Avoidance 4.6.7 WRED Profiles: DSCP-Based WRED (AF) In DSCP, the following parameters identify Assured Forwarding (AF) PHB based on: Packets requiring AF PHB should be marked with DSCP value aaadd0, where aaa is the number of the class and dd is the drop probability, or drop preference, of the traffic class. There are four defined AF classes. Each class should be treated independently and have bandwidth allocated that is based on the QoS policy. For the AF DiffServ traffic class, WRED configures itself by default for three different profiles, depending on the drop preference DSCP marking bits. Therefore, AF traffic should be classified into the three possible classes, such as AF high drop, AF medium drop, and AF low drop. These three classes are based on the sensitivity to packet drops of the application or applications represented by the class. This would mean that the mission-critical class would have an AF low drop, an AF medium drop, and an AF high drop. Configuring DSCP-Based CBWRED
The random-detect dscp-based command is used to enable DSCP-based WRED on an interface. Changing WRED weighting to values based on DSCP increases the number of WRED traffic profiles to 64 (as compared to 8 profiles for precedence-based WRED). Figure shows the command syntax. You can configure WRED as part of the policy for a standard class or the default class. The WRED random-detect command and the WFQ queue-limit command are mutually exclusive for class policy. If you configure WRED, its packet-drop capability is used to manage the queue when packets exceeding the configured maximum count are enqueued. If you configure the WFQ queue-limit command for class policy, tail drop is used. Changing the WRED Traffic Profile
When DSCP-based WRED is enabled, default values are selected for each traffic profile based on DSCP. Network administrators can then modify these default values to match their specific administrative QoS policy goals. Figure shows the command syntax. When you are modifying the default WRED profile for DSCP, the same values described previously for IP precedence are configurable, except that the minimum threshold, the maximum threshold, and the mark probability denominator will be applied to DSCP classified packets. CBWRED Using DSCP: Example
This example in Figure shows DSCP-based CBWRED. Remember that the DiffServ model itself provides defined traffic classes and their associated PHBs. DiffServ-based classification is used in this example as follows: To enforce this service policy, a router uses CBWFQ to perform bandwidth sharing and uses WRED within service classes to perform differentiated drop. Figure shows sample drop profile graphs and outputs. The configuration example shows how traffic classification is performed using DSCP-based classes, representing the mission-critical class as the AF1 class, and the bulk class as the AF2 class. WRED DSCP-based parameters are set reflecting the class-dependent drop strategy: All other traffic is part of the default class and is fair-queued, with default WRED parameters. Note
When you enable WRED with the random-detect command, the parameters are set to their default values. The weight factor is 9. For all precedences, the mark probability denominator is 10, and maximum threshold is based on the output buffering capacity and the transmission speed for the interface.
Content 4.6 Congestion Avoidance 4.6.8 Monitoring CBWRED The show policy-map interface command displays the configuration of all classes configured for all service policies on the specified interface. This includes all WRED parameters implementing the drop policy on the specified interface. Figure shows the command syntax.The “Key Fields in the show policy-map interface Command Output” table explains some of the key fields of the output of the show policy-map interface command. Key Fields in the show policy-map interface Command Output
Field Description Service-policy output Name of the output service policy applied to the specified interface or virtual circuit (VC). Class-map Class of traffic being displayed. Output is displayed for each configured class in the policy. The choice for implementing class matches (for example, match-all or match-any) can also appear next to the traffic class. Match Match criteria specified for the class of traffic. Choices include criteria such as IP precedence, IP DSCP value, Multiprotocol Label Switching (MPLS) experimental (EXP) value, access groups, and QoS groups. exponential weight Exponent used in the average queue size calculation for a WRED parameter group. mean queue depth Average queue depth based on the actual queue depth on the interface and the exponential weighting constant. It is a fluctuating average. The minimum and maximum thresholds are compared against this value to determine drop decisions.
Content 4.7 Introducing Traffic Policing and Shaping 4.7.1 Traffic Policing and Shaping Overview Both traffic shaping and policing mechanisms are traffic-conditioning mechanisms that are used in a network to control the traffic rate. Both mechanisms use classification so that they can differentiate traffic. They both measure the rate of traffic and compare that rate to the configured traffic-shaping or traffic-policing policy. Figure illustrates these mechanisms. The difference between traffic shaping and policing can be described in terms of their implementation. Traffic policing drops excess traffic to control traffic flow within specified rate limits. Traffic policing does not introduce any delay to traffic that conforms to traffic policies. Traffic policing can cause more TCP retransmissions, because traffic in excess of specified limits is dropped. Traffic-policing mechanisms such as class-based policing or committed access rate (CAR) also have marking capabilities in addition to rate-limiting capabilities. Instead of dropping the excess traffic, traffic policing can mark and then send the excess traffic. This feature allows the excess