642-785 - Maintaining Cisco Service Provider Quality of Service (MSPQS)
Go back to Cisco
Which is the preferred queuing method provided by Cisco IOS Software for a converged VoIP network?
When the customer LAN is managed by the service provider and the network is oversubscribed, what is the best location for use of congestion avoidance mechanisms?
on output interfaces from the customer LAN leading into the service provider network
Within a QoS policy-map configuration, one of the traffic classes is configured with the random detect dscp-based configuration command. What will that command accomplish?
decreases the probability of congestion by selectively dropping TCP packets before the queue is full
When the RFC 3270 Uniform mode is used, the PHB at the egress router is based on which of these?
The topmost MPLS label EXP value.
What are three benefits of IntServ and RSVP? (Choose three.)
RSVP helps network devices identify dynamic port numbers.
IntServ networks will reject or downgrade new RSVP sessions if all reservable bandwidth is booked somewhere in a path.
They enable the network to guarantee necessary QoS to individual data flows.
Which tool is best suited to provide incident management or trouble ticketing?
Where do service providers apply policy when traffic is leaving the customer network? (Choose two.)
Apply input policy on the PE interface toward CE, if CE is unmanaged.
Apply output policy on the CE interface toward PE, if CE is managed.
The provider you work for supports a wide range of Cisco router models and Cisco IOS Software releases. You are in the process of moving to a SNMP-based tool to monitor QoS configurations and statistics. What are some of the challenges to this approach?
Older versions of Cisco IOS Software do not maintain index persistency, which results in significant changes in the information that is collected unless properly configured.
Which Cisco IOS command will show a summary of probes?
show ip sla monitors
Which Cisco IOS command lists all traffic classifications?
An MPLS VPN customer reports that the VoIP voice quality is very bad for calls across the MPLS VPN. What could be the problem?
Traffic shaping is applied to the voice traffic class
Which Cisco IOS IP SLA Probe would you select to identify bottlenecks in the network?
ICMP Path Echo
What are two benefits of the DiffServ model? (Choose two.)
Service quality can be tailored to the needs of many different types of traffic.
It does not require setting up an end-to-end signaling mechanism.
What are three fundamental QoS mechanisms? (Choose three.)
Which two match statements are correct in Cisco IOS Software? (Choose two.)
class-map class1 match not protocol ip
class-map match-any class3 match dscp af31 match dscp af41
A customer reports drops for traffic in the BIZ-CRITICAL class. As part of the verification process you identify that there is a high number of random drops. Which of these can help address this problem?
Increase the WRED min threshold value for the traffic class.
A customer reports that the packets from their custom TCP application, which is currently marked as DSCP AF43, is being dropped more frequently than desired. All of the links in service provider network are using DSCP-based WRED. What can be changed to best improve the drop rate of the customer's custom application without causing major problems for the other traffic classes?
Mark the custom application traffic to DSCP AF41 instead.
Which MQC capabilities are unique to Cisco IOS XR Software?
When the RFC 3270 Short Pipe mode is used, the PHB at the egress router is based on which of these?
The Differentiated Services Field value in the IP packet header
You work for service provider that employs a four-class service provider QoS model (best effort, bronze, silver, gold). A customer recently reported an issue with high drop rate on a missioncritical database application in the silver class. The silver class is currently at capacity. The gold class is currently only used for voice and video conferencing, and has available bandwidth. The application was remapped to the gold class. Which statements are valid alternative options to this decision? (Choose two.)
The provider should consider moving to a six-class QoS model to provide more available classes of service to differentiate traffic.
Additional bandwidth should have been added to the silver class from one of the lower classes before the traffic class was moved.
Which three can be monitored by IP SLA? (Choose three.)
A customer with an unmanaged CE router has reported experiencing poor quality on the video conferencing and broadcast multimedia applications. To address the issue, the provider has implemented shaping on the PE router to "smooth out" speed mismatches in the network. Which of these is a valid concern with this approach?
While shaping may help the broadcast multimedia application, it could cause additional delay in the video conferencing application.
Which statement about the MQC configuration shown here for the class-default traffic class is true? policy-map CE-EDGE-4CLASS-OC3-POS-1 class class-default bandwidth percent 25 !
The maximum number of packets the queue for the class-default class can hold is 64 packets using the default queue-limit.
When troubleshooting a QoS problem, which two key statistics should you look for? (Choose two.)
When is a best-effort QoS policy most appropriate?
when the customer will only transport delay-tolerant data
What are the benefits of using differentiated services in the service provider backbone? (Choose two.)
less bandwidth required to achieve the same SLA
more aggregate traffic supported
What three statements about WRED are true? (Choose three.)
WRED is a congestion avoidance technique.
Using WRED on the edge devices (CE, PE) is an effective method of implementing congestion avoidance in the service provider network, because it avoids the congestion prior to reaching the core.
WRED employs a weighted method for determining which packets to drop to avoid congestion. Packets with a higher precedence are less likely to be dropped
Which two methods can be used to monitor QoS performance? (Choose two.)
A service provider recently changed from RFC 3270 Uniform mode to Pipe mode. Which three considerations about this change are true? (Choose three.)
If the provider moved to a Short Pipe mode model, it would allow them to implement customerspecific QoS policy on the PE egress.
The Pipe mode provides the isolation between the customer and provider DiffServ domains, while maintaining a simpler support model.
In Uniform mode, when the topmost EXP value is changed, the changed EXP value will propagate to the original IP precedence or DSCP.
Which statement about the rates and statistics (such as match, transmit, drop, and police) shown in the show policy-map interface output is true?
The rates are computed as a moving average of instantaneous rates.
A customer reports that during peak hours, FTP performance is unacceptable. Currently, the service provider maps the customer FTP traffic into the class-default traffic class and FTP traffic is marked with MPLS EXP1. The class-default class traffic is configured with MPLS EXP based WRED and also with the bandwidth command. What could be causing the FTP performance issues with the customer?
FTP packets are being more randomly dropped that the other TCP traffic within the classdefault traffic class, because the WRED thresholds for MPLS EXP 3 marked packets were configured to improper values.
What problem could occur when applying this policy map on CE? policy-map CE-SP-MODEL class VOIP priority class BROADCAST-VIDEO priority class SP-CRITICAL-DATA bandwidth percent 40 random-detect exp 3 15000 packets 45000 packets class class-default random-detect exp 1 2500 packets 7500 packets random-detect exp 0 1500 packets 10500 packets
No bandwidth or policing is configured for voice and video traffic, and other traffic could be impacted.
A MPLS VPN customer reports poor VoIP call quality across the MPLS VPN. What could be the problem?
The total end-to-end delay measured is constantly over 350 msec.
Where would a service provider apply this policy?
PE interface toward CE
Which three components would you use to define and apply traffic classification? (Choose three.)
To classify and prioritize traffic according to importance, which three statements are correct? (Choose three.)
Voice traffic should be given top priority.
Email should be placed in a best-effort class.
Critical applications should be given sufficient bandwidth to meet their requirements.
What would be the first step when defining an MQC?
Define a traffic class by using the class-map command.
Which of these correctly describes traffic classification using qos-group?
qos-group is local to the router.
Which statement about when you mix TCP and UDP traffic in the same class and overflow occurs is true?
TCP traffic is starved, because UDP never slows down when packets get dropped.
Where would you apply WRED to avoid congestion and poor QoS from oversubscription of the network?
wherever there is a potential bottleneck and most traffic is TCP
An MPLS VPN customer with managed CE services recently made some changes to their call manager configurations. Now the customer reports that their VoIP traffic voice quality is very poor. What could be a cause of the problem?
The customer changed the codec used for voice compression from G.729 to G.711.
Which two items can be obtained from the CISCO-NETFLOW-MIB? (Choose two.)
A customer is experiencing poor quality on voice connections coming from a particular branch office. Data on the service provider network capacity indicates that the devices serving that location are heavily oversubscribed. It will be months before planned expansions can be implemented. Which class-based traffic-shaping technique would be appropriate to consider implementing?
shaping to the configured average rate
A service provider will apply this policy at which of these interfaces? interface Gi1/1 service-policy output SP-POLICY ! policy-map SP-POLICY class TRAFFIC-CLASS-1 police rate percent 20 conform-action transmit exceed-action drop priority ! class TRAFFIC-CLASS-2 bandwidth percent 50 queue-limit 800 ms random-detect exp 3 280 ms 490 ms ! class class-default queue-limit 500 ms random-detect exp 1 200 ms 350 ms random-detect exp 0 100 ms 200 ms ! end-policy-map ! class-map match-any TRAFFIC-CLASS-1 match dscp ipv4 45 ef cs6 match mpls experimental topmost 5 6 end-class-map ! class-map match-any TRAFFIC-CLASS-2 match dscp ipv4 cs2 af21 af23 af31 af32 af33 35 37 match mpls experimental topmost 3 end-class-map !
PE interface toward P
What can be used to provide read-only access to QoS configuration information and statistics from a network management station?
Which two statements about IP SLA are correct? (Choose two.)
Cisco IOS IP SLA is Layer 2 transport independent.
IOS IP SLA uses active traffic monitoring which allows users to analyze service levels for IP applications.
Which Cisco IOS feature is used to enable the following class map configuration? class-map match-all test-cm match protocol gnutella file-transfer "*"
A provider plans to use IP SLA to help diagnose future performance issues in their PE routers in the network. Which two should be considered by the provider? (Choose two.)
IP SLA should be implemented on dedicated hardware to properly simulate network endpoints
IP SLA should use a mix of source and target points.
Which three issues are important challenges for monitoring service quality in the SP core network? (Choose three.)
Multiple flows compete for a limited amount of bandwidth
Packets have to traverse many network devices and links that add up to the overall delay.
Packets may have to be dropped when a link is congested
Which of these describes Short Pipe mode in QoS for MPLS VPN service providers?
Service provider does not remark customer DSCP values. (Service provider uses independent MPLS EXP markings.) Final PE-to-CE policies are based on customer markings.