Configuring IS-IS Interface-Specific Parameters

You can change IS-IS interface–specific parameters; most can be configured independently of other attached routers. You are not required to alter any interface parameters; however, some parameters must be consistent across all routers in your network. If you change certain values from the defaults, you must configure them on multiple interfaces and routers.

In the following command guidelines, many parameters are preset to a default value. If that parameter has been modified from its default, use the no version of the command to restore its default value.

Configuring Authentication

You can set a password to authenticate IS-IS hello packets, and you can configure HMAC MD5 authentication for IS-IS interfaces.

isis authentication-key

isis message-digest-key

Configuring Link-State Metrics

You can configure the routing metric (cost) for an IS-IS interface. Routes with lower total path metrics are preferred over those with higher path metrics.

isis metric

Configuring a Reference Bandwidth to Set a Default Metric

By default, all IS-IS interfaces without a configured metric have the same routing metric, 10, However, when you configure a reference bandwidth for IS-IS, the default metric is calculated differently for each IS-IS interface. The default routing metric in this case is the reference bandwidth divided by the bandwidth of the particular interface.

For example, if you set the IS-IS reference bandwidth to 50,000,000, the default metric for a 10-Mbps interface is calculated as 5. Interfaces with lower bandwidths have higher default metrics than this interface. Similarly, links with higher bandwidths have lower default metrics than this interface.


Setting the CSNP Interval

You can set the advertised complete sequence number PDU (CSNP) interval for an IS-IS interface.

isis csnp-interval

Configuring Hello Packet Parameters

You can set the hello interval and the hello multiplier for IS-IS hello packets.

isis hello-interval

isis hello-multiplier

Padding IS-IS Hello Packets

You can use the isis hello padding command to configure IS-IS hello packet padding. Padding the hello packets promotes early error detection due to transmission problems with large frames or due to mismatched MTUs on adjacent interfaces.

When disabled (default), IS-IS hello packets are padded to the full MTU size until an adjacency is formed with the adjacent interface. After the adjacency is formed, the hello packets are no longer padded. When enabled, IS-IS hello packets are always padded.

isis hello padding

Configuring LSP Parameters

You can configure the transmission interval, retransmission interval, and retransmission throttle interval for LSPs on an interface-specific basis.

isis lsp-interval

isis retransmit-interval

isis retransmit-throttle-interval

Setting the Designated Router Priority

You can set the priority for the designated IS-IS router that you have elected to use.

isis priority

Configuring Passive Interfaces

You can configure an IS-IS passive interface. A passive interface only advertises its IP address in its LSPs; it does not send or receive IS-IS packets.

Optionally, you can set a route tag for an IS-IS passive interface by including the tag keyword and a numeric tag value in the passive-interface command.

Passive interfaces have a metric of zero by default. You can set a different metric for a particular passive interface by specifying the value along with the metric keyword. A global default metric set with the metric command does not affect any passive interface. Similarly, configuring a reference bandwidth for IS-IS has no effect on passive interfaces. Metrics specified for a passive interface apply to both level 1 and level 2 interfaces unless you restrict the metric to a single level.

If you configure an interface for IP processing without any explicit IP address assigned to it as an unnumbered interface and if you also define it to be a loopback interface to send packets back to the router for local processing by using the ip unnumbered loopback interfaceSpecifier command on a physical interface that is assigned to be a passive interface in an IS-IS instance, the IS-IS application verifies whether the index of the interface is unnumbered. If the IS-IS instance detects that the IS-IS passive interface is an unnumbered interface, IS-IS does not perform a lookup on the local address time. The IS-IS application performs a lookup to check whether the passive interface has a valid IP address and subnet mask only if it is a numbered interface.

If you configure an interface with only an IPv6 address in IS-IS Router Configuration mode, the UID of the interface is used to perform a binding to the IPv6 interface table. This binding determines if the interface entry is present in the table. If the interface already exists, the interface is configured as a passive interface.


Configuring Adjacency

You can configure the type (level) of adjacency you want to use on an IS-IS interface.

isis circuit-type

Configuring Route Tags for IS-IS Interfaces

To configure a route tag for the IP addresses on an IS-IS interface:

  1. Specify an IS-IS routing process, and access Router Configuration mode.
    host1(config)#router isis engineering host1(config-router)#
  2. Configure a NET for the IS-IS process.
    host1(config-router)#net 47.0010.0000.0000.0000.0001.0001.1111.1111.1111.00
  3. Configure the router to accept and generate only new-style TLV tuples with a wider metric field. New-style TLV tuples include TLV type 135, which contains the route tag.
    host1(config-router)#metric-style wide
  4. Exit Router Configuration mode.
  5. Specify the interface on which you want to route IS-IS.

    The procedure assumes that at least one IP address is already configured on this interface.

    host1(config)#interface atm 2/2.1
  6. Configure a route tag for the interface.
    host1(config-subif)#isis tag 221
  7. Specify the IS-IS process to apply to the interface.
    host1(config-subif)#ip router isis engineering
  8. (Optional) Access Privileged Exec mode, and verify the route tag assignment.
    host1(config-subif)#exit host1(config)#exit host1#show isis database detail

isis tag

Configuring Point-to-Point-over-LAN Circuits

You can deploy IS-IS on broadcast and point-to-point circuits. IS-IS treats these circuits differently in several ways, such as when establishing neighbor adjacencies or flooding link-state information.

Broadcast circuits use designated routers and are represented as virtual nodes in the network topology. They require periodic database synchronization. By default, IS-IS treats the broadcast link as LAN media and tries to bring up the LAN adjacency even when the interface is configured as unnumbered or only a single neighbor exists on that link.

In contrast, point-to-point circuits have less overhead, because they do not use designated routers, the link-state database has no representation of the pseudonode or network LSA, and they do not require periodic database synchronization. However, if more than two routers are connected on the LAN media, routing information in the network is reduced.

Although broadcast circuits are intended to handle more than two devices, in some circumstances you might connect only two routers over the physical or virtual LAN. Even though only two routers are connected, IS-IS treats the circuit as a broadcast circuit that has many more connected routers, with all the associated broadcast overhead but without the benefits of reduced routing information and of optimized flooding that result from having more than two routers on the LAN.

You can use the isis network point-to-point command to configure IS-IS to operate using point-to-point connections on a broadcast circuit when only two routers are on the circuit. This configuration is known as a point-to-point-over-LAN or P2P circuit. This interface configuration tears down the current LAN adjacency that IS-IS has over this interface. IS-IS then reestablishes the adjacency as a point-to-point connection and regenerates the LSPs. The broadcast link is thereafter treated as simple point-to-point interface.

Treating the LAN as a P2P circuit reduces the amount of information that IS-IS has to maintain and manage. For example, there is no need to elect a designated router for the interface. LSP flooding is performed as in P2P links without the need for using periodic CSNPs.

This circuit configuration can be advantageous even when many routers are on the LAN. For example, you might want to organize the routers into multiple smaller VLANs so that you can assign different costs to the IS-IS neighbors. You can apply this configuration to any such VLAN that has only two routers. IS-IS then views the LAN as a mesh of point-to-point connections.

The use of IP unnumbered interfaces makes the most of scarce IP address resources and provides for simpler network management and configuration. This configuration enables IP processing on a point-to-point interface without an explicit IP address. The IP unnumbered interface borrows the IP address of another interface on the node. Point-to-point-over-LAN circuits separate the concept of network type from media type, and enable you to apply unnumbered interface configurations to LANs.

The point-to-point-over-LAN feature requires the following:

isis network point-to-point

Summary Example

host1(config-router)#passive-interface loopback 0 host1(config-if)#interface atm 8/0 host1(config-if)#isis tag 55 host1(config-if)#isis metric 20 level-2 host1(config-if)#isis csnp-interval 30 level-1 host1(config-if)#isis hello-interval 6 level-1 host1(config-if)#isis hello-multiplier 10 level-1 host1(config-if)#isis lsp-interval 100 host1(config-if)#isis retransmit-interval 60 host1(config-if)#isis retransmit-throttle-interval 300 host1(config-if)#isis priority 80 level-1 host1(config-if)#isis circuit-type level-2-only host1(config-intf)#no isis network point-to-point