Usage Guidelines

The template should be enabled to create port-channel active/active mode. The device restarts after enabling the sdm prefer enable_portchannel_qos_multiple_active command. After a successful reboot, verify the configuration using the command show sdm prefer current .

Usage Guidelines

Usage Guidelines

Enables this template to achieve 3072 QoS TCAM entries. Use the sdm prefer enable_qos_scale command.

Usage Guidelines

Use the sendqdm command to send a message to a specific QDM client. For example, entering thesendqdmclient9messagehello command will send the message “hello” to client ID 9.

Use the sendqdmmessage message-text command to send a message to all QDM clients. For example, entering the sendqdmmessagehello command sends the message “hello” to all open QDM clients.

Usage Guidelines

The maximum number of service groups that are supported on a router is determined by the router at system-startup time. For the Cisco 7600 series router, the valid entry range for the service-group-identifier argument is 1 to 32768.

Usage Guidelines

The table below shows which configuration mode to choose based on the intended use of the command.

You can attach a single policy map to one or more interfaces or to one or more VCs to specify the service policy for those interfaces or VCs.

A service policy specifies class-based weighted fair queueing (CBWFQ). The class policies that make up the policy map are then applied to packets that satisfy the class map match criteria for the class.

Before you can attach a policy map to an interface or ATM VC, the aggregate of the configured minimum bandwidths of the classes that make up the policy map must be less than or equal to 75 percent (99 percent on the Cisco 10008 router) of the interface bandwidth or the bandwidth allocated to the VC.

Before you can enable low latency queueing (LLQ) for Frame Relay (priority queueing [PQ]/CBWFQ), you must first enable Frame Relay traffic shaping (FRTS) on the interface using the frame-relay traffic-shaping command in interface configuration mode. You then attach an output service policy to the Frame Relay VC using the service-policy command in Static maps class configuration mode.

To attach a policy map to an interface or ATM VC, the aggregate of the configured minimum bandwidths of the classes that make up the policy map must be less than or equal to 75 percent of the interface bandwidth or the bandwidth allocated to the VC. For a Frame Relay VC, the total amount of bandwidth allocated must not exceed the minimum committed information rate (CIR) configured for the VC less any bandwidth reserved by the frame-relay voice bandwidth or frame-relay ip rtp priority Static maps class configuration mode commands. If these values are not configured, the minimum CIR defaults to half of the CIR.

Configuring CBWFQ on a physical interface is possible only if the interface is in the default queueing mode. Serial interfaces at E1 (2.048 Mbps) and below use weighted fair queueing (WFQ) by default. Other interfaces use first-in first-out (FIFO) by default. Enabling CBWFQ on a physical interface overrides the default interface queueing method. Enabling CBWFQ on an ATM permanent virtual circuit (PVC) does not override the default queueing method.

When you attach a service policy with CBWFQ enabled to an interface, commands related to fancy queueing such as those pertaining to fair queueing, custom queueing, priority queueing, and Weighted Random Early Detection (WRED) are available using the modular quality of service CLI (MQC). However, you cannot configure these features directly on the interface until you remove the policy map from the interface.

Note

Beginning in Cisco IOS Release 12.4(18e), you cannot configure the traffic-shape rate and MQC shaping on the same interface at the same time. You must remove the traffic-shape rate configured on the interface before you attach the service policy. For example, if you try to enter the service-policy {input | output } policy-map-name command when the traffic-shape rate command is already in effect, this message is displayed: Remove traffic-shape rate configured on the interface before attaching the service-policy. If the MQC shaper is attached first, and you enter the legacy traffic-shape rate command on the same interface, the command is rejected and an error message is displayed.

You can modify a policy map attached to an interface or VC, changing the bandwidth of any of the classes that make up the map. Bandwidth changes that you make to an attached policy map are effective only if the aggregate of the bandwidth amount for all classes that make up the policy map, including the modified class bandwidth, is less than or equal to 75 percent of the interface bandwidth or the VC bandwidth. If the new aggregate bandwidth amount exceeds 75 percent of the interface bandwidth or VC bandwidth, the policy map is not modified.

After you apply the service-policy command to set a class of service (CoS) bit to an Ethernet interface, the policy remains active as long as there is a subinterface that is performing 8021.Q or Inter-Switch Link (ISL) trunking. Upon reload, however, the service policy is removed from the configuration with the following error message:

Process "set" action associated with class-map voip failed: Set cos supported only with IEEE 802.1Q/ISL interfaces.

Note	The service-policy input and service-policy output commands cannot be configured if the ip subscriber interface command is already configured on the interface; these commands are mutually exclusive.

Usage Guidelines

You can attach a single policy map to one or more classes to specify the service policy for those classes. This command is only available for the output interface, which is assumed.

Usage Guidelines

After entering the control-plane command, use the service-policy command to configure a quality of service (QoS) policy. This policy is attached to the control plane interface for aggregate or distributed control plane services and controls the number or rate of packets that are going to the process level.

When you configure output policing on control-plane traffic, using the service-policyoutput policy-map-name command, a router is automatically enabled to silently discard packets. Output policing is supported as follows:

• Supported only in:

  • Cisco IOS Release 12.2(25)S and later Cisco IOS 12.2S releases.
  • Cisco IOS Release 12.3(4)T and later Cisco IOS 12.3T releases.
  • Cisco IOS Release 12.2(18)SXD1 and later Cisco IOS 12.2SX releases.
  • Cisco IOS XE Release 2.2 and later Cisco IOS XE releases.

• Not supported for attaching a QoS policy for distributed control-plane services.

• Not supported on the Cisco 6500 router, Cisco 7500 series, and Cisco 10720 Internet router.

The service-policyoutput command configures output policing, which is performed in silent mode to silently discard packets exiting from the control plane according to the attached QoS policy. Silent mode allows a router that is running Cisco IOS software to operate without sending any system messages. If a packet that is exiting from the control plane is discarded for output policing, you do not receive an error message.

Silent mode allows a router that is running Cisco IOS software to operate without sending any system messages. If a packet that is destined for the router is discarded for any reason, users will not receive an error message. Some events that will not generate error messages are as follows:

• Traffic that is being transmitted to a port to which the router is not listening

• A connection to a legitimate address and port that is rejected because of a malformed request

Usage Guidelines

This command is used to create hierarchical service policies in policy-map class configuration mode.

This command is different from the service-policy { input |output } policy-map-name command used in interface configuration mode. The purpose of the service-policy {input |output } policy-map-name is to attach service policies to interfaces.

The child policy is the previously defined service policy that is being associated with the new service policy through the use of the service-policy command. The new service policy using the preexisting service policy is the parent policy.

This command has the following restrictions:

• The priority command can be used in either the parent or the child policy, but not both policies simultaneously.

• The shape command can be used in either the parent or the child policy, but not both polices simultaneously on a subinterface.

• The fair-queue command cannot be defined in the parent policy.

• If the bandwidth command is used in the child policy, the bandwidth command must also be used in the parent policy. The one exception is for policies using the default class.

Usage Guidelines

The policy map must already exist and must contain the Quality of Service (QoS) feature to be applied to the service group, according to the provisions specified by the Service Level Agreement (SLA). To create and configure the policy map, use the Modular Quality of Service Command-Line Interface (CLI) (MQC). For more information about the MQC, see the Cisco IOS Quality of Service Solutions Configuration Guide .

Usage Guidelines

The input and output keywords indicate the direction in which the policy map will be applied.

The value for the policy-map-name argument represents a QoS policy map configured on the switch using the policy-map policy-map-name global configuration command.

Usage Guidelines

To disable this command, remove the service policy from the interface using the noservice-policy command.

The setatm-clp command works only on platforms that support one of the following adapters: the Enhanced ATM Port Adapter (PA-A3), the ATM Inverse Multiplexer over ATM Port Adapter with eight T1 ports (PA-A3-8T1IMA), or the ATM Inverse Multiplexer over ATM Port Adapter with eight E1 ports (PA-A3-8E1IMA). For more information, refer to the documentation for your specific router.

A policy map containing the setatm-clp command can be attached as an output policy only. The setatm-clp command does not support packets that originate from the router. A policy map containing ATM set CLP bit quality of service (QoS) cannot be attached to PPP over X (PPPoX) sessions. The policy map is accepted only if you do not specify the setatm-clp command.

Usage Guidelines

CoS packet marking is supported only in the Cisco Express Forwarding switching path.

The setcos command should be used by a router if a user wants to mark a packet that is being sent to a switch. Switches can leverage Layer 2 header information, including a CoS value marking.

The setcos command can be used only in service policies that are attached in the output direction of an interface. Packets entering an interface cannot be set with a CoS value.

The matchcos and setcos commands can be used together to allow routers and switches to interoperate and provide quality of service (QoS) based on the CoS markings.

Layer 2 to Layer 3 mapping can be configured by matching on the CoS value because switches already can match and set CoS values. If a packet that needs to be marked to differentiate user-defined QoS services is leaving a router and entering a switch, the router should set the CoS value of the packet because the switch can process the Layer 2 header.

Using This Command with the Enhanced Packet Marking Feature

You can use this command as part of the Enhanced Packet Marking feature to specify the “from-field” packet-marking category to be used for mapping and setting the CoS value. The “from-field” packet-marking categories are as follows:

• Precedence

• Differentiated services code point (DSCP)

If you specify a “from-field” category but do not specify the table keyword and the applicable table-map-nam e argument, the default action will be to copy the value associated with the “from-field” category as the CoS value. For instance, if you configure the setcosprecedence command, the precedence value will be copied and used as the CoS value.

You can do the same for the DSCP marking category. That is, you can configure the setcosdscp command, and the DSCP value will be copied and used as the CoS value.

Note	If you configure the setcosdscp command, only the first three bits (the class selector bits) of the DSCP field are used.

Usage Guidelines

This command is supported on the Gigabit Ethernet WAN interfaces on Cisco 7600 series routers that are configured with an Optical Service Module (OSM)-2+4GE-WAN+ OSM module only.

OSMs are not supported on Cisco 7600 series routers that are configured with a Supervisor Engine 32.

The 802.1P prioritization bits are used in the VLAN tag for QoS processing.

When the router copies the double-tagged QinQ packets to the destination interface, by default it uses the P bits from the outer (provider) VLAN tag. To preserve the P bits that are in the inner (customer) VLAN tag, use the setcoscos-inner command.

For the setcoscos-inner command to be effective, you must configure the appropriate interface or subinterface as a trusted interface using the mlsqostrust command. Otherwise, the interface or subinterface defaults to being untrusted, where the Layer 2 interface zeroes out the P bits of the incoming packets before the setcoscos-inner command can copy them to the outer VLAN tag.

The setcoscos-inner command is supported only for the subinterfaces that are configured with an inner (customer) VLAN. The setcoscos-inner command is not supported for the subinterfaces that use the out-range keyword on the bridge-domain (subinterface configuration) command or that are not configured with any form of the bridge-domain (subinterface configuration) command.

This behavior remains when you configure the setcoscos-inner command on a policy that is applied to a main interface. Thesetcoscos-inner command affects the subinterfaces that are configured with a specific inner VLAN but it does not affect the subinterfaces that are not configured with any VLAN or that are configured with the out-range keyword.

Usage Guidelines

This command was introduced in Cisco IOS Release 12.2(33)SRA to support marking of the inner CoS value when using multipoint bridging (MPB) features on the Enhanced FlexWAN module, and when using MPB features on SPAs with the Cisco 7600 SIP-200 and Cisco 7600 SIP-400 on the Cisco 7600 series router.

This command is not supported on the Cisco 7600 SIP-600.

On the Cisco 7600 SIP-200, this command is not supported with the setcos command on the same interface.

For more information about QoS and the forms of marking commands supported by the SIPs on the Cisco 7600 series router, refer to the Cisco 7600 Series SIP, SSC, and SPA Software Configuration Guide .

Usage Guidelines

This command was introduced in Cisco IOS Release 12.2(33)SRB and is limited to policies that are applied to the EVC service instances.

For classification, the reference to the outer and inner tags is made to the frames as seen on the wire - that is, for ingress frames, tags prior to the "rewrite", while the for egress, it is after the "rewrite" of the tags, if any.

For marking, the reference to the outer COS at the ingress is to the DBUS-COS and reference to the inner is to the COS in the first tag on the frame; while, at the egress, the reference to outer and inner COS is to the ones in the frame.

Usage Guidelines

The discard class value indicates the discard portion of the PHB. Use the setdiscard-class command only in DiffServ Tunneling Pipe mode. The discard class value is required when the input PHB marking will be used to classify packets on the output interface.

You can also use this command to specify the type of traffic that will be dropped when there is congestion.

Cisco 10000 Series Router

This command is supported only on the ESR-PRE2.

Usage Guidelines

Once the DSCP bit is set, other quality of service (QoS) features can then operate on the bit settings.

DSCP and Precedence Values Are Mutually Exclusive

The set dscp command cannot be used with the set precedence command to mark the same packet. The two values, DSCP and precedence, are mutually exclusive. A packet can have one value or the other, but not both.

Precedence Value and Queueing

The network gives priority (or some type of expedited handling) to marked traffic. Typically, you set the precedence value at the edge of the network (or administrative domain); data then is queued according to the precedence. Weighted fair queueing (WFQ) can speed up handling for high-precedence traffic at congestion points. Weighted Random Early Detection (WRED) ensures that high-precedence traffic has lower loss rates than other traffic during times of congestion.

Use of the “from-field” Packet-Marking Category

If you are using this command as part of the Enhanced Packet Marking feature, it can specify the “from-field” packet-marking category to be used for mapping and setting the DSCP value. The “from-field” packet-marking categories are as follows:

• Class of service (CoS)

• QoS group

If you specify a “from-field” category but do not specify the table keyword and the applicable table-map-name argument, the default action will be to copy the value associated with the “from-field” category as the DSCP value. For instance, if you configure the set dscp cos command, the CoS value will be copied and used as the DSCP value.

Note	The CoS field is a 3-bit field, and the DSCP field is a 6-bit field. If you configure the set dscp cos command, only the three bits of the CoS field will be used.

If you configure the set dscp qos-group command, the QoS group value will be copied and used as the DSCP value.

The valid range for the DSCP value is a number from 0 to 63. The valid value range for the QoS group is a number from 0 to 99. Therefore, when configuring the set dscp qos-group command, note the following points:

• If a QoS group value falls within both value ranges (for example, 44), the packet-marking value will be copied and the packets will be marked.

• If QoS group value exceeds the DSCP range (for example, 77), the packet-marking value will not be copied and the packet will not be marked. No action is taken.

Setting DSCP Values in IPv6 Environments

When this command is used in IPv6 environments, the default match occurs on both IPv4 and IPv6 packets. However, the actual packets set by this function are only those that meet the match criteria of the class map containing this function.

Setting DSCP Values for IPv6 Packets Only

To set DSCP values for IPv6 values only, you must also use the match protocol ipv6 command. Without that command, the precedence match defaults to match both IPv4 and IPv6 packets.

Setting DSCP Values for IPv4 Packets Only

To set DSCP values for IPv4 values only, you must use the appropriate match ip command. Without this command, the class map may match both IPv6 and IPv4 packets, depending on the other match criteria, and the DSCP values may act upon both types of packets.

Usage Guidelines

To disable this command in a traffic policy, use the nosetfr-de command in policy-map class configuration mode of the traffic policy.

If the DE bit is already set to 1, no changes are made to the frame.

Usage Guidelines

You can enter reserved keywords EF (expedited forwarding), AF11 (assured forwarding class AF11), and AF12 (assured forwarding class AF12) instead of numeric values for ip-dscp-value .

After the IP DSCP bit is set, other quality of service (QoS) features can operate on the bit settings.

You cannot mark a packet by the IP precedence using the setipprecedence (policy-map configuration) command and then mark the same packet with an IP DSCP value using the setipdscp command.

The network gives priority (or some type of expedited handling) to marked traffic. Typically, you set IP precedence at the edge of the network (or administrative domain); data is queued based on the precedence. Weighted Fair Queueing (WFQ) can speed up handling for high-precedence traffic at congestion points. Weighted Random Early Detection (WRED) ensures that high-precedence traffic has lower loss rates than other traffic during traffic congestion.

The setipprecedence (policy-map configuration) command is applied when you create a service policy in QoS policy-map configuration mode. This service policy is not attached to an interface or to an ATM virtual circuit. See the service-policy command for information on attaching a service policy to an interface.

When configuring policy-map class actions, note the following:

• For hardware-switched traffic, Policy Feature Card (PFC) QoS does not support the bandwidth , priority , queue-limit , or random-detect policy-map class commands. You can configure these commands because they can be used for software-switched traffic.

• PFC QoS does not support the setmpls or setqos-group policy-map class commands.

• PFC QoS supports the setipdsc p and setipprecedence policy-map class commands (see the “Configuring Policy Map Class Marking” section in the Cisco 7600 Series Router Cisco IOS Software Configuration Guide).

• You cannot do all three of the following in a policy-map class:

  • Mark traffic with the setipdscp or setipprecedence (policy-map configuration) commands
  • Configure the trust state
  • Configure policing

In a policy-map class, you can either mark traffic with the setipdscp or setipprecedence (policy-map configuration) commands or do one or both of the following:

• • Configure the trust state
  • Configure policing

Usage Guidelines

It is possible to configure L2TPv3 (or GRE) tunnel marking and the ip tos commands at the same time. However, Modular Quality of Service (QoS) Command-Line Interface (CLI) (MQC) (L2TPv3 or GRE) tunnel marking has higher priority over ip tos commands, meaning that tunnel marking always rewrites the IP header of the tunnel packet and overwrites the values set by ip tos commands. The order of enforcement is as follows when these commands are used simultaneously:

1. set ip dscp tunnel or set ip precedence tunnel (L2TPv3 or GRE tunnel marking)

2. ip tos reflect

3. ip tos tos-value

We recommend that you configure only L2TPv3 (or GRE) tunnel marking and reconfigure any peers configured with the ip tos commands to use L2TPv3 (or GRE) tunnel marking.

Note	For Cisco IOS Release 12.4(15)T2, marking GRE-tunneled packets is supported only on platforms equipped with a Cisco RPM-XF.

Usage Guidelines

The table below lists the value definitions for precedence values in the IP header. They are listed from least to most important.

After the IP precedence bits are set, other quality of service (QoS) features, such as Weighted Fair Queueing (WFQ) and Weighted Random Early Detection (WRED), operate on the bit settings.

The network priorities (or some type of expedited handling) mark traffic through the application of WFQ or WRED at points downstream in the network. Typically, you set IP precedence at the edge of the network (or administrative domain); data is queued based on the precedence. WFQ can speed up handling for certain precedence traffic at congestion points. WRED can ensure that certain precedence traffic has lower loss rates than other traffic during traffic congestion.

The setipprecedence command is applied when you create a service policy in policy-map configuration mode. This service policy is not attached to an interface or to an ATM virtual circuit. See the service-policy command for information on attaching a service policy to an interface.

Usage Guidelines

The table below lists the values for the number argument and the corresponding name argument for precedence values in the IP header. They are listed from least to most important.

You can set the precedence using either a number or the corresponding name. Once the IP Precedence bits are set, other QoS services such as weighted fair queueing (WFQ) and Weighted Random Early Detection (WRED) then operate on the bit settings.

The network gives priority (or some type of expedited handling) to marked traffic through the application of WFQ or WRED at points downstream in the network. Typically, you set IP Precedence at the edge of the network (or administrative domain); data then is queued based on the precedence. WFQ can speed up handling for certain precedence traffic at congestion points. WRED can ensure that certain precedence traffic has lower loss rates than other traffic during times of congestion.

The mapping from arguments such as routine and priority to a precedence value is useful only in some instances. That is, the use of the precedence bit is evolving. You can define the meaning of a precedence value by enabling other features that use the value. In the case of the high-end Internet QoS available from Cisco, IP Precedences can be used to establish classes of service that do not necessarily correspond numerically to better or worse handling in the network.

Use the route-map (IP) global configuration command with the match and set route-map configuration commands to define the conditions for redistributing routes from one routing protocol into another, or for policy routing. Each route-map command has an associated list of match and set commands. The match commands specify the match criteria--the conditions under which redistribution or policy routing is allowed for the current route-map command. The set commands specify the set actions--the particular redistribution or policy routing actions to perform if the criteria enforced by the match commands are met. The no route-map command deletes the route map.

The set route-map configuration commands specify the redistribution set actions to be performed when all of the match criteria of a route map are met.

Usage Guidelines

It is possible to configure L2TPv3 (or GRE) tunnel marking and the ip tos commands at the same time. However, Modular Quality of Service (QoS) Command-Line Interface (CLI) (MQC) (L2TPv3 or GRE) tunnel marking has higher priority over ip tos commands, meaning that tunnel marking always rewrites the IP header of the tunnel packet and overwrites the values set by ip tos commands. The order of enforcement is as follows when these commands are used simultaneously:

1. set ip dscp tunnel or set ip precedence tunnel (L2TPv3 or GRE tunnel marking)

2. ip tos reflect

3. ip tos tos-value

We recommend that you configure only L2TPv3 (or GRE) tunnel marking and reconfigure any peers configured with the ip tos commands to use L2TPv3 (or GRE) tunnel marking.

Note	For Cisco IOS Release 12.4(15)T2, marking GRE-tunneled packets is supported only on platforms equipped with a Cisco RPM-XF.

Usage Guidelines

This command allows you to set four bits in the ToS byte header. The table below shows the format of the four bits in binary form.

The T3 bit sets the delay. Setting T3 to 0 equals normal delay, and setting it to 1 equals low delay.

The T2 bit sets the throughput. Setting this bit to 0 equals normal throughput, and setting it to 1 equals maximum throughput. Similarly, the T1 and T0 bits set reliability and cost, respectively. Therefore, as an example, if you want to set a packet with the following requirements:

minimum delay T3 = 1

normal throughput T2 = 0

normal reliability T1 = 0

minimum monetary cost T0 = 1

You would set the ToS to 9, which is 1001 in binary format.

Use the route-map (IP) global configuration command with the match and set (route-map) configuration commands to define the conditions for redistributing routes from one routing protocol into another, or for policy routing. Each route-map command has an associated list of match andset commands. The match commands specify the match criteria--the conditions under which redistribution or policy routing is allowed for the current route-map command. Theset commands specify the set actions--the particular redistribution or policy routing actions to perform if the criteria enforced by the match commands are met. The no route-map command deletes the route map.

The set (route-map) commands specify the redistribution set actions to be performed when all of the match criteria of a route map are met.

Usage Guidelines

Command Compatibility

If a router is loaded with an image from Cisco IOS Release 12.2(13)T that contained an old configuration, the setip precedence command is still recognized. However, the set precedence command will be used in place of the set ip precedence command.

The set precedence command cannot be used with theset dscp command to mark the same packet. The two values, differentiated services code point (DSCP) and precedence, are mutually exclusive. A packet can have one value or the other, but not both.

Bit Settings

Once the precedence bits are set, other quality of service (QoS) features such as weighted fair queueing (WFQ) and Weighted Random Early Detection (WRED) then operate on the bit settings.

Precedence Value

The network gives priority (or some type of expedited handling) to marked traffic through the application of WFQ or WRED at points downstream in the network. Typically, you set the precedence value at the edge of the network (or administrative domain); data then is queued according to the specified precedence. WFQ can speed up handling for certain precedence traffic at congestion points. WRED can ensure that certain precedence traffic has lower loss rates than other traffic during times of congestion.

Using This Command with the Enhanced Packet Marking Feature

If you are using this command as part of the Enhanced Packet Marking feature, you can use this command to specify the “from-field” packet-marking category to be used for mapping and setting the precedence value. The “from-field” packet-marking categories are as follows:

• CoS

• QoS group

If you specify a “from-field” category but do not specify the table keyword and the applicable table-map-name argument, the default action will be to copy the value associated with the “from-field” category as the precedence value. For instance, if you configure the set precedence cos command, the CoS value will be copied and used as the precedence value.

You can do the same for the QoS group-marking category. That is, you can configure the set precedence qos-group command, and the QoS group value will be copied and used as the precedence value.

The valid value range for the precedence value is a number from 0 to 7. The valid value range for the QoS group is a number from 0 to 99. Therefore, when configuring the setprecedenceqos-group command, note the following points:

• If a QoS group value falls within both value ranges (for example, 6), the packet-marking value will be copied and the packets will be marked.

• If QoS group value exceeds the precedence range (for example, 10), the packet-marking value will not be copied, and the packet will not be marked. No action is taken.

Precedence Values in IPv6 Environments

When this command is used in IPv6 environments, it can set the value in both IPv4 and IPv6 packets. However, the actual packets set by this function are only those that meet the match criteria of the class map containing this function.

Setting Precedence Values for IPv6 Packets Only

To set the precedence values for IPv6 packets only, the match protocol ipv6 command must also be used in the class map that classified packets for this action. Without the match protocol ipv6 command, the class map may classify both IPv6 and IPv4 packets (depending on other match criteria), and the set precedence command will act upon both types of packets.

Setting Precedence Values for IPv4 Packets Only

To set the precedence values for IPv4 packets only, use a command involving the ip keyword like the match ip precedence or match ip dscp command or include the match protocol ip command along with the others in the class map. Without the additional ip keyword, the class map may match both IPv6 and IPv4 packets (depending on the other match criteria) and the set precedence or set dscp command may act upon both types of packets.

Usage Guidelines

The set qos-group command allows you to associate a group ID with a packet. The group ID can be used later to classify packets into QoS groups based as prefix, autonomous system, and community string.

A QoS group and discard class are required when the input per-hop behavior (PHB) marking will be used for classifying packets on the output interface.

Using This Command with the Enhanced Packet Marking Feature

If you are using this command as part of the Enhanced Packet Marking feature, you can use this command to specify the “from-field” packet-marking category to be used for mapping and setting the precedence value.

If you specify a “from-field” category but do not specify the table keyword and the applicable table-map-name argument, the default action will be to copy the value associated with the “from-field” category as the precedence value. For instance, if you enter set qos-groupprecedence , the precedence value will be copied and used as the QoS group value.

A packet is marked with a QoS group value only while it is being processed within the router. The QoS group value is not included in the packet’s header when the packet is transmitted over the output interface. However, the QoS group value can be used to set the value of a Layer 2 or Layer 3 field that is included as part of the packet’s headers (such as the MPLS EXP, CoS, and DSCP fields).

Note	The set qos-group cos and set qos-group precedence commands are equivalent to the mls qos trust cos and mls qos trust prec commands.

Tip

The set qos-group command cannot be applied until you create a service policy in policy-map configuration mode and then attach the service policy to an interface or ATM virtual circuit (VC). For information on attaching a service policy, refer to the “Modular Quality of Service Command-Line Interface Overview” chapter of the Cisco IOS Quality of Service Solutions Configuration Guide.

Usage Guidelines

Although multiple VLANs are allowed under a single PVC, the locally generated packets including the Address Resolution Protocol (ARP) packets are sent out with the class default VLAN ID only. The setvlaninner command must be applied within the class default.

Usage Guidelines

Traffic shaping limits the rate of transmission of data. In addition to using a specifically configured transmission rate, you can use Generic Traffic Shaping (GTS) to specify a derived transmission rate based on the level of congestion.

You can specify two types of traffic shaping; average rate shaping and peak rate shaping. Average rate shaping limits the transmission rate to the CIR. Using the CIR ensures that the average amount of traffic being sent conforms to the rate expected by the network.

Peak rate shaping configures the router to send more traffic than the CIR. To determine the peak rate, the router uses the following formula:

peak rate = CIR(1 + Be / Bc)

where:

• Be is the Excess Burst size.

• Bc is the Committed Burst size.

Peak rate shaping allows the router to burst higher than average rate shaping. However, using peak rate shaping, the traffic sent above the CIR (the delta) could be dropped if the network becomes congested.

If your network has additional bandwidth available (over the provisioned CIR) and the application or class can tolerate occasional packet loss, that extra bandwidth can be exploited through the use of peak rate shaping. However, there may be occasional packet drops when network congestion occurs. If the traffic being sent to the network must strictly conform to the configured network provisioned CIR, then you should use average traffic shaping.

Usage Guidelines

Committed Information Rate

This command calculates the committed information rate (CIR) on the basis of a percentage of the available bandwidth on the interface. Once a policy map is attached to the interface, the equivalent CIR value in bits per second (bps) is calculated on the basis of the interface bandwidth and the percent value entered with this command. The showpolicy-mapinterface command can then be used to verify the CIR bps value calculated.

The calculated CIR bps rate must be in the range of 8000 and 154,400,000 bps. If the rate is less than 8000 bps, the associated policy map cannot be attached to the interface. If the interface bandwidth changes (for example, more is added), the CIR bps values are recalculated on the basis of the revised amount of bandwidth. If the CIR percentage is changed after the policy map is attached to the interface, the bps value of the CIR is recalculated.

Conform Burst and Peak Burst Sizes in Milliseconds

This command also allows you to specify the values for the conform burst size and the peak burst size in milliseconds. If you want bandwidth to be calculated as a percentage, the conform burst size and the peak burst size must be specified in milliseconds (ms).

The traffic shape converge rate depends on the traffic pattern and the time slice (Tc) parameter, which is directly affected by the bc that you configured. The Tc and the average rate configured are used to calculate bits per interval sustained. Therefore, to ensure that the shape rate is enforced, use a bc that results in a Tc greater than 10 ms.

Hierarchical Policy Maps

The shape (percent) command, when used in “child” (hierarchical) policy maps, is not supported on the Cisco 7500, the Cisco 7200, or lower series routers. Therefore, the shape (percent) command cannot be configured for use in hierarchical policy maps on these routers.

How Bandwidth Is Calculated

The shape (percent) command is often used in conjunction with the bandwidth and priority commands. The bandwidth and priority commands can be used to calculate the total amount of bandwidth available on an entity (for example, a physical interface). When the bandwidth and priority commands calculate the total amount of bandwidth available on an entity, the following guidelines are invoked:

• If the entity is a physical interface, the total bandwidth is the bandwidth on the physical interface.

• If the entity is a shaped ATM permanent virtual circuit (PVC), the total bandwidth is calculated as follows:

  • For a variable bit rate (VBR) virtual circuit (VC), the sustained cell rate (SCR) is used in the calculation.
  • For an available bit rate (ABR) VC, the minimum cell rate (MCR) is used in the calculation.

For more information on bandwidth allocation, see the “Congestion Management Overview” chapter in the Cisco IOS Quality of Service Solutions Configuration Guide.

Note	This command cannot be used with the shapeadaptive command.

Usage Guidelines

The measurement interval is the committed burst size (Bc) divided by committed information rate (CIR). Bc cannot be set to 0. If the measurement interval is too large (greater than 128 milliseconds), the system subdivides it into smaller intervals.

If you do not specify the committed burst size (Bc) and the excess burst size (Be), the algorithm decides the default values for the shape entity. The algorithm uses a 4 milliseconds measurement interval, so Bc is CIR * (4 / 1000).

Burst sizes larger than the default committed burst size (Bc) need to be explicitly specified. The larger the Bc, the longer the measurement interval. A long measurement interval may affect voice traffic latency, if applicable.

When the excess burst size (Be) is not configured, the default value is equal to the committed burst size (Bc).

Traffic Shaping on the Cisco 10000 Series Performance Routing Engine

The Cisco 10000 series router does not support the peak keyword.

On the PRE2, you specify a shape rate and a unit for the rate. Valid values for the rate are from 1 to 2488320000 and units are bps, kbps, mbps, gbps. The default unit is kbps. For example:

shape 128000 bps

On the PRE3, you only need to specify a shape rate. Because the unit is always bps on the PRE3, the unit argument is not available. Valid values for the shape rate are from 1000 to 2488320000.

shape 1000

The PRE3 accepts the PRE2 shape command as a hidden command. However, the PRE3 rejects the PRE2 shape command if the specified rate is outside the valid PRE3 shape rate range (1000 to 2488320000).

Traffic Shaping Overhead Accounting for ATM (Cisco 7300 Series Router, Cisco 7600 Series Router, and Cisco 10000 Series Router)

When configuring ATM overhead accounting on the Cisco 7300 series router, the Cisco 7600 series router, or the Cisco 10000 series router, you must specify the BRAS-DSLAM, DSLAM-CPE, and subscriber line encapsulation types. The router supports the following subscriber line encapsulation types:

• snap-rbe

• mux-rbe

• snap-dot1q-rbe

• mux-dot1q-rbe

• snap-pppoa

• mux-pppoa

• snap-1483routed

• mux-1483routed

For hierarchical policies, configure ATM overhead accounting in the following ways:

• Enabled on parent--If you enable ATM overhead accounting on a parent policy, you are not required to enable accounting on the child policy.

• Enabled on child and parent--If you enable ATM overhead accounting on a child policy, then you must enable ATM overhead accounting on the parent policy.

The encapsulation types must match for the child and parent policies.

The user-defined offset values must match for the child and parent policies.

Usage Guidelines

If traffic shaping is not enabled, this command has no effect.

When continuous BECN messages are received, the shape entity immediately decreases its maximum shape rate by one-fourth for each BECN message received until it reaches the lower bound committed information rate (CIR). If, after several intervals, the interface has not received another BECN and traffic is waiting in the shape queue, the shape entity increases the shape rate back to the maximum rate by 1/16 for each interval. A shape entity configured with the shapeadaptive mean-rate-lower-bound command will always be shaped between the mean rate upper bound and the mean rate lower bound.

Note	The shapeadaptive command cannot be used with the shape(percent) command.

Usage Guidelines

When the downstream Frame Relay switch is congested, a Frame Relay interface or point-to-point interface receives a Frame Relay message with the FECN bit on. This message may be an indication that no traffic is waiting to carry a BECN to the far end (voice/multimedia traffic is one-way). When the shapefecn-adapt command is configured, a small buffer is allocated and a Frame Relay TEST RESPONSE is built on behalf of the Frame Relay switch. The Frame Relay TEST RESPONSE is equipped with the triggering data-link connection identifier (DLCI) of the triggering mechanism. It also sets the BECN bit and sends it out to the wire.

Usage Guidelines

You can specify the maximum number of buffers allowed on shaping queues for each class configured to use Generic Traffic Shaping (GTS).

You configure this command under a class in a policy map. However, the shapemax-buffers command is not supported for HQF in Cisco IOS Release 12.4(20)T. Use the queue-limit command, which provides similar functionality.