Cisco SCE 2000 and SCE 1000 Software Configuration Guide, Release 4.1.x

Configuring the Line Interface

Introduction

This module describes how to configure the physical line interfaces (ports) as well how to configure those interfaces for tunneling, VLAN translation, TOS marking, and traffic rules.

•Line Interfaces

•Tunneling Protocols

•Configuring VLAN Translation

•Configuring Traffic Rules and Counters

•TOS Marking

•Counting the Dropped Packets

Line Interfaces

The Line Interfaces (Subscriber and Network) are used to connect the SCE platform to the network. See the description of network topologies in the Topology section of the relevant SCE platform installation guide (Cisco SCE 2000 Installation and Configuration Guide or Cisco SCE 1000 2xGBE Installation and Configuration Guide.)

The SCE 1000 2xGBE and the SCE 2000 4xGBE have Gigabit Ethernet line interfaces. You should configure auto-negotiate for these interfaces.

Flow Control and Bandwidth Considerations

Note By design, the SCE platform reacts to Ethernet flow control and does not activate it. Therefore, it is possible for a situation to arise in which flow control actually stalls the SCE platform by overflowing the SCE platform queues, thereby causing traffic to be dropped on the Rx interfaces. If this situation persists for more than five seconds, it may trigger the internal sanity checks mechanism within the SCE platform, which may in turn trigger a reload of the SCE platform in an attempt to recover

Note Using the SCE platform shapers ('Setting the Total Interface BW') with low values (on the order of tens or a few hundred Mbps) may result in system congestion and potential service loss. This is a known system limitation.

Configuring the Gigabit Ethernet Line Interfaces

Note The maximum packet size supported by the SCE 2000 and SCE 1000 platforms is 1600 bytes

How to Configure a Specific Gigabit Ethernet Line Interface

Step 1 At the SCE# prompt, type configure, and press Enter.

Enters Global Configuration mode.

Step 2 At the SCE(config)# prompt, type interface GigabitEthernet 0/portnumber, and press Enter, where portnumber is the number of the selected port (1-4).

Enters Interface Configuration mode for the selected GBE interface.

Step 3 At the SCE(config if)# prompt, type auto-negotiate, and press Enter,

Enables auto-negotiation.

Step 4 At the SCE(config if)# prompt, type exit and press Enter.

Exits to global configuration mode, from which you can access a different Gigabit Ethernet interface.

Step 5 Repeat steps 2 and 3 for the remaining Gigabit Ethernet interfaces.

Auto-negotiation must enabled separately and explicitly for each Gigabit Ethernet interface.

How to Configure a Range of Gigabit Ethernet Line Interfaces

Step 1 At the SCE# prompt, type configure and press Enter.

Enters Global Configuration mode.

Step 2 At the SCE(config)# prompt, type interface range GigabitEthernet 0/port range and press Enter.

Where port range is defined as follows:

•SCE 2000: any range of the interface numbers between 1 and 4

•SCE 1000: 1-2

Enters Interface Configuration mode for the selected range of GBE interfaces.

Step 3 At the SCE(config if range)# prompt, type auto-negotiate, and press Enter,

Enables autonegotiation for all the interfaces in the specified range.

Step 4 At the SCE(config if range)# prompt, type exit and press Enter.

Exits to global configuration mode, from which you can access a different Gigabit Ethernet interface.

Configuring a Range of Gigabit Ethernet Line Interfaces: Example

This example illustrates how to configure autonegotiation on ports 1 and 2 of an SCE platform.

SCE>configure 
SCE(config)interface range GigabitEthernet 0/1-2 
SCE(config if range)auto-negotiate 
SCE(config if range)exit 
SCE(config)

Tunneling Protocols

•Managed VPNs

•Selecting the Tunneling Mode

•Asymmetric L2 Support

•Displaying the Tunneling Configuration

Tunneling technology is used across various telecommunications segments to solve a wide variety of networking problems. The SCE platform is designed to recognize and process various tunneling protocols in several ways. The SCE platform is able to either ignore the tunneling protocols ("skip" the header) or treat the tunneling information as subscriber information ("classify"). A special case of classification by tunneling information is MPLS/VPN with private IP support

Table 6-1 shows the support for the various tunneling protocols (the default behavior for each protocol is in bold).

When the tunneling information is ignored, the subscriber identification is the subscriber IP of the IP packet carried inside the tunnel.

Asymmetric Tunneling

Some tunneling modes are symmetric and some are asymmetric (see Table 6-1). Any time that one of the asymmetric tunneling modes is enabled, the entire system is automatically set to asymmetric flow open mode. In this mode, flows are opened earlier than in symmetric flow open mode, and the first packet of each direction of the flow (upstream and downstream) reaches the software. This is required to support redirect and block operations over asymmetric layer 2 protocols. However, it also has some performance impact, so that a certain performance degradation should be expected in any asymmetric mode.

It is also possible to explicitly configure the system to treat all flows as having asymmetric layer 2 characteristics (including Ethernet, VLAN, MPLS, and L2TP), for the purpose of packet injection (such as block flow and redirect flow operations).

To view the effective flow open mode, use the show interface linecard 0 flow-open-mode command.

Note For directions on how to configure the asymmetric tunneling option, see Asymmetric L2 Support

L2TP

L2TP is an IP-based tunneling protocol, therefore the system must be specifically configured to recognize the L2TP flows, given the UDP port used for L2TP. The SCE platform can then skip the external IP, UDP, and L2TP headers, reaching the internal IP, which is the actual subscriber traffic. If L2TP is not configured, the system treats the external IP header as the subscriber traffic, thus all the flows in the tunnel are seen as a single flow.

VLAN

A single VLAN tag is supported per packet (no QinQ support).

Subscriber classification by VLAN tag is supported only in symmetric VLAN environments - i.e. where the upstream and downstream tags of a flow are identical.

MPLS

MPLS/VPN-based subscribers with private IP support are a special case of classification by tunneling information.

MPLS labels are supported up to a maximum of 15 labels per packet.

For more information regarding MPLS/VPN-based subscribers, see the MPLS/VPN Support

Managed VPNs

•Private IP Addresses

•Capacity

•Limitations for VPN mode

A managed VPN is a named entity, introduced in the same way that a subscriber is introduced, and containing VPN mappings.

A managed VPN may contain several MPLS mappings or a single VLAN mapping. A VPN-based subscriber contains a set of mappings of the form: IP@VpnName, where IP can be either a single IP address or a range of addresses.

Managed VPN entities can be configured only via the SM. The SCE platform CLI can be used to view VPN-related information, but not to configure the VPNs.

Private IP Addresses

Private IP addresses are supported only in the following modes, as these modes provide information regarding the higher-level entity (VLAN or VPN) to which the IP addresses of the flow belong:

•MPLS VPN auto-learn

•VLAN symmetric classify

Capacity

The system supports:

•2015 VPNs

•80,000 IP mappings over VPNs

Limitations for VPN mode

Mutually exclusive system modes

When the system is working in VPN mode, the following modes are not supported:

•TCP Bypass-establishment

•DDoS

•Value Added Services (VAS) mode

Number of MPLS labels

•The choice of the unique VPN site must be based on the BGP label only. The BGP label must be the innermost label.

•The MPLS/VPN solution supports various combinations of labels.

•The system does not support VPNs for which other MPLS-related features, such as MPLS-TE or MPLS-FRR, are enabled

Subscriber-related limitations

•The SM must be configured to operate in Push mode.

•Introduced subscriber aging is not supported when using VPN-based subscribers

Topology-related limitations

•An asymmetrical routing topology in which the traffic may be unidirectional, is not supported, since the identification of the VPN relies on the bidirectional nature of the traffic for various mechanisms.

TCP-related requirements

•Number of Upstream TCP Flows - There must be enough TCP flows opening from the subscriber side on each PE-PE route in each period of time. The higher the rate of TCP flows from the subscriber side, the higher the accuracy of the mechanism can be.

VPN configuration requirements

•Two VPN sites must be aggregated into one VPN if the following conditions are both true:

–They are both connected to the same SCE platform †

–They both communicate with a common remote site using the same upstream labels and P router.

•In MPLS-based VPNs (MPLS auto-learn mode), a subscriber MAY NOT have IP mappings over more than one VPN

•In VLAN-based VPNs (VLAN symmetric classify mode), a subscriber may have IP mappings over more than one VPN, but only if the IP mappings are the full range of the VPN (0.0.0.0/0). (This option is provided for backwards compatibility, supporting legacy multi-VLAN subscribers.)

Selecting the Tunneling Mode

Use these commands to configure tunneling:

•ip-tunnel

•vlan

•mpls

•L2TP identify-by

Configuring L2TP IP Tunnels

By default, IP tunnel recognition is disabled.

The IP tunneling mode is mutually exclusive with VPN-based classification.

How to Enable IP Tunnels

Use this command to configure recognition of L2TP tunnels and skipping into the internal IP packet.

Step 1 From the SCE(config if)# prompt, type ip-tunnel L2TP skip and press Enter.

How to Disable IP Tunnels

Step 1 From the SCE(config if)# prompt, type no ip-tunnel and press Enter.

IPinIP Tunneling

IPinIP is an IP-based tunneling protocol; therefore the system must be specifically configured to recognize the flows inside the tunnel. The SCE platform will then skip the external IP header, reaching the internal IP, which is the actual subscriber traffic. When IPinIP skip is disabled, the system treats the external IP header as the subscriber traffic, resulting in all IPinIP traffic being reported as generic IP.

Guidelines for configuring IPinIP tunnels:

•IPinIP and other tunnels: IPinIP is supported simultaneously with plain IP traffic and any other tunneling protocol supported by the SCE platform.

•Overlapping IP addresses: There is no support for overlapping IP addresses within different IPinIP tunnels.

•DSCP marking: For IPinIP traffic, DSCP marking can be done on either the external or the internal IP header exclusively.

Fragmentation

Fragmentation should be avoided whenever possible. If it is not possible to avoid fragmentation, it is recommended to opt for internal fragmentation. If that is also not possible, the SCE platform can be operated under conditions of external fragmentation.

How to Enable IPinIP Tunneling

By default, IP tunnel recognition is disabled. Use this command to configure recognition of IPinIP tunnels and skipping into the internal IP packet.

Step 1 Shut down the linecard. (This is a root level command.)

From the SCE(config if)#> prompt, type shutdown and press Enter.

Step 2 Enable IPinIP tunneling.

From the SCE(config if)# prompt, type ip-tunnel IPinIP skip and press Enter.

Step 3 Restart the linecard.

From the SCE(config if)#> prompt, type no shutdown and press Enter.

How to Disable IPinIP Tunneling

Step 1 Shut down the linecard. (This is a root level command.)

From the SCE(config if)#> prompt, type shutdown and press Enter.

Step 2 Disable IPinIP tunneling.

From the SCE(config if)# prompt, type no ip-tunnel IPinIP skip and press Enter.

Step 3 Restart the linecard.

From the SCE(config if)#> prompt, type no shutdown and press Enter.

Configuring DSCP Marking for IPinIP Tunnels

DSCP marking modifies the DSCP bits of the IPv4 header. In IPinIP tunnels there are at least two IP headers. By default, DSCP marking is performed only on the external IP header (refer Figure 6-1).

Figure 6-1 DSCP Marking for IPinIP Tunnels

Note DSCP marking should be enabled and configured through SCA BB console. See the "How to Manage DSCP ToS Marker Values" section in the "Using the Service Configuration Editor: Traffic Control" chapter of the Cisco Service Control Application for Broadband User Guide for further information.

Use this command to configure the SCE platform to mark the DSCP bits of the internal IP header. This command takes effect only when IPinIP skip is enabled.

Step 1 Shut down the linecard. (This is a root level command.)

From the SCE(config if)#> prompt, type shutdown and press Enter.

Step 2 Configure the DSCP marking (.

From the SCE(config if)#> prompt, type ip-tunnel IPinIP DSCP-marking-skip and press Enter.

Enables DSCP marking on the internal IP header of IPinIP traffic.

Step 3 Restart the linecard.

From the SCE(config if)#> prompt, type no shutdown and press Enter.

To perform DSCP marking on the external IP header, use the following command:

Step 1 Shut down the linecard. (This is a root level command.)

From the SCE(config if)#> prompt, type shutdown and press Enter.

Step 2 Configure the DSCP marking.

From the SCE(config if)#> prompt, type no ip-tunnel IPinIP DSCP-marking-skip and press Enter.

Enables DSCP marking on the external IP header of IPinIP traffic.

Step 3 Restart the linecard.

From the SCE(config if)#> prompt, type no shutdown and press Enter.

Configuring the VLAN Environment

Note The SCE 1000 and SCE 2000 platforms support a maximum of 2048 VLAN tags.

Use this command to configure the VLAN environment.

•Options

•Configuring the VLAN Environment: Example

Options

There are three options:

•symmetric classify

•symmetric skip (default)

•a-symmetric skip

Symmetric environment refers to an environment in which the same VLAN tags are used for carrying a transaction in the upstream and downstream directions

Setting the mode to classify means that VPN and flow classification will use the VLAN tag. Using VLAN classification is mutually exclusive with other tunnel-based classification or IP tunnels.

An a-symmetric environment is an environment in which the VLAN tags might not be the same in the upstream and downstream directions of the same flow.

The SCE platform is configured by default to work in symmetric environments. A specific command should be used to allow correct operation of the SCE platform in asymmetric environments and instruct it to take into consideration that the upstream and downstream of each flow has potentially different VLAN tags.

Note Using the a-symmetric skip value incurs a performance penalty.

Step 1 From the SCE(config if)# prompt, type vlan {symmetric classify | symmetric skip |a-symmetric skip} and press Enter.

Specify the desired VLAN mode.

Configuring the VLAN Environment: Example

The following example selects VLAN-based classification.

SCE(config if)#vlan symmetric classify

Configuring the MPLS Environment

Use this command to set the MPLS environment.

Note For more information regarding the Cisco Service Control MPLS/VPN solution, see the "MPLS/VPN Support " chapter.

•Options

•Configuring the MPLS Environment: Example

Options

The following options are available:

•traffic-engineering skip (default)—Use when all IP addresses are unique and MPLS labels are not mandatory for routing.

•VPN skip —Use when all IP addresses are unique, but MPLS labels are mandatory for routing.

•VPN auto-learn —Use in an MPLS/VPN environment where auto-learning is required due to the existence of private IP addresses and/or VPN based subscribers.

When this option is configured, both ip-tunnel and VLAN must be set to their default values.

Use the VPN keyword when the labels are mandatory in the traffic, otherwise use traffic-engineering (default).

Note that using the VPN value incurs a performance penalty.

The MPLS VPN auto-learn option is required in an MPLS/VPN environment.

Step 1 From the SCE(config if)# prompt, type mpls {traffic-engineering skip|vpn skip|vpn auto-learn} and press Enter.

Specify the desired MPLS mode.

Configuring the MPLS Environment: Example

The following example selects the MPLS/VPN tunnel environment.

SCE(config if)#mpls vpn auto-learn

About Changing the Managed VPN Modes

Managed VPNs can only exist in either VLAN symmetric classify or MPLS VPN auto-learn mode, but these two modes cannot be enabled simultaneously. When changing from one of these VPN-related modes to another, keep the following guidelines in mind:

•All VPN-based subscribers must be cleared to change the tunneling mode. If the connection with the SM is down, use the no subscriber all with-vpn-mappings CLI command (see About VPN-Based Subscribers)

•All VPN mappings must also be removed. This can only be done via the SM CLU (which means that the connection with the SM must be up). (See How to Manage VPN Mappings

How to Restore the Default VLAN or MPLS Environment

Use this command to restore the default VLAN or MPLS configuration.

It is not usually necessary to explicitly restore the default environment, as this is done automatically when executing a VLAN or MPLS command. When such an automatic reset to default occurs, a warning message appears similar to the following:

Warning Disabled previously configured IP-tunnel support or tunneling classification mode

Step 1 From the SCE(config if)# prompt, type default {mpls | vlan} and press Enter.

Configuring the L2TP Environment

Use this command to set the port number that the LNS and LAC use for L2TP tunnels.

Note Non-first fragments of pure IP traffic (not tunneled) are not handled correctly when the system is in L2TP skip mode. Incorrect UDP/TCP ports are assumed, and the fragment is mapped to the wrong flow.

•External Fragmentation in the L2TP Environment

•Options

External Fragmentation in the L2TP Environment

If external fragmentation exists in the L2TP environment, it is required to configure a ' quick-forwarding-ignore ' Traffic Rule (see Configuring Traffic Rules and Counters) that bypasses all IP traffic targeted to either the LNS or LAC IP address. This will make sure that any packets not having the L2TP port indication (i.e. non-first fragments) will not require handling by the traffic processors.

In addition, in order to prevent reordering of L2TP tunneled fragments, it is advised to define a ' quick-forwarding ' traffic-rule for all the L2TP traffic. This can be done based on the IP ranges in use by the internal IPs in the tunnel (as allocated by the LNS), or simply for all of the traffic passing through the SCE platform.

Note By enabling quick-forwarding, the Cisco SCE can only perform traffic monitoring for externally-fragmented L2TP traffic. It cannot perform flow redirection, flow blocking, or rate-limiting.

Options

The following option is available:

•portnumber —The port number that the LNS and LAC use for L2TP tunnels.

Default port# = 1701

Step 1 From the SCE(config if)# prompt, type L2TP identify-by port-number portnumber and press Enter.

Enables privileged EXEC mode.

Asymmetric L2 Support

You should enable asymmetric layer 2 support in cases where the following conditions apply for any flows:

•Each direction of the flow has a different pair of MAC addresses.

•The routers do not accept packets with the MAC address of the other link.

Note 'Asymmetric routing topology' support and 'asymmetric tunneling support' are two separate features. Asymmetric routing topology refers to topologies where the SCE platform might see some flows only in one direction (upstream/downstream). Asymmetric tunneling support (asymmetric L2 support) refers to the ability to support topologies where the SCE platform sees both directions of all flows, but some of the flows may have different layer 2 characteristics (like MAC addresses, VLAN tags, MPLS labels and L2TP headers), which the SCE platform must specifically take into account when injecting packets into the traffic (such as in block and redirect operations). Note as well, that in order to support asymmetric layer 2, the SCE platform switches to asymmetric flow open mode, which incurs a certain performance penalty. This is NOT the case for asymmetric routing topology.

Step 1 From the SCE(config if)# prompt, type asymmetric-L2-support and press Enter.

Displaying the Tunneling Configuration

Step 1 From the SCE# prompt, type show interface linecard 0 MPLS|VLAN|L2TP|IP-tunnel and press Enter.

Displays the current configuration for the specified tunnel option.

How to Display the IPinIP Configuration

Step 1 From the SCE# prompt, type show interface linecard 0 ip-tunnel IPinIP and press Enter.

Displays the current configuration for the specified tunnel option.

How to Display the Logged-in VPNs

Options

The following options are available

•vpn-name —The name of a specific currently logged-in VPN for which to display details.

•all-names —Use this keyword to display all the VPN names that are currently logged into the system.

Step 1 From the SCE> prompt, type show interface linecard 0 VPN {name vpn-name | all-names} and press Enter.

How to Display the Asymmetric L2 Support Mode

Step 1 From the SCE# prompt, type show interface linecard 0 asymmetric-L2-support and press Enter.

Configuring VLAN Translation

•VLAN Translation Features and Limitations

•Setting the VLAN Translation Constant

•How to Disable VLAN Translation

•How to Monitor VLAN Translation

Some topologies require the SCE platform to be able to translate between different VLAN tags.

Figure 6-2 illustrates an example of such a system, in which one router acts as a dispatcher, forwarding traffic and performing load balancing between two SCE 2000 platforms.

Figure 6-2 VLAN Translation

In this example, traffic enters the router via the access ports; it is forwarded to an EtherChannel, which is configured as a trunk, and enters the SCE 2000 platforms.

As can be seen from this drawing, the subscriber side VLAN tags must be different from those on the network side, or the router will simply forward the traffic to the opposite port. This can be supported very simply by having the SCE platform replace the VLAN tags according to a preset configuration.

VLAN Translation Features and Limitations

Features

•Configuration of an increment or decrement constant.

•Configuration of the constant is global for the line card.

•The configured operation (either increment or decrement) is applied to the network side.

•The subscriber side automatically performs the opposite operation. That is, if the VLAN is incremented by X on the network side, it is decremented by X on the subscriber side.

•VLAN tagged packets are changed (incremented or decremented) before transmission.

•Non-tagged packet are not changed.

•This feature allows seamless processing with non-VLAN traffic.

Limitations

•LIC Bypass not supported - Translation is done in the transmission. Therefore, in LIC bypass, where there is no transmission, there is also no translation.

This means that in general, installations using the VLAN translation feature should rely on cutoff on failure and at upgrade (use redundant SCE platform).

•STP hazard - VLAN translation may interfere with Spanning Tree Protocol. This should be taken in consideration when deploying the solution.

•The maximum offset that can be configured is 2047. Note that there is no protection for wraparound.

Setting the VLAN Translation Constant

Use this command to define the VLAN translation constant. Make sure that the same VLAN translation constant is configured for all SCE platforms in the system.

•Options

•Setting the VLAN Translation Constant: Example

Options

The following options are available:

•increment | decrement —Keywords indicating whether to increment or decrement the VLAN by the specified value.

•value — Integer value by which the VLAN is to incremented or decremented.

The configured translation is applied to the network port side. The reverse operation is performed at the subscriber side.

For example, if "increment 5" is defined, at the network port the VLAN is incremented by 5, and at the subscriber port the VLAN is decremented by 5.

In this case, the network side VLAN tags might be 105, 205, 305, and the subscriber side the VLAN tags would then be 100, 200, 300.

Default = 0

Maximum = 2047 (Note that there is no protection for wraparound of the VLAN value.)

Step 1 From the SCE(config if)# prompt, type vlan translation increment|decrement value value and press Enter.

Sets the VLAN translation constant.

Setting the VLAN Translation Constant: Example

The following example sets the translation constant to 10, decremented at the network side.

SCE(config if)#vlan translation decrement value 10

How to Disable VLAN Translation

Step 1 From the SCE# prompt, type no vlan translation and press Enter

Disables VLAN translation.

How to Monitor VLAN Translation

Step 1 From the SCE# prompt, type show interface linecard 0 vlan translation and press Enter.

Displays current VLAN translation configuration.

Configuring Traffic Rules and Counters

•Information About Traffic Rules and Counters

•Configuring Traffic Counters

•Configuring Traffic Rules

•Managing Traffic Rules and Counters

Information About Traffic Rules and Counters

•What are Traffic Rules and Counters?

•Traffic Rules

•Traffic Counters

What are Traffic Rules and Counters?

Traffic rules and counters may be configured by the user. This functionality enables the user to define specific operations on the traffic flowing through the SCE Platform, such as blocking or ignoring certain flows or counting certain packets. The configuration of traffic rules and counters is independent of the application loaded by the SCE platform, and thus is preserved when the application being run by the SCE platform is changed.

Possible uses for traffic rules and counters include:

•Enabling the user to count packets according to various criteria. Since the traffic counters are readable via the SCE SNMP MIB, these might be used to monitor up to 32 types of packets, according to the requirements of the installation.

•Ignoring certain types of flows. When a traffic rules specifies an "ignore" action, packets matching the rule criteria will not open a new flow, but will pass through the SCE platform without being processed. This is useful when a particular type of traffic should be ignored by the SCE platform.

Possible examples include ignoring traffic from a certain IP range known to require no service, or traffic from a certain protocol.

•Blocking certain types of flows. When a traffic rules specifies a "block" action, packets matching the rule criteria (and not belonging to an existing flow) will be dropped and not passed to the other interface. This is useful when a particular type of traffic should be blocked by the SCE platform.

Possible examples include performing ingress source address filtering (dropping packets originating from a subscriber port whose IP address does not belong to any defined subscriber-side subnet), or blocking specific ports.

It should be noted that using traffic rules and counters does not affect performance. It is possible to define the maximum number of both traffic rules and counters without causing any degradation in the SCE platform performance.

Traffic Rules

A traffic rule specifies that a defined action should be taken on packets processed by the SCE Platform that meet certain criteria. The maximum number of rules is 128, which includes not only traffic rules configured via the SCE platform CLI, but also any additional rules configured by external management systems, such as SCA BB. Each rule is given a name when it is defined, which is then used when referring to the rule.

Packets are selected according to user-defined criteria, which may be any combination of the following:

•IP address — A single address or a subnet range can be specified for each of the line ports (Subscriber / Network).

•Protocol — TCP/UDP/ICMP/IGRP/EIGRP/IS-IS/OSPF/Other

•TCP/UDP Ports — A single port or a port range can be specified for each of the line ports (Subscriber / Network). Valid for the TCP/UDP protocols only.

•Direction (Upstream/Downstream) (TCP only).

The possible actions are:

•Count the packet by a specific traffic counter

•Block the packet (do not pass it to the other side)

•Ignore the packet (do not provide service for this packet — No bandwidth metering, transaction reporting etc. is done)

•Quick-forward the packet with service — forward delay-sensitive packets through the fast path while maintaining serviceability for these packets

•Quick-forward the packet with no service — forward delay-sensitive packets through the fast path with no service provided for these packets.

•flow-capture — Capture the flow configured by this rule. No service to this flow

Block and Ignore actions affect only packets that are not part of an existing flow.

Note that Block and Ignore are mutually exclusive. However, blocked or ignored packets can also be counted.

It is possible for a single packet to match more that one rule (The simplest way to cause this is to configure two identical rules with different names). When this happens, the system operates as follows:

•Any counter counts a specific packet only once. This means that:

–If two rules specify that the packet should be counted by the same counter, it is counted only once.

–If two rules specify that the packet should be counted by different counters, it is counted twice, once by each counter.

•Block takes precedence over Ignore — If one rule specifies Block, and another rule specifies Ignore, the packet is blocked.

Traffic Counters

Traffic counters count the traffic as specified by the traffic rules. The maximum number of counters is 32. Each counter is given a name when it is defined, which is then used when referring to the counter.

A traffic counter can be configured in one of two ways:

•Count packets — the counter is incremented by 1 for each packet it counts.

•Count bytes — the counter is incremented by the number of bytes in the packet for each packet it counts.

Configuring Traffic Counters

A traffic counter must be created before it can be referenced in a traffic rule. Use the following commands to create and delete traffic counters.

•How to Create a Traffic Counter

•How to Delete a Traffic Counter

•How to Delete all Existing Traffic Counters

How to Create a Traffic Counter

Options

The following options are available:

•name —The name of the counter

•Count packets — the counter is incremented by 1 for each packet it counts.

•Count bytes — the counter is incremented by the number of bytes in the packet for each packet it counts.

Step 1 From the SCE(config if)# prompt, type traffic-counter name name count-bytes|count-packets and press Enter.

Adds a traffic counter with the specified name and counting mode.

How to Delete a Traffic Counter

Step 1 From the SCE(config if)# prompt, type no traffic-counter name name and press Enter.

Note that a traffic counter cannot be deleted if it is used by any existing traffic rule.

How to Delete all Existing Traffic Counters

Step 1 From the SCE(config if)# prompt, type no traffic-counter all and press Enter.

Removes all traffic counters.

Note that a traffic counter cannot be deleted if it is used by any existing traffic rule.

Configuring Traffic Rules

Use the following commands to create and delete traffic rules.

•How to Create a Traffic Rule

•How to Delete a Traffic Rule

•How to Delete all Traffic Rules

•How to Delete All Flow Control Traffic Rules

How to Create a Traffic Rule

•Options

•Configuring Traffic Rules: Examples

Options

The following options are available:

IP specification:

all|([all-but] (ip-address|ip-range))

•ip-address is a single IP address in dotted-decimal notation, such as 10.1.2.3

•ip-range is an IP subnet range, in the dotted-decimal notation followed by the number of significant bits, such as 10.1.2.0/24.

•Use the all-but keyword to exclude the specified IP address or range of IP addresses

protocol:

Any one of the following protocols:

TCP/UDP/ICMP/IGRP/EIGRP/IS-IS/OSPF/Other

port specification:

all|([all-but] (port#|port-range)

•Specify the ports only if the protocol is either TCP or UDP.

•Specify the port or port range for both the subscriber-side and the network-side.

•Specify a range of ports using the form MinPort:MaxPort.

•Use the all-but keyword to exclude the specified port or range of ports.

tunnel id specification:

all|([all-but] tunnel id)

•tunnel id is an 8-bit Hex value range, in the format '(HEX) Tunnel-id ' or '(HEX) MinTunnelId :(HEX) MaxTunnelId ', which reflects the lower eight bits of the VLAN tag.

•Tunnel-ID-based rules can only be used in " VLAN symmetric classify " mode see, and only when tunnel id mode is enabled.

Use the traffic-rule tunnel-id-mode command.

Note that the VLAN tag itself is a 12-bit value, and therefore aliasing of the lower 8 bits can occur, depending on the VLAN tags used.

direction:

Any of the following:

upstream/downstream/both

traffic-counter:

Either of the following:

•name <name of an existing traffic counter>— Packets meeting the criteria of the rule are to be counted in the specified counter. If a counter name is defined, the "count" action is also defined implicitly. The keyword name must appear as well as the actual name of the counter.

•none — If none is specified, then an action must be explicitly defined via the action option.

action: (not required if the action is count only)

One of the following:

•block — Block the specified traffic

•ignore — Bypass the specified traffic; traffic receives no service

•quick-forwarding — Forward delay-sensitive packets through the fast path while maintaining serviceability for these packets

•quick-forwarding-ignore — Forward delay-sensitive packets through the fast path with no service provided for these packets.

•flow-capture — Capture the flow configured by this rule. No service to this flow

Step 1 From the SCE(config if)# prompt, type traffic-rule name name IP-addresses (all|(subscriber-side <IP specification> network-side <IP specification>)) protocol protocol [ports subscriber-side <port specification> network-side <port specification>] [tunnel-id <tunnel-id specification>] direction direction traffic-counter <traffic-counter>[action action ]

Enables privileged EXEC mode.

Configuring Traffic Rules: Examples

•Example 1

•Example 2

•Example 3

•Example 4

Example 1

This example creates the following traffic rule:

•Name = rule1

•IP addresses: subscriber side = all IP addresses, network side = 10.10.10.10 only

•Protocol = other

•Direction = both

•Traffic counter = counter1

•The only action performed will be counting

SCE(config if)# traffic-rule rule1 IP-addresses subscriber-side all network-side 
10.10.10.10 protocol other direction both traffic-counter name counter1

Example 2

This example creates the following traffic rule:

•Name = rule2

•IP addresses: subscriber side = all IP addresses, network side = all IP addresses EXCEPT the subnet 10.10.10.0/24

•Protocol = TCP

•Ports: subscriber-side = 100-200, network-side = all

•Tunnel id = all

•Direction = downstream

•Traffic counter = counter2

•Action = Block

•The actions performed will be counting and blocking

The first command enables tunnel id mode.

SCE(config if)#traffic-rule tunnel-id-mode  
SCE(config if)# traffic-rule rule2 IP-addresses subscriber-side all network-side all-but 
10.10.10.0/24 protocol tcp ports subscriber-side 100:200 network-side all tunnel-id all 
direction downstream traffic-counter name counter2 action block

Example 3

This example creates the following traffic rule:

•Name = rule3

•IP addresses: all

•Protocol = IS-IS

•Direction = upstream

•Traffic counter = none

•Action = ignore (required since traffic-counter = none)

•The only action performed will be Ignore.

SCE(config if)# traffic-rule rule3 IP-addresses all protocol IS-IS direction upstream 
traffic-counter none action ignore

Example 4

The following example illustrates how to configure a traffic rule that will be used as a recording rule using the flow-capture option. All flows that match this rule will be recorded when the flow capture process is in operation.

1. Name = FlowCaptureRule

2. IP addresses: subscriber side = all IP addresses, network side = all IP addresses

3. Direction = both

4. Protocol = 250

5. Traffic counter name = counter2

6. Action = flow-capture

7. The actions performed will be counting and flow capture.

SCE>enable 10 
Password:<cisco> 
SCE#configure 
SCE(config)#interface linecard 0 
SCE(config if)#traffic-rule name FlowCaptureRule ip-addresses subscriber-side all 
network-side all protocol 250 direction both traffic-counter name counter2 action 
flow-capture  
SCE(config if)#

How to Delete a Traffic Rule

Step 1 From the SCE(config if)# prompt, type no traffic-rule name name and press Enter.

Removes the specified traffic rule.

How to Delete all Traffic Rules

Step 1 From the SCE(config if)# prompt, type no traffic-rule all and press Enter.

Removes all existing traffic rules.

How to Delete All Flow Control Traffic Rules

Step 1 From the SCE(config if)# prompt, type no traffic-rule capture and press Enter.

Removes all flow capture traffic rules.

Managing Traffic Rules and Counters

Use these commands to display existing traffic rule configuration, as well as traffic counter configuration (packets/bytes and the name of the rule using the counter) and traffic counter value.

You can also reset a specific counter or all counters.

•How to View a Specified Traffic Rule

•How to View all Traffic Rules

•How to View a Specified Traffic Counter

•How to View all Traffic Counters

•How to Reset a Specified Traffic Counter

•How to Reset all Traffic Counters

How to View a Specified Traffic Rule

Step 1 From the SCE# prompt, type show interface linecard 0 traffic-rule name rule-name and press Enter.

Displays the configuration of the specified traffic rule.

How to View all Traffic Rules

Step 1 From the SCE# prompt, type show interface linecard 0 traffic-rule all and press Enter.

Displays the configuration of all existing traffic rules.

How to View a Specified Traffic Counter

Step 1 From the SCE# prompt, type show interface linecard 0 traffic-counter name counter-name and press Enter.

Displays the value of the specified counter and lists the traffic rules that use it.

Viewing a Traffic Counter: Example

The following example displays information for the traffic counter "cnt".

SCE# show interface linecard 0 traffic-counter name cnt 
Counter 'cnt' value: 0 packets. Rules using it: None.

How to View all Traffic Counters

Step 1 From the SCE# prompt, type show interface linecard 0 traffic-counter all and press Enter.

Displays the value of the each counter and lists the traffic rules that use it.

Viewing the Traffic Counters: Example

The following example displays information for all existing traffic counters.

SCE# show interface linecard 0 traffic-counter all 
Counter 'cnt' value: 0 packets. Rules using it: None. 
Counter 'cnt2' value: 0 packets. Rules using it: Rule2. 
2 counters listed out of 32 available.

How to Reset a Specified Traffic Counter

Step 1 From the SCE# prompt, type clear interface linecard 0 traffic-counter name counter-name and press Enter.

Displays the value of the specified counter and lists the traffic rules that use it.

How to Reset all Traffic Counters

Step 1 From the SCE# prompt, type clear interface linecard 0 traffic-counter all and press Enter.

Displays the value of the each counter and lists the traffic rules that use it.

TOS Marking

TOS marking is used in IP networks as a means to signal the priority of a flow between network elements. The Cisco Service Control solution supports the TOS classification on a per-service, per-package level via the SCA BB application. The SCE platform TOS marking feature enables marking the TOS field in the IP header of each packet according to the policy configured via the SCA BB console. The actual TOS value set in the IP header is determined according to the value defined in a configurable TOS translation table.

TOS marking configuration is performed via the SCA BB console, The SCE platform CLI allows you to view the state of TOS marking (enabled or disabled) for each interface and to display the TOS translation table.

For information on configuring TOS marking, see the Cisco Service Control Application for Broadband User Guide.

Note TOS marking in release 3.1.5 is not backwards compatible with any previous SCOS releases.

How to Display the TOS Marking Configuration

Use this command to display the state of TOS marking (enabled or disabled) per interface and the TOS translation table.

Step 1 From the SCE> prompt, type show interface linecard 0 ToS-marking and press Enter.

Counting the Dropped Packets

By default, the SCE platform hardware drops red packets (packets that are marked to be dropped due to BW control criteria). However, this presents a problem for the user who needs to know the number of dropped packets per service. To be able to count dropped packets per service, the traffic processor must see all dropped packets for all flows. However, if the hardware is dropping red packets, the traffic processor will not be able to count all dropped packets and the user will not get proper values on the relevant MIB counters (tpTotalNumWredDiscardedPackets).

Note The MIB object tpTotalNumWredDiscardedPackets counts dropped packets. The value in this counter is absolute only when hardware packet drop is disabled (not the default mode). When hardware packet drop is enabled (default mode), this MIB counter provides only a relative value indicating the trend of the number of packet drops, with a factor of approximately 1:6.

The user can disable the drop-red-packets-by-hardware mode. This allows the application to access existing per-flow counters. The application can then retrieve the number of dropped packets for every flow and provide the user with better visibility into the exact number of dropped packets and their distribution.

Note that counting all dropped packets has a considerable effect on system performance, and therefore, by default, the drop-red-packets-by-hardware mode is enabled.

Configuring Hardware Packet Drop

By default hardware packet drop is enabled.

Note Disabling this feature may have both delay and performance implications.

How to Disable Hardware Packet Drop

Use this command to disable the drop-red-packets-by-hardware mode, enabling the software to count all dropped packets.

Step 1 From the SCE(config if)# prompt, type no accelerate-packet-drops and press Enter.

How to Enable Hardware Packet Drop

Use this command to enable the drop-red-packets-by-hardware mode.

Step 1 From the SCE(config if)# prompt, type accelerate-packet-drops and press Enter.