flows fairly by making bandwidth available to all flows, it does not support a fixed bandwidth guarantee.
Question #4:A. 4000 bpsAnswer:Explanation:peak_rate = CIR * (1 + Be/Bc)In this example: peak_rate = 8000 * (1 + 2000/2000) = 16,000 bps. Note that if the Bc and Be values are calculated by IOS rather than being statically configured, Bc will always equal Be, which means that the peak rate will be twice the CIR.Question #5:A. WRED is a congestion management mechanism.Answer:B, CExplanation:For more free QoS practice questions, check out the QoS page of my website.
You configure CB-Shaping by issuing the command shape peak 8000 2000 2000. This configuration shapes to what peak rate?
B. 8000 bps
C. 16000 bps
D. 32000 bps
C
In the syntax, the 8000 represents the Committed Information Rate (CIR). The first 2000 is the Committed Burst (Bc), and the second 2000 is the Excess Burst (Be). When configuring CB-Shaping, you can either shape to “average” or shape to “peak.” When shaping to average, traffic rates don’t exceed the CIR. However, when shaping to peak, traffic rates can burst above the CIR, while some of that excess traffic could be dropped by the service provider. When shaping to peak, the peak shaping rate is calculated by the formula:
Identify two of the following statements which are correct regarding WRED. (Choose 2)
B. WRED uses an interface’s average queue depth when determining if a packet should be dropped.
C. WRED uses a packet’s IP Precedence or DSCP marking when determining if a packet should be dropped.
D. WRED’s Mark Probability Denominator (MPD) parameter equals the probably of discard when the queue depth equals the RED profile’s maximum threshold value.
Weighted Random Early Detection (WRED) is a congestion avoidance mechanism. Different IP Precedence and DSCP values each have RED profiles, which determine the probability of packet discard based on an interface’s average queue depth. A RED profile specifies the minimum threshold, maximum threshold, and probability of discard when the average queue depth equals the maximum threshold. When the average queue depth is at or below the minimum threshold, there is zero probability of discard (i.e. the “no drop” mode). When the average queue depth is greater than the maximum threshold, there is a 100 percent probability of discard (i.e. the “full drop” mode). When the average queue depth is greater than the minimum threshold and equal to or less than the maximum threshold, there is a possibility of packet discard (i.e. the “random drop” mode). The probably of discard when the average queue depth equals the maximum threshold equals 1/MPD, where MPD is the configurable Mark Probability Denominator. For example, if the MPD equals 4, the probably of discard when the average queue depth equals the maximum threshold is 1/4 (i.e. 25 percent).
Hope you’ve enjoyed the collection of QoS videos and review questions this month. Stay tuned, because I’ll announce November’s exam focus very soon!
