-
Cisco IOS Quality of Service Solutions Command Reference, Release 12.3 T
-
Introduction
-
Quality of Service Commands: A through F
-
Quality of Service Commands: I
-
Quality of Sevice Commands: M
-
Quality of Service Commands: O through P
-
Quality of Service Commands: Q through R
-
Quality of Service Commands: send qdm message through show ip nbar pdlm
-
Quality of Service Commands: show ip nbar port-map through show qdm status
-
Quality of Service Commands: show queue through vc-hold-queue
-
Table Of Contents
random-detect discard-class-based
random-detect exponential-weighting-constant
random-detect flow average-depth-factor
qos pre-classify
To enable quality of service (QoS) preclassification, use the qos pre-classify command in interface configuration mode. To disable the QoS preclassification feature, use the no form of this command.
qos pre-classify
no qos pre-classify
Syntax Description
This command has no arguments or keywords.
Defaults
Disabled
Command Modes
Interface configuration
Command History
Usage Guidelines
This command is restricted to tunnel interfaces, virtual templates, and crypto maps. The qos pre-classify command is unavailable on all other interface types.
The qos pre-classify command can be enabled for IP packets only.
Examples
The following example enables the QoS for Virtual Private Networks (VPNs) feature on tunnel interfaces and virtual templates:
Router(config-if)# qos pre-classifyThe following example enables the QoS for VPNs feature on crypto maps:
Router(config-crypto-map)# qos pre-classifyRelated Commands
queue-limit
To specify or modify the maximum number of packets the queue can hold for a class policy configured in a policy map, use the queue-limit command in policy-map class configuration mode. To remove the queue packet limit from a class, use the no form of this command.
queue-limit number-of-packets
no queue-limit number-of-packets
Syntax Description
number-of-packets
A number in the range from 1 to 64 specifying the maximum number of packets that the queue for this class can accumulate.
Defaults
On the Versatile Interface Processor (VIP)-based platforms, the default value is chosen as a function of the bandwidth assigned to the traffic class. The default value is also based on available buffer memory. If sufficient buffer memory is available, the default queue-limit value is equal to the number of 250-byte packets that would lead to a latency of 500 milliseconds (ms) when the packets are delivered at the configured rate. For example, if two 250-byte packets are required to lead to a latency of 500 ms, the default number-of-packets value would be 2.
On all other platforms, the default is 64.
Command Modes
Policy-map class configuration
Command History
Usage Guidelines
Weighted fair queueing (WFQ) creates a queue for every class for which a class map is defined. Packets satisfying the match criteria for a class accumulate in the queue reserved for the class until they are sent, which occurs when the queue is serviced by the fair queueing process. When the maximum packet threshold you defined for the class is reached, enqueueing of any further packets to the class queue causes tail drop or, if Weighted Random Early Detection (WRED) is configured for the class policy, packet drop to take effect.
Overriding Queue Limits Set by the Bandwidth Command
The bandwidth command can be used with the Modular Command-Line Interface (MQC) to specify the bandwidth for a particular class. When used with the MQC, the bandwidth command uses a default queue limit for the class. This queue limit can be modified using the queue-limit command, thereby overriding the default set by the bandwidth command.
Note
Using the queue-limit command to modify the default queue-limit is especially important for higher-speed interfaces, in order to meet the minimum bandwidth guarantees required by the interface.
Examples
The following example configures a policy map called policy11 to contain policy for a class called acl203. Policy for this class is set so that the queue reserved for it has a maximum packet limit of 40.
policy-map policy11class acl203bandwidth 2000queue-limit 40Related Commands
queue-list default
To assign a priority queue for those packets that do not match any other rule in the queue list, use the queue-list default command in global configuration mode. To restore the default value, use the no form of this command.
queue-list list-number default queue-number
no queue-list list-number default queue-number
Syntax Description
list-number
Number of the queue list. Any number from 1 to 16 that identifies the queue list.
queue-number
Number of the queue. Any number from 1 to 16.
Defaults
Disabled
The default number of the queue list is queue number 1.
Command Modes
Global configuration
Command History
Usage Guidelines
When you use multiple rules, remember that the system reads the queue-list commands in order of appearance. When classifying a packet, the system searches the list of rules specified by queue-list commands for a matching protocol or interface type. When a match is found, the system assigns the packet to the appropriate queue. The system searches the list in the order specified, and the first matching rule terminates the search.
Queue number 0 is a system queue. It is emptied before any of the other queues are processed. The system enqueues high-priority packets, such as keepalives, to this queue.
Use the show interfaces command to display the current status of the output queues.
Examples
In the following example, the default queue for list 10 is set to queue number 2:
queue-list 10 default 2Related Commands
queue-list interface
To establish queueing priorities on packets entering on an interface, use the queue-list interface command in global configuration mode. To remove an entry from the list, use the no form of this command.
queue-list list-number interface interface-type interface-number queue-number
no queue-list list-number interface interface-type interface-number queue-number
Syntax Description
Defaults
No queueing priorities are established.
Command Modes
Global configuration
Command History
Usage Guidelines
When you use multiple rules, remember that the system reads the queue-list commands in order of appearance. When classifying a packet, the system searches the list of rules specified by queue-list commands for a matching protocol or interface type. When a match is found, the system assigns the packet to the appropriate queue. The list is searched in the order specified, and the first matching rule terminates the search.
Examples
In the following example, queue list 4 establishes queueing priorities for packets entering on interface tunnel 3. The queue number assigned is 10.
queue-list 4 interface tunnel 3 10Related Commands
queue-list lowest-custom
To set the lowest number for a queue to be treated as a custom queue, use the queue-list lowest-custom command in global configuration mode. To restore the default value, use the no form of this command.
queue-list list-number lowest-custom queue-number
no queue-list list-number lowest-custom queue-number
Syntax Description
list-number
Number of the queue list. Any number from 1 to 16 that identifies the queue list.
queue-number
Number of the queue. Any number from 1 to 16.
Defaults
The default number of the lowest custom queue is 1.
Command Modes
Global configuration
Command History
Usage Guidelines
All queues from queue 0 to the queue prior to the one specified in the queue-list lowest-custom command use the priority queue. (Queue 0 has the highest priority.)
All queues from the one specified in the queue-list lowest-custom command to queue 16 use a round-robin scheduler.
Use the show queueing custom command to display the current custom queue configuration.
Examples
In the following example, the lowest custom value is set to 2 for queue list 4:
queue-list 4 lowest-custom 2Related Commands
queue-list protocol
To establish queueing priority based upon the protocol type, use the queue-list protocol command in global configuration mode. To remove an entry from the list, use the no form of this command.
queue-list list-number protocol protocol-name queue-number queue-keyword keyword-value
no queue-list list-number protocol protocol-name queue-number queue-keyword keyword-value
Syntax Description
list-number
Number of the queue list. Any number from 1 to 16.
protocol-name
Protocol type: aarp, appletalk, arp, bridge (transparent), clns, clns_es, clns_is, cmns, compressedtcp, decnet, decnet_node, decnet_routerl1, decnet_routerl2, dlsw, ip, ipx, pad, rsrb, stun and x25.
queue-number
Number of the queue. Any number from 1 to 16.
queue-keyword keyword-value
Possible keywords are fragments, gt, list, lt, tcp, and udp. See Table 12 of the priority-list protocol command.
Defaults
No queueing priorities are established.
Command Modes
Global configuration
Command History
Usage Guidelines
When you use multiple rules for a single protocol, remember that the system reads the queue-list commands in order of appearance. When classifying a packet, the system searches the list of rules specified by queue-list commands for a matching protocol. When a match is found, the system assigns the packet to the appropriate queue. The system searches the list in the order specified, and the first matching rule terminates the search.
The decnet_router-l1 keyword refers to the multicast address for all level 1 routers, which are intra-area routers, and the decnet_router-l2 keyword refers to all level 2 routers, which are interarea routers.
The dlsw, rsrb, and stun keywords refer only to direct encapsulation.
Use Table 12, Table 13, and Table 14 in the priority-list protocol command section to configure the queueing priorities for your system.
Examples
The following example assigns 1 as the custom queue list, specifies DECnet as the protocol type, and assigns 3 as a queue number to the packets sent on this interface:
queue-list 1 protocol decnet 3The following example assigns DECnet packets with a size greater than 200 bytes to queue number 2:
queue-list 2 protocol decnet 2 gt 200The following example assigns DECnet packets with a size less than 200 bytes to queue number 2:
queue-list 4 protocol decnet 2 lt 200The following example assigns traffic that matches IP access list 10 to queue number 1:
queue-list 1 protocol ip 1 list 10The following example assigns Telnet packets to queue number 2:
queue-list 4 protocol ip 2 tcp 23The following example assigns User Datagram Protocol (UDP) Domain Name Service packets to queue number 2:
queue-list 4 protocol ip 2 udp 53The following example assigns traffic that matches Ethernet type code access list 201 to queue number 1:
queue-list 1 protocol bridge 1 list 201Related Commands
queue-list queue byte-count
To specify how many bytes the system allows to be delivered from a given queue during a particular cycle, use the queue-list queue byte-count command in global configuration mode. To return the byte count to the default value, use the no form of this command.
queue-list list-number queue queue-number byte-count byte-count-number
no queue-list list-number queue queue-number byte-count byte-count-number
Syntax Description
Defaults
This command is disabled by default. The default byte count is 1500 bytes.
Command Modes
Global configuration
Command History
Examples
In the following example, queue list 9 establishes the byte count as 1400 for queue number 10:
queue-list 9 queue 10 byte-count 1400Related Commands
queue-list queue limit
To designate the queue length limit for a queue, use the queue-list queue limit command in global configuration mode. To return the queue length to the default value, use the no form of this command.
queue-list list-number queue queue-number limit limit-number
no queue-list list-number queue queue-number limit limit-number
Syntax Description
Defaults
The default queue length limit is 20 entries.
Command Modes
Global configuration
Command History
Examples
In the following example, the queue length of queue 10 is increased to 40:
queue-list 5 queue 10 limit 40Related Commands
random-detect discard-class
To configure the weighted random early detection (WRED) parameters for a discard-class value for a class policy in a policy map, use the random-detect discard-class command in policy-map class configuration mode. To disable this feature, use the no form of this command.
random-detect discard-class value min-threshold max-threshold mark-prob-denominator
no random-detect discard-class value min-threshold max-threshold mark-prob-denominator
Syntax Description
Defaults
To return the values to the default for the discard class, use the no form of this command.
Command Modes
Policy-map class configuration
Command History
Usage Guidelines
When you configure the random-detect discard-class command on an interface, packets are given preferential treatment based on the discard class of the packet. Use the random-detect discard-class command to adjust the discard class for different discard class values.
Examples
The following example shows that if the discard class is 2, there is a 10 percent chance that packets will be dropped if there are more packets than the minimum threshold of 100 packets or there are fewer packets than the maximum threshold of 200 packets:
policy-map set-MPLS-PHBclass IP-AF11bandwidth percent 40random-detect discard-class-basedrandom-detect-discard-class 2 100 200 10Related Commands
random-detect discard-class-based
To base weighted random early detection (WRED) on the discard class value of a packet, use the random-detect discard-class-based command in policy-map class configuration mode. To disable this feature, use the no form of this command.
random-detect discard-class-based
no random-detect discard-class-based
Syntax Description
This command has no arguments or keywords.
Defaults
The defaults are router-dependent.
Command Modes
Policy-map class configuration
Command History
Usage Guidelines
Enter this command so that WRED is based on the discard class instead of on the IP precedence field.
Examples
The following example shows that random detect is based on the discard class value of a packet:
policy-map nameclass-namebandwidth percent 40random-detect discard-class-basedRelated Commands
random-detect dscp
To change the minimum and maximum packet thresholds for the differentiated services code point (DSCP) value, use the random-detect dscp command in interface configuration mode. To return the minimum and maximum packet thresholds to the default for the DSCP value, use the no form of this command.
random-detect dscp dscpvalue min-threshold max-threshold [mark-probability-denominator]
no random-detect dscp dscpvalue min-threshold max-threshold [mark-probability-denominator]
Syntax Description
Defaults
If WRED is using the DSCP value to calculate the drop probability of a packet, all entries of the DSCP table are initialized with the default settings shown in Table 16 in the "Usage Guidelines" section of this command.
Command Modes
Interface configuration
Command History
Usage Guidelines
The random-detect dscp command allows you to specify the DSCP value. The DSCP value can be a number from 0 to 63, or it can be one of the following keywords: ef, af11, af12, af13, af21, af22, af23, af31, af32, af33, af41, af42, af43, cs1, cs2, cs3, cs4, cs5, or cs7.
This command must be used in conjunction with the random-detect (interface) command.
Additionally, the random-detect dscp command is available only if you specified the dscp-based argument when using the random-detect (interface) command.
Table 16 lists the default settings used by the random-detect dscp command for the DSCP value specified. Table 16 lists the DSCP value, and its corresponding minimum threshold, maximum threshold, and mark probability. The last row of the table (the row labeled "default") shows the default settings used for any DSCP value not specifically shown in the table.
Examples
The following example enables WRED to use the DSCP value af22. The minimum threshold for DSCP value af22 is 28, the maximum threshold is 40, and the mark probability is 10.
random-detect dscp af22 20 40 10Related Commands
Enables WRED or DWRED.
show queueing
Lists all or selected configured queueing strategies.
show queueing interface
Displays the queueing statistics of an interface or VC.
random-detect (interface)
To enable Weighted Random Early Detection (WRED) or distributed WRED (DWRED), use the random-detect command in interface configuration mode. To configure WRED as class policy in a policy map, use the random-detect interface and policy-map class configuration command. To disable WRED or DWRED, use the no form of this command.
random-detect [dscp-based | prec-based]
no random-detect [dscp-based | prec-based]
Syntax Description
Defaults
WRED and DWRED are disabled by default.
If you choose not to use either the dscp-based or the prec-based keywords, WRED uses the IP Precedence value (the default method) to calculate drop probability for the packet.
Command Modes
Interface configuration when used on an interface
Policy-map class configuration when used to specify class policy in a policy map
Command History
Usage Guidelines
WRED is a congestion avoidance mechanism that slows traffic by randomly dropping packets when congestion exists. DWRED is similar to WRED but uses the Versatile Interface Processor (VIP) instead of the Route Switch Processor (RSP). WRED and DWRED are most useful with protocols like TCP that respond to dropped packets by decreasing the transmission rate.
The router automatically determines parameters to use in the WRED calculations. To change these parameters, use the random-detect precedence command.
The DWRED feature is supported only on Cisco 7000 series routers with an RSP7000 card and Cisco 7500 series routers with a VIP2-40 or greater interface processor. A VIP2-50 interface processor is strongly recommended when the aggregate line rate of the port adapters on the VIP is greater than DS3. A VIP2-50 interface processor is required for OC-3 rates.
To use DWRED, distributed Cisco Express Forwarding (dCEF) switching must first be enabled on the interface. For more information on dCEF, refer to the Cisco IOS Switching Services Configuration Guide and the Cisco IOS Switching Services Command Reference.
WRED in a Policy Map
You can configure WRED as part of the policy for a standard class or the default class. The WRED random-detect command and the weighted fair queueing (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.
To configure a policy map and create class policies, use the policy-map and class (policy-map) commands. When specifying class policy within a policy map, you can use the random-detect command with either of the following commands:
•
•
Note that if you use WRED packet drop instead of tail drop for one or more classes composing a policy map, you must ensure that WRED is not configured for the interface to which you attach that service policy.
The DWRED feature is not supported for class policy.
Two Methods for Calculating the Drop Probability of a Packet
This command includes two optional keywords, dscp-based and prec-based, that determine the method WRED uses to calculate the drop probability of a packet.
Note the following points when deciding which method to instruct WRED to use:
•
•
•
•
Examples
The following example configures WRED on the High-Speed Serial Interface (HSSI) 0/0/0 interface:
interface Hssi0/0/0random-detectThe following example configures the policy map called policy1 to contain policy specification for the class called class1. During times of congestion, WRED packet drop is used instead of tail drop.
! The following commands create the class map called class1:class-map class1match input-interface fastethernet0/1! The following commands define policy1 to contain policy specification for class1:policy-map policy1class class1bandwidth 1000random-detectThe following example enables WRED to use the DSCP value 8. The minimum threshold for the DSCP value 8 is 24 and the maximum threshold is 40. This configuration was performed at the interface level.
Router(config)# interface serial0/0Router(config-if)# random-detect dscp-basedRouter(config-if)# random-detect dscp 8 24 40The following example enables WRED to use the DSCP value 8 for class c1. The minimum threshold for DSCP value 8 is 24 and the maximum threshold is 40. The last line attaches the service policy to the output interface or virtual circuit (VC) p1.
Router(config-if)# class-map c1Router(config-cmap)# match access-group 101Router(config-if)# policy-map p1Router(config-pmap)# class c1Router(config-pmap-c)# bandwidth 48Router(config-pmap-c)# random-detect dscp-basedRouter(config-pmap-c)# random-detect dscp 8 24 40Router(config-if)# service-policy output p1Related Commands
random-detect (per VC)
To enable per-virtual circuit (VC) Weighted Random Early Detection (WRED) or per-VC VIP-distributed WRED (DWRED), use the random-detect command in VC submode mode. To disable per-VC WRED and per-VC DWRED, use the no form of this command.
random-detect [attach group-name]
no random-detect [attach group-name]
Syntax Description
Defaults
WRED and DWRED are disabled by default.
Command Modes
VC submode
Command History
Usage Guidelines
WRED is a congestion avoidance mechanism that slows traffic by randomly dropping packets when congestion exists. DWRED is similar to WRED but uses the Versatile Interface Processor (VIP) instead of the Route Switch Processor (RSP). WRED and DWRED are most useful with protocols like TCP that respond to dropped packets by decreasing the transmission rate.
WRED and DWRED are configurable at the interface and per-VC levels. The VC-level WRED or DWRED configuration will override the interface-level configuration if WRED or DWRED is also configured at the interface level.
Use this command to configure a single ATM VC or a VC that is a member of a bundle.
Note the following points when using the random-detect (per VC) command:
•
•
When this command is used to configure an interface-level WRED or DWRED group to include per-VC WRED or DWRED as a drop policy, the configured WRED or DWRED group parameters are inherited under the following conditions:
•
•
When an interface-level WRED or DWRED group configuration is removed, per-VC WRED or DWRED parameters are removed from any VC that inherited them from the configured interface-level WRED or DWRED group.
When an interface-level WRED or DWRED group configuration is modified, per-VC WRED or DWRED parameters are modified accordingly if the WRED or DWRED parameters were inherited from the configured interface-level WRED or DWRED group configuration.
This command is only supported on interfaces that are capable of VC-level queueing. The only currently supported interface is the Enhanced ATM port adapter (PA-A3).
The DWRED feature is only supported on Cisco 7000 series routers with an RSP7000 card and Cisco 7500 series routers with a VIP2-40 or greater interface processor. A VIP2-50 interface processor is strongly recommended when the aggregate line rate of the port adapters on the VIP is greater than DS3. A VIP2-50 interface processor is required for OC-3 rates.
To use DWRED, distributed Cisco Express Forwarding (dCEF) switching must first be enabled on the interface. For more information on dCEF, refer to the Cisco IOS Switching Services Configuration Guide and the Cisco IOS Switching Services Command Reference.
Examples
The following example configures per-VC WRED for the permanent virtual circuit (PVC) called cisco. Because the attach keyword was not used, WRED uses default parameters.
pvc cisco 46random-detectThe following example creates a DWRED group called Rome and then applies the parameter group to an ATM PVC:
! The following commands create the DWRED parameter group Rome:random-detect-group Romeprecedence rsvp 46 50 10precedence 1 32 50 10precedence 2 34 50 10precedence 3 36 50 10precedence 4 38 50 10precedence 5 40 50 10precedence 6 42 50 10precedence 7 44 50 10exitexit! The following commands create a PVC on an ATM interface and then apply the! DWRED group Rome to that PVC:interface ATM2/0.23 point-to-pointip address 10.9.23.10 255.255.255.0no ip mroute-cachepvc vc1 201/201random-detect attach Romevbr-nrt 2000 1000 200encapsulation aal5snapThe following show queueing command displays the current settings for each of the IP Precedences following configuration of per-VC DWRED:
Router# show queueing random-detect interface atm2/0.23 vc 201/201random-detect group Rome:exponential weight 9class min-threshold max-threshold mark-probability----------------------------------------------------------0 30 50 1/101 32 50 1/102 34 50 1/103 36 50 1/104 38 50 1/105 40 50 1/106 42 50 1/107 44 50 1/10rsvp 46 50 1/10Related Commands
random-detect ecn
To enable explicit congestion notification (ECN), use the random-detect ecn command in policy-map class configuration mode. To disable ECN, use the no form of this command.
random-detect ecn
no random-detect ecn
Syntax Description
This command has no arguments or keywords.
Defaults
By default, ECN is disabled.
Command Modes
Policy-map class configuration
Command History
Usage Guidelines
If ECN is enabled, ECN can be used whether Weighted Random Early Detection (WRED) is based on the IP precedence value or the differentiated services code point (DSCP) value.
Examples
The following example enables ECN in a policy map called "pol1":
Router(config)# policy-map pol1Router(config-pmap)# class class-defaultRouter(config-pmap)# bandwidth per 70Router(config-pmap-c)# random-detectRouter(config-pmap-c)# random-detect ecnRelated Commands
random-detect exponential-weighting-constant
To configure the Weighted Random Early Detection (WRED) and distributed WRED (DWRED) exponential weight factor for the average queue size calculation for the queue, use the random-detect exponential-weighting-constant command in interface configuration mode. To configure the exponential weight factor for the average queue size calculation for the queue reserved for a class, use the random-detect exponential-weighting-constant command in policy-map class configuration mode. To return the value to the default, use the no form of this command.
random-detect exponential-weighting-constant exponent
no random-detect exponential-weighting-constant
Syntax Description
Defaults
The default exponential weight factor is 9.
Command Modes
Interface configuration when used on an interface
Policy-map class configuration when used to specify class policy in a policy map, or when used in the Modular Quality of Service Command-Line Interface (MQC)
Command History
Usage Guidelines
WRED is a congestion avoidance mechanism that slows traffic by randomly dropping packets when congestion exists. DWRED is similar to WRED but uses the Versatile Interface Processor (VIP) instead of the Route Switch Processor (RSP). WRED and DWRED are most useful with protocols like TCP that respond to dropped packets by decreasing the transmission rate.
Use this command to change the exponent used in the average queue size calculation for the WRED and DWRED services. You can also use this command to configure the exponential weight factor for the average queue size calculation for the queue reserved for a class
Note
The DWRED feature is not supported for class policy.
The DWRED feature is only supported on Cisco 7000 series routers with an RSP7000 card and Cisco 7500 series routers with a VIP2-40 or greater interface processor. A VIP2-50 interface processor is strongly recommended when the aggregate line rate of the port adapters on the VIP is greater than DS3. A VIP2-50 interface processor is required for OC-3 rates.
To use DWRED, distributed Cisco Express Forwarding (dCEF) switching must first be enabled on the interface. For more information on dCEF, refer to the Cisco IOS Switching Services Configuration Guide and the Cisco IOS Switching Services Command Reference.
Examples
The following example configures WRED on an interface with a weight factor of 10:
interface Hssi0/0/0description 45Mbps to R1ip address 10.200.14.250 255.255.255.252random-detectrandom-detect exponential-weighting-constant 10The following example configures the policy map called policy1 to contain policy specification for the class called class1. During times of congestion, WRED packet drop is used instead of tail drop. The weight factor used for the average queue size calculation for the queue for class1 is 12.
! The following commands create the class map called class1:class-map class1match input-interface FE0/1! The following commands define policy1 to contain policy specification for class1:policy-map policy1class class1bandwidth 1000random-detectrandom-detect exponential-weighting-constant 12The following example configures policy for a traffic class named int10 to configure the exponential weight factor as 12. This is the weight factor used for the average queue size calculation for the queue for traffic class int10. WRED packet drop is used for congestion avoidance for traffic class int10, not tail drop.
policy-map policy12 class int10 bandwidth 2000random-detect exponential-weighting-constant 12Related Commands
random-detect flow
To enable flow-based Weighted Random Early Detection (WRED), use the random-detect flow command in interface configuration mode. To disable flow-based WRED, use the no form of this command.
random-detect flow
no random-detect flow
Syntax Description
This command has no arguments or keywords.
Defaults
Flow-based WRED is disabled by default.
Command Modes
Interface configuration
Command History
Usage Guidelines
You must use this command to enable flow-based WRED before you can use the random-detect flow average-depth-factor and random-detect flow count commands to further configure the parameters of flow-based WRED.
Before you can enable flow-based WRED, you must enable and configure WRED. For complete information, refer to the Cisco IOS Quality of Service Solutions Configuration Guide.
Examples
The following example enables flow-based WRED on serial interface 1:
interface Serial1random-detectrandom-detect flowRelated Commands
random-detect flow average-depth-factor
To set the multiplier to be used in determining the average depth factor for a flow when flow-based Weighted Random Early Detection (WRED) is enabled, use the random-detect flow average-depth-factor command in interface configuration mode. To remove the current flow average depth factor value, use the no form of this command.
random-detect flow average-depth-factor scaling-factor
no random-detect flow average-depth-factor scaling-factor
Syntax Description
Defaults
The default average depth factor is 4.
Command Modes
Interface configuration
Command History
Usage Guidelines
Use this command to specify the scaling factor that flow-based WRED should use in scaling the number of buffers available per flow and in determining the number of packets allowed in the output queue for each active flow. This scaling factor is common to all flows. The outcome of the scaled number of buffers becomes the per-flow limit.
If this command is not used and flow-based WRED is enabled, the average depth scaling factor defaults to 4.
A flow is considered nonadaptive—that is, it takes up too much of the resources—when the average flow depth times the specified multiplier (scaling factor) is less than the depth for the flow, for example:
average-flow-depth * (scaling factor) < flow-depth
Before you use this command, you must use the random-detect flow command to enable flow-based WRED for the interface. To configure flow-based WRED, you may also use the random-detect flow count command.
Examples
The following example enables flow-based WRED on serial interface 1 and sets the scaling factor for the average flow depth to 8:
interface Serial1random-detectrandom-detect flowrandom-detect flow average-depth-factor 8Related Commands
random-detect flow count
To set the flow count for flow-based Weighted Random Early Detection (WRED), use the random-detect flow count command in interface configuration mode. To remove the current flow count value, use the no form of this command.
random-detect flow count number
no random-detect flow count number
Syntax Description
Defaults
256
Command Modes
Interface configuration
Command History
Usage Guidelines
Before you use this command, you must use the random-detect flow command to enable flow-based WRED for the interface.
Examples
The following example enables flow-based WRED on serial interface 1 and sets the flow threshold constant to 16:
interface Serial1random-detectrandom-detect flowrandom-detect flow count 16Related Commands
random-detect-group
To define the Weighted Random Early Detection (WRED) or distributed WRED (DWRED) parameter group, use the random-detect group command in global configuration mode. To delete the WRED or DWRED parameter group, use the no form of this command.
random-detect-group group-name [dscp-based | prec-based]
no random-detect-group group-name [dscp-based | prec-based]
Syntax Description
Defaults
No WRED or DWRED parameter group exists.
If you choose not to use either the dscp-based or the prec-based keywords, WRED uses the IP Precedence value (the default method) to calculate drop probability for the packet.
Command Modes
Global configuration
Command History
Usage Guidelines
WRED is a congestion avoidance mechanism that slows traffic by randomly dropping packets when there is congestion. DWRED is similar to WRED but uses the Versatile Interface Processor (VIP) instead of the Route Switch Processor (RSP). WRED and DWRED are most useful when the traffic uses protocols such as TCP that respond to dropped packets by decreasing the transmission rate.
The router automatically determines parameters to use in the WRED calculations. If you want to change these parameters for a group, use the exponential-weighting-constant or precedence command.
Two Methods for Calculating the Drop Probability of a Packet
This command includes two optional arguments, dscp-based and prec-based, that determine the method WRED uses to calculate the drop probability of a packet.
Note the following points when deciding which method to instruct WRED to use:
•
•
•
•
Examples
The following example defines the WRED parameter group called sanjose:
random-detect-group sanjoseprecedence 0 32 256 100precedence 1 64 256 100precedence 2 96 256 100precedence 3 128 256 100precedence 4 160 256 100precedence 5 192 256 100precedence 6 224 256 100precedence 7 256 256 100The following example enables WRED to use the DSCP value 9. The minimum threshold for the DSCP value 9 is 20 and the maximum threshold is 50. This configuration can be attached to other virtual circuits (VCs) as required.
Router(config)# random-detect-group sanjose dscp-basedRouter(cfg-red-grp)# dscp 9 20 50Router(config-subif-vc)# random-detect attach sanjoseRelated Commands
random-detect precedence
To configure Weighted Random Early Detection (WRED) and distributed WRED (DWRED) parameters for a particular IP Precedence, use the random-detect precedence command in interface configuration mode. To configure WRED parameters for a particular IP Precedence for a class policy in a policy map, use the random-detect precedence command in policy-map class configuration mode. To return the values to the default for the precedence, use the no form of this command.
random-detect precedence {precedence | rsvp} min-threshold max-threshold mark-prob-denominator
no random-detect precedence {precedence | rsvp} min-threshold max-threshold mark-prob-denominator
Syntax Description
precedence
IP Precedence number. The value range is from 0 to 7. For Cisco 7000 series routers with an RSP7000 interface processor and Cisco 7500 series routers with a VIP2-40 interface processor (VIP2-50 interface processor strongly recommended), the precedence value range is from 0 to 7 only; see Table 17 in the "Usage Guidelines" section of this command.
rsvp
Indicates Resource Reservation Protocol (RSVP) traffic.
min-threshold
Minimum threshold in number of packets. The value range of this argument is from 1 to 4096. When the average queue length reaches the minimum threshold, WRED randomly drops some packets with the specified IP Precedence.
max-threshold
Maximum threshold in number of packets. The value range of this argument is from the value of the min-threshold argument to 4096. When the average queue length exceeds the maximum threshold, WRED drops all packets with the specified IP Precedence.
mark-prob-denominator
Denominator for the fraction of packets dropped when the average queue depth is at the minimum threshold. For example, if the denominator is 512, 1 out of every 512 packets is dropped when the average queue is at the minimum threshold. The value range is from 1 to 65536. The default is 10; 1 out of every 10 packets is dropped at the minimum threshold.
Defaults
For all precedences, the mark-prob-denominator default is 10, and the max-threshold is based on the output buffering capacity and the transmission speed for the interface.
The default min-threshold depends on the precedence. The min-threshold for IP Precedence 0 corresponds to half of the max-threshold. The values for the remaining precedences fall between half the max-threshold and the max-threshold at evenly spaced intervals. See Table 17 in the "Usage Guidelines" section of this command for a list of the default minimum threshold values for each IP Precedence.
Command Modes
Interface configuration when used on an interface
Policy-map class configuration when used to specify class policy in a policy map
Command History
Usage Guidelines
WRED is a congestion avoidance mechanism that slows traffic by randomly dropping packets when congestion exists. DWRED is similar to WRED but uses the Versatile Interface Processor (VIP) instead of the Route Switch Processor (RSP).
When you configure the random-detect command on an interface, packets are given preferential treatment based on the IP Precedence of the packet. Use the random-detect precedence command to adjust the treatment for different precedences.
If you want WRED or DWRED to ignore the precedence when determining which packets to drop, enter this command with the same parameters for each precedence. Remember to use reasonable values for the minimum and maximum thresholds.
Note that if you use the random-detect precedence command to adjust the treatment for different precedences within class policy, you must ensure that WRED is not configured for the interface to which you attach that service policy.
Table 17 lists the default minimum threshold value for each IP Precedence.
Note
The DWRED feature is supported only on Cisco 7000 series routers with an RSP7000 card and Cisco 7500 series routers with a VIP2-40 or greater interface processor. A VIP2-50 interface processor is strongly recommended when the aggregate line rate of the port adapters on the VIP is greater than DS3. A VIP2-50 interface processor is required for OC-3 rates.
To use DWRED, distributed Cisco Express Forwarding (dCEF) switching must first be enabled on the interface. For more information on dCEF, refer to the Cisco IOS Switching Services Configuration Guide and the Cisco IOS Switching Services Command Reference.
Note
Examples
The following example enables WRED on the interface and specifies parameters for the different IP Precedences:
interface Hssi0/0/0description 45Mbps to R1ip address 10.200.14.250 255.255.255.252random-detectrandom-detect precedence 0 32 256 100random-detect precedence 1 64 256 100random-detect precedence 2 96 256 100random-detect precedence 3 120 256 100random-detect precedence 4 140 256 100random-detect precedence 5 170 256 100random-detect precedence 6 290 256 100random-detect precedence 7 210 256 100random-detect precedence rsvp 230 256 100The following example configures policy for a class called acl10 included in a policy map called policy10. Class acl101 has these characteristics: a minimum of 2000 kbps of bandwidth are expected to be delivered to this class in the event of congestion and a weight factor of 10 is used to calculate the average queue size. For congestion avoidance, WRED packet drop is used, not tail drop. IP Precedence is reset for levels 0 through 4.
policy-map policy10class acl10bandwidth 2000random-detectrandom-detect exponential-weighting-constant 10random-detect precedence 0 32 256 100random-detect precedence 1 64 256 100random-detect precedence 2 96 256 100random-detect precedence 3 120 256 100random-detect precedence 4 140 256 100Related Commands
rate-limit
To configure committed access rate (CAR) and distributed committed access rate (DCAR) policies, use the rate-limit command in interface configuration mode. To remove the rate limit from the configuration, use the no form of this command.
rate-limit {input | output} {bps | access-group acl-index | [rate-limit] rate-limit-acl-index] | dscp dscp-value | qos-group qos-group-number} burst-normal burst-max conform-action conform-action exceed-action exceed-action
no rate-limit {input | output} {bps | access-group acl-index | [rate-limit] rate-limit-acl-index] | dscp dscp-value | qos-group qos-group-number} burst-normal burst-max conform-action conform-action exceed-action exceed-action
Syntax Description
Defaults
CAR and DCAR are disabled.
Command Modes
Interface configuration
Command History
Usage Guidelines
Use this command to configure your CAR policy on an interface. To specify multiple policies, enter this command once for each policy.
CAR and DCAR can be configured on an interface or subinterface.
Policing Traffic with CAR
CAR embodies a rate-limiting feature for policing traffic. When policing traffic with CAR, Cisco recommends the following values for the normal and extended burst parameters:
normal burst = configured rate * (1 byte)/(8 bits) * 1.5 seconds
extended burst = 2 * normal burst
With the listed choices for parameters, extensive test results have shown CAR to achieve the configured rate. If the burst values are too low, then the achieved rate is often much lower than the configured rate.
For more information about using CAR to police traffic, see the "Policing with CAR" section of the "Policing and Shaping Overview" in the Cisco IOS Quality of Service Configuration Guide.
Examples
In the following example, the rate is limited by the application in question:
•
rate-limit input rate-limit access-group 101 20000000 3750000 7500000 conform-actionset-mpls-exp-transmit 5 exceed-action set-mpls-exp-transmit 0access-list 101 permit tcp any any eq www•
rate-limit input access-group 102 10000000 1875000 3750000conform-action set-mpls-exp-transmit 5 exceed-action dropaccess-list 102 permit tcp any any eq ftp•
rate-limit input 8000000 1500000 3000000 conform-action set-mpls-exp-transmit 5exceed-action dropNotice that two access lists are created to classify the web and FTP traffic so that they can be handled separately by the CAR feature.
Router(config)# interface Hssi0/0/0Router(config-if)# description 45Mbps to R2Router(config-if)# rate-limit input rate-limit access-group 101 20000000 3750000 7500000conform-action set-mpls-exp-transmit 5 exceed-action set-mpls-exp-transmit 0Router(config-if)# rate-limit input access-group 102 10000000 1875000 3750000conform-action set-mpls-exp-transmit 5 exceed-action dropRouter(config-if)# rate-limit input 8000000 1500000 3000000 conform-actionset-mpls-exp-transmit 5 exceed-action dropRouter(config-if)# ip address 200.200.14.250 255.255.255.252!Router(config-if)# access-list 101 permit tcp any any eq wwwRouter(config-if)# access-list 102 permit tcp any any eq ftpIn the following example, the MPLS experimental field is set, and the packet is transmitted:
Router(config)# interface FastEthernet1/1/0Router(config-if)# rate-limit input 8000 1500 3000 access-group conform-actionset mpls-exp-transmit 5 exceed-action set-mpls-exp-transmit 5In the following example, any packet with a DSCP of 1 can apply the rate limit:
Router(config)# interface serial6/1/0Router(config-if)# rate-limit output dscp 1 8000 1500 3000 conform-action transmit exceed-action dropRelated Commands
