|
|
The International Organization for Standardization (ISO) Connectionless Network Service (CLNS) protocol is a standard for the network layer of the Open Systems Interconnection (OSI) model.
This chapter describes how to configure ISO CLNS. For a complete description of the commands in this chapter, refer to the "ISO CLNS Commands" chapter of the Router Products Command Reference publication. For historical background and a technical overview of ISO CLNS, see the Internetworking Technology Overview publication.
The Cisco routing software supports packet forwarding and routing for ISO CLNS on networks using a variety of data link layers: Ethernet, Token Ring, Fiber Distributed Data Interface (FDDI), and serial.
You can use CLNS routers on serial interfaces with High-Level Data Link Control (HDLC), Point-to-Point Protocol (PPP), Link Access Procedure, Balanced (LAPB), X.25, Switched Multimegabit Data Service (SMDS), or Frame Relay encapsulation. To use HDLC encapsulation, you must have a router at both ends of the link. If you use X.25 encapsulation, you must manually enter the NSAP-to-X.121 mapping. The LAPB, SMDS, Frame Relay, and X.25 encapsulations interoperate with other vendors.
Cisco's CLNS implementation also is compliant with the Government Open Systems Interconnection Profile (GOSIP) Version 2.
As part of its CLNS support, Cisco routers fully support these ISO and American National Standards Institute (ANSI) standards:
Both the ISO-developed IS-IS and Cisco's ISO Interior Gateway Routing Protocol (IGRP) dynamic routing protocols are supported for dynamic routing of ISO CLNS. In addition, static routing for ISO CLNS is supported.
To configure ISO CLNS, complete the tasks in the following sections:
See the end of this chapter for configuration examples.
In the following section you will learn how to assign NETs, or addresses, for areas and domains. Addressing background material is provided first, followed by a description of how to assign these addresses. The topic of addressing includes the following sections:
Addresses in the ISO network architecture are referred to as network service access point (NSAP) addresses and network entity titles (NETs). Each node in an OSI network has one or more NETs. In addition, each node has many NSAP addresses. Each NSAP address differs from one of the NETs for that node in only the last byte (see Figure 19-1). This byte is called the n-selector. Its function is similar to the port number in other protocol suites.
Our implementation supports all NSAP address formats that are defined by ISO 8348/Ad2; however, we provide dynamic routing (ISO-IGRP or IS-IS routing) only for NSAP addresses that conform to the address constraints defined in the ISO standard for IS-IS (ISO 10589).
An NSAP address consists of two major fields:
The key difference between the ISO-IGRP and IS-IS NSAP addressing schemes is in the definition of area addresses. Both use the system ID for Level 1 routing. However, they differ in the way addresses are specified for area routing. An ISO-IGRP NSAP address includes three separate levels for routing: the domain, area, and system ID. An IS-IS address includes two fields: a single continuous area field comprising the domain and area fields defined for ISO-IGRP and the system ID.
Figure 19-1 illustrates the ISO-IGRP NSAP addressing structure.
The ISO-IGRP NSAP address is divided into three parts: a domain part, an area address, and a system ID. Domain routing is performed on the domain part of the address. Area routing for a given domain uses the area address. System ID routing for a given area uses the system ID part. The NSAP address is laid out as follows:
Our ISO-IGRP routing implementation interprets the bytes from the AFI up to (but not including) the area field in the DSP as a domain identifier. The area field specifies the area, and the system ID specifies the system.
Figure 19-2 illustrates the IS-IS NSAP addressing structure.
An IS-IS NSAP address is divided into two parts: an area address (AA) and a system ID. Level 2 routing uses the AA. Level 1 routing uses the system ID address. The NSAP address is laid out as follows:
The IS-IS routing protocol interprets the bytes from the AFI up to (but not including) the system ID field in the DSP as an area identifier. The system ID specifies the system.
All NSAP addresses must obey the following constraints:
The following are examples of OSI network and GOSIP NSAP addresses using the ISO-IGRP implementation.
The following is the OSI network NSAP address format:
The following is an example of the GOSIP NSAP address structure. This structure is mandatory for addresses allocated from the International Code Designator (ICD) 0005 addressing domain. Refer to the GOSIP document, U.S. Government Open Systems Interconnection Profile (GOSIP), Draft Version 2.0, April 1989, for more information.
Routes are entered by specifying pairs (NSAP prefix and next-hop NET). NETs are similar in function to NSAP addresses. In the routing table, the best match means the longest NSAP prefix entry that matches the beginning of the destination NSAP address. In the following sample static routing table, Table 19-1, the next-hop NETs are listed for completeness, but are not necessary to understand the routing algorithm. Table 19-2 offers examples of how the longest matching NSAP prefix can be matched with routing table entries in Table 19-1.
Table 19-1 Sample Routing Table Entries
| Entry | NSAP Address Prefix | Next-Hop NET |
|---|---|---|
Table 19-2 Hierarchical Routing Examples
| Datagram Destination NSAP Address | Table Entry Number Used |
|---|---|
Octet boundaries must be used for the internal boundaries of NSAP addresses and NETs.
The first task you have to perform to enable CLNS routing is to assign addresses or NETs for your domains and areas. Assign domain and area addresses the same way, following the addressing rules described at the beginning of this section.
First, establish domains. The domain address uniquely identifies the routing domain. All routers within a given domain are given the same domain address. Within each routing domain, you can set up one or more areas. Determine which routers are to be assigned to which areas. The area address uniquely identifies the routing area.
A router can have one or more area addresses. The concept of multiple area addresses is described in the section that follows, "Assign Multiple Area Addresses to IS-IS Areas."
Note ISO-IGRP and IS-IS should not be configured for the same area. Do not specify an NSAP address where all bytes up to (but not including) the system ID are the same when enabling both ISO-IGRP and IS-IS routing.
Note If you are going to configure a routing process, you might want to read the section "Configure a Routing Process," which appears after the discussion of addressing.
To configure NETs, perform the following tasks in the order listed:
| 1This command is documented in the "Interface Commands" chapter of the Router Products Command Reference. |
See the "Basic Static Routing Examples" and "Static Intradomain Routing Example" sections at the end of this chapter for examples of configuring NETs.
IS-IS routing supports assignment of multiple area addresses on the same router. This concept is referred to as multihoming. You must statically assign the multiple area addresses on the router. All of the addresses must have the same system ID. For example, you can assign one address (area1 plus system ID), and two additional addresses in different areas (area2 plus system ID and area3 plus system ID) where the system ID is the same. The number of areas allowed in a domain is unlimited.
Multihoming provides a mechanism for smoothly migrating network addresses:
A router can dynamically learn about any adjacent router. As part of this process, the routers inform each other of their area addresses. If two routers share at least one area address, the set of area addresses of the two routers are merged. The merged set cannot contain more than three addresses. If there are more than three, the three addresses with the lowest numerical values are kept, and all others are dropped.
To configure multiple area addresses in IS-IS areas statically, perform the following task in the order listed:
| Step 2. Enable IS-IS routing, which places you in router configuration mode. | |
| Step 3. Configure NETs for the routing process. You can perform this step as many as three times. |
| 1This command is documented in the "Interface Commands" chapter of the Router Products Command Reference. |
See the "Configuring NETs Examples" section at the end of this chapter for examples of configuring NETs and multiple area addresses.
You must assign static addresses if you have configured the router to support ISO CLNS but you are not using a routing protocol.
A CLNP packet sent to any of the defined NSAP addresses or NETs will be received by the router. The router uses the following algorithm to select the NET to use when it sends a packet:
To assign an address to the router if you are not dynamically routing CLNS packets, perform the following task in global configuration mode:
| Task | Command |
|---|---|
Assign an address to the router when the router is not configured to dynamically route CLNS packets using ISO-IGRP or IS-IS. |
Conceptually, each end system (ES) lives in one area. It discovers the nearest intermediate system (IS) router by listening to ES-IS packets. Each ES must be able to communicate directly with an IS in its area.
When an ES wants to communicate with another ES, it sends the packet to any IS on the same medium. The IS looks up the destination NSAP address and forwards the packet along the best route. If the destination NSAP address is for an ES in another area, the Level 1 IS sends the packet to the nearest Level 2 IS. The Level 2 IS forwards the packet along the best path for the destination area until it gets to a Level 2 IS that is in the destination area. This IS then forwards the packet along the best path inside the area until it is delivered to the destination ES.
End systems need to know how to get to a Level 1 IS for their area, and Level 1 ISs need to know all of the ESs that are directly reachable through each of their interfaces. To provide this information, the routers support the ES-IS protocol. A router dynamically discovers all ESs running the ES-IS protocol. ESs that are not running the ES-IS protocol must be statically configured.
It is sometimes desirable for a router to have a neighbor entry statically configured rather than learned through ES-IS, ISO-IGRP, or IS-IS.
Perform the following tasks in interface configuration mode, as needed, to statically enter mapping information between the NSAP protocol addresses and the subnetwork point of attachment (SNPA) addresses for end systems or intermediate systems:
Note It is necessary to use static mapping only for those end systems that do not support ES-IS. The router continues to dynamically discover those end systems that do support ES-IS.
If there are systems on the Ethernet that do not use ES-IS, or if X.25 is being used and no dynamic routing protocol is running over the X.25 network, you specify the NSAP/NET (protocol address) to SNPA (media address) mappings by performing the following tasks in interface configuration mode:
| 1This command is documented in the "X.25 and LAPB Commands" chapter of the Router Products Command Reference publication. |
If you have configured interfaces for ISO-IGRP or IS-IS, the ES-IS routing software automatically turns ES-IS on for those interfaces.
You can define a name-to-NSAP address mapping. This name can then be used in place of typing the long set of numbers associated with an NSAP address.
To define a name-to-NSAP address mapping, perform the following task in global configuration mode:
The assigned NSAP name is displayed, where applicable, in show and debug EXEC commands.
There are some effects and requirements associated with using names to represent NETs and NSAP addresses, however; they include the following:
The commands that are affected by these requirements include:
If your router has both ISO CLNS and IP enabled, you can use the Domain Name System (DNS) to query ISO CLNS addresses by using the NSAP address type, as documented in RFC 1348. This feature is useful for the ISO CLNS ping EXEC command and when making Telnet connections. This feature is enabled by default.
To enable or disable DNS queries for ISO CLNS addresses, perform the following tasks in global configuration mode:
| Task | Command |
|---|---|
The basic function of a router is to forward packets: to receive a packet in one interface and send it out another (or the same) interface to the proper destination. All routers do this by looking up the destination address in a table. The tables can be built either dynamically or statically. If you are configuring all of the entries in the table yourself, you are using static routing. If you have a routing process building the tables, you are using dynamic routing. It is possible, and sometimes necessary, to use both static and dynamic routing simultaneously.
When you configure only ISO CLNS and not routing protocols, the router only makes forwarding decisions. It does not perform other routing-related functions. In such a configuration, the router compiles a table of adjacency data, but does not advertise this information. The only information that is inserted into the routing table is the NSAP and network entity title (NET) addresses of this router, static routes, and adjacency information.
Static routing is used when it is not possible or desirable to use dynamic routing. The following are some instances of when you would use static routing:
We support two dynamic routing protocols for CLNP networks:
Both routing protocols support the concept of areas. Within an area, all routers know how to reach all of the system IDs. Between areas, routers know how to reach the proper area.
IS-IS supports two levels of routing: station routing (within an area) and area routing (between areas). ISO-IGRP supports three levels of routing: system routing, area routing, and interdomain routing. Routing across domains (interdomain routing) can be done either statically or dynamically with ISO-IGRP.
Some intermediate systems keep track of how to communicate with all of the end systems in their areas and thereby function as Level 1 routers (also referred to as local routers). Other intermediate systems keep track of how to communicate with other areas in the domain, functioning as Level 2 routers (sometimes referred to as area routers). Our routers are always Level 1 and Level 2 routers when routing ISO-IGRP; they can be configured to be Level 1 only, Level 2 only, or both Level 1 and Level 2 routers when routing IS-IS.
End systems communicate with intermediate systems using the ES-IS protocol. Level 1 and Level 2 intermediate systems communicate with each other using either ISO IS-IS or our ISO-IGRP protocol.
This section describes the tasks associated with each routing protocol. When dynamically routing, you can choose either ISO-IGRP or IS-IS, or you can route both routing protocols at the same time.
You do not need to explicitly specify a routing process to use static routing facilities. If you choose static routing, the configuration process begins with enabling CLNS routing on the router. CLNS routing is enabled by default when you configure either of the routing protocols.
CLNS static routing is configured when you do the following:
Step 2 Assign a static NET address for the router.
Step 3 Enable ISO CLNS for each interface.
Step 4 Enter a specific static route.
Step 5 Configure other, optional variations.
Each of these steps is described in the following sections.
See the "Basic Static Routing Examples," "Static Intradomain Routing Example," and " Static Interdomain Routing Example" sections at the end of this chapter for examples of configuring static routes.
To configure CLNS on the router, perform the following task in global configuration mode:
If you have configured a router to support ISO CLNS but you have not configured it to route CLNS packets dynamically using ISO-IGRP or IS-IS, then you must assign an address to the router.
A CLNP packet sent to any of the defined NSAP addresses or NETs will be received by the router. The router uses the following algorithm to select the NET to use when it sends a packet:
To assign an address to the router, perform the following task in global configuration mode:
| Task | Command |
|---|---|
Assign an address to the router when the router is not configured to dynamically route CLNS packets using ISO-IGRP or IS-IS. |
You also must enable ISO CLNS for each interface. This is done automatically when you configure IS-IS or ISO-IGRP routing on an interface; however, if you do not intend to perform any dynamic routing on an interface but intend to pass ISO CLNS packet traffic to end systems, you must enable CLNS yourself.
Enable ISO CLNS when you want to pass ISO CLNS packet traffic to end systems but do not want to perform any dynamic routing on an interface. Perform the following task in interface configuration mode:
You can enter a specific static route and apply it globally even when you are dynamically routing. NSAP addresses that start with the NSAP prefix you specify are forwarded to the next-hop node.
To apply a specific static route globally, perform the following task in global configuration mode:
You also can perform the following tasks that use variations of the clns route global configuration command:
The following list shows how to perform each of these tasks. Perform these tasks in global configuration mode:
| Task | Command |
|---|---|
Explicitly tell a router to discard packets with the specified NSAP prefix. |
|
Configure a default prefix rather than specify an NSAP prefix. |
CLNS routing is enabled by default on the router when you configure ISO-IGRP. All you need to do to specify an ISO-IGRP routing process is to enable the ISO-IGRP routing process, identify the address for the router, and specify the interfaces that are to route ISO-IGRP. Optionally, you can set a level for your routing updates when you configure the interfaces. You can specify up to ten ISO-IGRP processes.
To configure ISO-IGRP dynamic routing, perform the following tasks in the order listed:
Although IS-IS allows you to configure multiple NETs, ISO-IGRP allows only one NET per process.
You can configure an interface to advertise Level 2 information only. This option reduces the amount of router-to-router traffic by telling the router to send out only Level 2 routing updates on certain interfaces. Level 1 information is not passed on the interfaces for which the Level 2 option is set.
The additional tasks that follow allow you to customize ISO-IGRP.
See the "Dynamic Routing within the Same Area Example," "Dynamic Routing in More Than One Area Example," and "Dynamic Routing in Overlapping Areas Example" sections at the end of this chapter for examples of configuring dynamic routing.
You can also configure the following ISO-IGRP parameters:
You have the option of altering the default behavior of ISO-IGRP routing and metric computations. This allows, for example, tuning of system behavior to allow for transmissions via satellite. Although ISO-IGRP metric defaults were carefully selected to provide excellent operation in most networks, you can adjust the metric.
Note Adjusting the ISO-IGRP metric can dramatically affect network performance, so ensure that all metric adjustments are made carefully. Due to the complexity of this task, it is not recommended unless it is done with guidance from an experienced system designer.
You can use different metrics for the ISO-IGRP routing protocol on CLNS. By performing the following task, you can configure the metric constants used in the ISO-IGRP composite metric calculation of reliability and load. Perform this task in router configuration mode:
Two additional ISO-IGRP metrics can be configured. These are the bandwidth and delay associated with an interface. Refer to the "Interface Commands" chapter of the Router Products Command Reference publication for details about the bandwidth and delay interface configuration commands used to set these metrics, and to the "Configuring Interfaces" chapter of this manual for configuration information.
Note Using the bandwidth and delay commands to change the values of the ISO-IGRP metrics also will change the values of IP IGRP metrics.
The basic timing parameters for ISO-IGRP are adjustable. Because the ISO-IGRP routing protocol executes a distributed, asynchronous routing algorithm, it is important that these timers be the same for all routers in the network.
To adjust ISO-IGRP timing parameters, perform the following task in router configuration mode:
Split horizon blocks information about routes from being advertised out the interface from which that information originated. This feature usually optimizes communication among multiple routers, particularly when links are broken.
To either enable or disable split horizon for ISO-IGRP updates, perform the following tasks in interface configuration mode:
| Task | Command |
|---|---|
The default for all LAN interfaces is for split horizon to be enabled; the default for WAN interfaces on X.25, Frame Relay, or SMDS networks is for split horizon to be disabled.
You can configure a router to do interdomain dynamic routing by putting it into two domains and configuring it to redistribute the routing information between the domains. Routers configured this way are referred to as border routers. If you have a router that is in two routing domains, you might want to redistribute routing information between the two domains.
Also, you can conditionally control the redistribution of routes between routing domains by defining route maps between the two domains. Route maps allow you to use tags in routes to influence route redistribution. These methods of specifying route redistribution are listed in the following tables.
Static routes by default are redistributed into ISO-IGRP routing domains.
Perform the following tasks in router configuration mode:
| Task | Command |
|---|---|
Redistribute routes from one routing protocol into other routing domains. |
|
Redistribute static routes by causing the specified routing process to advertise static CLNS routes. |
Perform the following task in global configuration mode:
| Task | Command |
|---|---|
When multiple routing processes are running in the same router for CLNS, it is possible for the same route to be advertised by more than one routing process. The router always picks the route whose routing protocol has the lowest administrative distance.The lower the value of the distance, the more preferred the route.
Default administrative distances are already set. By default, the following administrative distances are assigned:
However, if you need to change an administrative distance for a route, perform the following task in router configuration mode:
If you want an ISO-IGRP prefix route to override a static route, you must set the distance for the routing process to be lower than 10.
CLNS routing is enabled by default on the router when you configure IS-IS dynamic routing. All you need to do to specify an IS-IS routing process is to enable the process, identify the address for the router, and specify the interfaces that are to route IS-IS. You can specify only one IS-IS process per router.
To configure IS-IS dynamic routing, perform the following required tasks in the order listed:
For IS-IS, multiple NETs per router are allowed, with a maximum of three. However, only one IS-IS process is allowed, whether you run it in integrated mode, ISO CLNS only, or IP only.
See the "IS-IS Routing Configuration Examples" section at the end of this chapter for examples of configuring IS-IS routing.
Our IS-IS implementation allows you to customize certain interface-specific IS-IS parameters. You can perform the following optional tasks:
You are not required to alter any of these parameters, but some interface parameters must be consistent across all routers in an attached network. Therefore, be sure that if you do configure any of these parameters, the configurations for all routers on the network have compatible values.
You can configure a cost for a specified interface. The default metric is used as a value for the IS-IS metric. This is the value assigned when there is no quality of service (QOS) routing performed. The only metric that is supported by the router and that you can configure is the default-metric, which you can configure for Level 1 and/or Level 2 routing.
To configure the link state metric, perform the following task in interface configuration mode:
| Task | Command |
|---|---|
isis metric default-metric delay-metric expense-metric error-metric {level-1 | level-2} |
You can specify the length of time in seconds between hello packets that the Cisco IOS software sends on the interface. You can also change the default hello packet multiplier used on the interface to determine the hold time transmitted in IS-IS hello packets (the default is 3).
The hold time determines how long a neighbor waits for another hello packet before declaring the neighbor down. This time determines how quickly a failed link or neighbor is detected so that routes can be recalculated.
To set the advertised hello interval and multiplier, perform the following task in interface configuration mode:
The hello interval can be configured independently for Level 1 and Level 2, except on serial point-to-point interfaces. (Because there is only a single type of hello packet sent on serial links, it is independent of Level 1 or Level 2.) Specify an optional level for X.25, SMDS, and Frame Relay multiaccess networks.
Use the isis hello-multiplier command in circumstances where hello packets are lost frequently and IS-IS adjacencies are failing unnecessarily. You can raise the hello multiplier and lower the hello interval (isis hello-interval command) correspondingly to make the hello protocol more reliable without increasing the time required to detect a link failure.
Complete sequence number PDUs (CSNPs) are sent by the designated router to maintain database synchronization.
You can configure the IS-IS CSNP interval for the interface by performing the following task in interface configuration mode:
This feature does not apply to serial point-to-point interfaces. It does apply to WAN connections if the WAN is viewed as a multiaccess meshed network.
You can configure the number of seconds between retransmission of IS-IS link state PDUs (LSPs) for point-to-point links.
To set the retransmission level, perform the following task in interface configuration mode:
| Task | Command |
|---|---|
Configure the number of seconds between retransmission of IS-IS LSPs for point-to-point links. |
The value you specify should be an integer greater than the expected round-trip delay between any two routers on the attached network. The setting of this parameter should be conservative, or needless retransmission will result. The value should be larger for serial lines and virtual links.
You can configure the priority to use for designated router election. Priorities can be configured for Level 1 and Level 2 individually.
To configure the priority to use for designated router election, perform the following task in interface configuration mode:
You can specify adjacency levels on a specified interface.
To configure the adjacency for neighbors on the specified interface, perform the following task in interface configuration mode:
| Task | Command |
|---|---|
Configure the type of adjacency desired for neighbors on the specified interface (specify the interface circuit type). |
If you specify Level 1, a Level 1 adjacency might be established if there is at least one area address common to both this system and its neighbors.
If you specify both Level 1 and Level 2, a Level 1 and 2 adjacency is established if the neighbor is also configured as both Level 1 and Level 2 and there is at least one area in common. If there is no area in common, a Level 2 adjacency is established. This is the default value.
If you specify Level 2, a Level 2 adjacency is established.
Note that it is seldom necessary to configure an interface as Level 1 only or Level 2 only—the protocols will determine the adjacency type automatically.
You can assign different passwords for different routing levels. By default, authentication is disabled. Specifying Level 1 or Level 2 disables the password only for Level 1 or Level 2 routing, respectively. If you do not specify a level, the default is Level 1.
To configure an authentication password for an interface, perform the following task in interface configuration mode:
| Task | Command |
|---|---|
You can configure the following IS-IS parameters:
If you have a router that is in two routing domains, you might want to redistribute routing information between the two domains. First, you specify the destination routing protocol, then you define the routing protocol that is to be redistributed into the destination routing protocol.
Redistribution only occurs for Level 2 routing.
You also can conditionally control the redistribution of routes between routing domains by defining route maps between them.
Additionally, you can cause the specified routing process to advertise static CLNS routes. Static routes are always redistributed into IS-IS unless you explicitly disable this feature.
The methods of specifying route redistribution are listed in the following tables.
Perform the following tasks in router configuration mode:
| Task | Command |
|---|---|
Redistribute routes from one routing protocol into other routing domains. |
|
Redistribute static routes by causing the specified routing process to advertise static CLNS routes. |
Perform the following task in global configuration mode:
| Task | Command |
|---|---|
See the "Route-Map Examples" section at the end of this chapter for examples of configuring route maps.
When multiple routing processes are running in the same router for CLNS, it is possible for the same route to be advertised by more than one routing process. The router always picks the route whose routing protocol has the lowest administrative distance.The lower the value of the distance, the more preferred the route.
By default the following administrative distances are assigned:
However, if you need to change an administrative distance for a route, perform the following task in router configuration mode:
If you want an IS-IS prefix route to override a static route, you must set the distance for the routing process to be lower than 10.
You can configure the router to act as a Level 1 (intra-area) router, as both a Level 1 router and a Level 2 (interarea) router, or as an interarea router only.
To configure the IS-IS level, perform the following task in router configuration mode:
Note that it is seldom necessary to configure the IS type because the IS-IS protocol will automatically establish this.
You can assign passwords to areas and domains. An area password is inserted in Level 1 (station router level) LSPs, CSNPs, and partial sequence number PDUs (PSNPs). A routing domain authentication password is inserted in Level 2 (the area router level) LSP, CSNP, and PSNP PDUs.
To configure area or domain passwords, perform the following tasks in router configuration mode:
| Task | Command |
|---|---|
You can allow the router to ignore IS-IS link state packets (LSPs) that are received with internal checksum errors rather than purging the LSPs. LSPs are used by the receiving routers to maintain their routing tables.
The IS-IS protocol definition requires that a received LSP with an incorrect data link checksum be purged by the receiver, which causes the initiator of the LSP to regenerate it. However, if a network has a link that causes data corruption while still delivering LSPs with correct data link checksums, a continuous cycle of purging and regenerating large numbers of LSPs can occur. Because this could render the network nonfunctional, use the ignore-lsp-errors to ignore these LSPs rather than purge the LSPs.
To allow the router to ignore LSPs with an internal checksum error, perform the following task in router configuration mode:
You can allow IS-IS to generate a log message when an IS-IS adjacency changes state (up or down). This may be very useful when monitoring large networks. Messages are logged using the system error message facility. Messages are of the form:
%CLNS-5-ADJCHANGE: ISIS: Adjacency to 0000.0000.0034 (Serial0) Up, new adjacency
%CLNS-5-ADJCHANGE: ISIS: Adjacency to 0000.0000.0034 (Serial0) Down, hold time expired
To generate log messages when an IS-IS adjacency changes state, perform the following task in router configuration mode:
Under normal condition, the default MTU size should be sufficient. However, if the MTU of a link is lowered to less than 1500 bytes, the LSP MTU must be lowered accordingly on each router in the network. If this is not done, routing will become unpredictable.
 | Caution The CLNS MTU of a link (which is the applicable value for IS-IS, even if it is being used to route IP) may differ from the IP MTU. To be certain about a link MTU as it pertains to IS-IS, use the show clns interface command to display the value. |
To set the maximum transmission unit (MTU) size of IS-IS link state packets (LSPs), perform the following task in router configuration mode:
| Task | Command |
|---|---|
Specify the maximum LSP packet size, in bytes. The size must be less than or equal to the smallest MTU of any link in the network. The default size is 1497 bytes. |
Note This rule applies for all routers in a network. If any link in the network has a reduced MTU, all routers must be changed, not just to the routers directly connected to such a link.
You can configure ES-IS parameters for communication between end systems and routers. In general, you should leave these parameters at their default values.
When configuring an ES-IS router, be aware of the following:
ISs and ESs periodically send out hello packets to advertise their availability. The frequency of these hello packets can be configured.
The recipient of a hello packet creates an adjacency entry for the system that sent it. If the next hello packet is not received within the interval specified, the adjacency times out and the adjacent node is considered unreachable.
A default rate has been set for hello packets; however, you can change the default by performing the following task in global configuration mode:
| Task | Command |
|---|---|
A default rate has been set for packet validity; however, you can change the default by performing the following task in global configuration mode:
| Task | Command |
|---|---|
Allow the sender of an ESH or ISH packet to specify the length of time you consider the information in these packets to be valid. |
A default rate has been set for the ES Configuration Timer (ESCT) option; however, you can change the default by performing the following task in interface configuration mode:
You can build powerful CLNS filter expressions, or access lists, that can be used to control either the forwarding of frames through router interfaces or the establishment of adjacencies with or the application of filters to any combination of ES or IS neighbors, ISO-IGRP neighbors, or IS-IS neighbors.
CLNS filter expressions are complex logical combinations of CLNS filter sets. CLNS filter sets are lists of address templates against which CLNS addresses are matched. Address templates are CLNS address patterns that are either simple CLNS addresses that match just one address, or match multiple CLNS addresses through the use of wildcard characters, prefixes, and suffixes. Frequently used address templates can be given aliases for easier reference.
To establish CLNS filters, perform the following tasks in global configuration mode:
| Task | Command |
|---|---|
Build filter sets of multiple address template permit and deny conditions. |
|
Build filter expressions, combining the use of one or more filter sets. |
Perform the following tasks in interface configuration mode:
See the "CLNS Filter Examples" section at the end of this chapter for examples of configuring CLNS filters.
This section provides general information about running ISO CLNS over WANs. For more information, see the relevant chapters describing the specific types of encapsulation that might be used.
You can use CLNS routers on serial interfaces with HDLC, PPP, LAPB, X.25, Frame Relay, DDR, or SMDS encapsulation. To use HDLC encapsulation, you must have a router at both ends of the link. If you use X.25 encapsulation, and if IS-IS or ISO-IGRP is not used on an interface, you must manually enter the NSAP-to-X.121 mapping. The LAPB, SMDS, Frame Relay, and X.25 encapsulations interoperate with other vendors.
Both ISO-IGRP and IS-IS can be configured over WANs.
X.25 is not a broadcast medium; therefore, ES-IS generally is not used to automatically advertise and record mappings between NSAP/NET (protocol addresses) and subnetwork points of attachment (SNPA) (media addresses). (With X.25, the SNPAs are the X.25 network addresses [X.121 addresses]. These are usually assigned by the X.25 network provider.) If you use static routing, you must configure the NSAP-to-X.121 mapping.
Configuring a serial line to use CLNS over X.25 requires configuring the general X.25 information and the CLNS-specific information. First, configure the general X.25 information. Then, enter the static mapping information.
You can specify nondefault packet and window sizes, reverse charge information, and so on. The X.25 facilities information that can be specified is exactly the same as in the x25 map interface configuration command described in the "Configuring X.25 and LAPB" chapter.
See the "Configuring ISO CLNS over X.25 Example" section at the end of this chapter for an example of configuring CLNS over X.25.
Perform the optional tasks in the following sections to configure miscellaneous features of an ISO CLNS network:
You can assign an NSAP address for a specific interface. This allows the router to advertise different addresses on each interface. This is useful if you are doing static routing and need to control the source NET used by the router on each interface.
To assign an NSAP address for a specified interface, perform the following task in interface configuration mode:
DECnet Phase V cluster aliasing allows multiple systems to advertise the same system ID in end-system hello packets. The router does this by caching multiple ES adjacencies with the same NSAP address, but different SNPA addresses. When a packet is destined to the common NSAP address, the router splits the packet loads among the different SNPA addresses. A router that supports this capability forwards traffic to each system. You can do this on a per-interface basis.
To configure cluster aliases, perform the following task in interface configuration mode:
| Task | Command |
|---|---|
Allow multiple systems to advertise the same system ID in end-system Hello packets on a per-interface basis. |
If DECnet Phase V cluster aliases are disabled on an interface, ES Hello packet information is used to replace any existing adjacency information for the NSAP address. Otherwise, an additional adjacency (with a different SNPA) is created for the same NSAP address.
See the "Configuring DECnet Cluster Aliases Example" section at the end of this chapter for an example of configuring DECnet OSI cluster aliases.
If you have an old DECnet implementation of ES-IS in which the NSAP address advertised in an ISH does not have the N-selector byte present, you may want to configure the router to allow ISHs sent and received to ignore the N-selector byte. The n-selector byte is the last byte of the NSAP address.
To enable Digital-compatible mode, perform the following task in interface configuration mode:
By default, the router will discard any packets it sees as set with security options. You can disable this behavior; that is, allow such packets to pass through.
Perform the following task in global configuration mode:
The ISO CLNS routing software ignores the Record Route option, the Source Route option, and the QOS (quality of service) option other than congestion experienced. The security option causes a packet to be rejected with a bad option indication.
Generally, you do not need to change the default settings for CLNS packet switching, but there are some modifications you can make when you decide to make changes in your network's performance. This section describes the following ISO CLNS parameters that you can change:
See the "Customizing Performance Parameters Example" section at the end of this chapter for examples of configuring various performance parameters.
All interfaces have a default maximum packet size. You can, however, set the maximum transmission unit (MTU) size of the packets sent on the interface to reduce fragmentation. The minimum value is 512; the default and maximum packet size depends on the interface type.
Changing the MTU value of packets sent on the interface can affect the CLNS MTU value. If the CLNS MTU is at its maximum given the interface MTU, then the CLNS MTU will change with the interface MTU. However, the reverse is not true: changing the CLNS MTU value has no effect on the mtu value of other packets sent on the interface.
To set the CLNS MTU packet size for a specified interface, perform the following task in interface configuration mode:
The CTR card does not support the switching of frames larger than 4472 bytes. Interoperability problems might occur if CTR cards are intermixed with other Token Ring cards on the same network. These problems can be minimized by lowering the CLNS MTU sizes to be the same on all devices on the network.
When the ISO CLNS routing software sources a CLNS packet, by default it generates checksums. You can disable this function.
Perform the following task in interface configuration mode:
Note Enabling checksum generation has no effect on routing packets (ES-IS, ISO-IGRP, and IS-IS) sourced by the system. It applies to pings and trace-route packets.
Fast switching through the cache is enabled by default for all supported interfaces. You can disable fast switching by performing the following task in interface configuration mode:
Note The cache still exists and is used after the no clns route-cache command is used; the software just does not do fast switching through the cache.
If a router configured for CLNS experiences congestion, it sets the congestion-experienced bit. You can set the congestion threshold on a per-interface basis. By setting this threshold, you cause the system to set the congestion-experienced bit if the output queue has more than the specified number of packets in it.
To set the congestion threshold, perform the following task in interface configuration mode:
When a CLNS packet is received, the routing software looks in the routing table for the next hop. If it does not find one, the packet is discarded and an error protocol data unit (ERPDU) might be sent.
You can set an interval between ERPDUs. Doing so reduces bandwidth if this feature is disabled. When you determine the minimum interval between ERPDUs, the router does not send ERPDUs more frequently than one per interface per ten milliseconds.
To transmit ERPDUs, perform the following tasks in interface configuration mode:
| Task | Command |
|---|---|
Send an ERPDU when the routing software detects an error in a data PDU; this is enabled by default. |
|
Determine the minimum interval, in milliseconds, between ERPDUs. |
If a packet is sent out the same interface it came in on, a redirect PDU (RDPDU) also can be sent to the sender of the packet. You can control RDPDUs in the following ways:
To control RDPDUs, perform either of the following tasks in interface configuration mode:
| Task | Command |
|---|---|
Allow CLNS to send redirect PDUs when a better route for a given host is known. |
|
Determine the minimum interval time, in milliseconds, between RDPDUs. |
Note SNPA masks are never sent, and RDPDUs are ignored by the router when the router is acting as an IS.
To configure parameters for packets sourced by a specified router, perform either of the following tasks in global configuration mode:
| Task | Command |
|---|---|
Globally specify in seconds the initial lifetime for locally generated packets for the specified router. |
|
Specify whether to request error PDUs on packets sourced by the router. |
It is a good idea to set the packet lifetime low in an internetwork that has frequent loops.
Note The clns want-erpdu command has no effect on routing packets (ES-IS, ISO-IGRP, and IS-IS) sourced by the system. It applies to pings and trace route packets.
Use the EXEC commands described in this section to monitor and maintain the ISO CLNS caches, tables, and databases.
The following sections provide configuration examples of both intra- and interdomain static and dynamic routing using static, ISO-IGRP, and IS-IS routing techniques:
The following are simple examples of configuring NETs for both ISO-IGRP and IS-IS.
The following example illustrates specifying an NET:
The following example illustrates using a name for an NET:
The use of this net router configuration command configures the system ID, area address, and domain address. Only a single NET per routing process is allowed.
The following example illustrates specifying a single NET:
The following example illustrates using a name for an NET:
The following example illustrates the assignment of three separate area addresses for a single router using net commands. Traffic received that includes an area address of 47.0004.004d.0001, 47.0004.004d.0002, or 47.0004.004d.0003, and that has the same system ID, is forwarded to this router.
Configuring FDDI, Ethernets, Token Rings, and serial lines for CLNS can be as simple as just enabling CLNS on the interfaces. This is all that is ever required on serial lines using HDLC encapsulation. If all systems on an Ethernet or Token Ring support ISO 9542 ES-IS, then nothing else is required.
In the following example, an Ethernet and a serial line can be configured as follows:
The following is a more complete example of CLNS static routing on a system with two Ethernet interfaces. After configuring routing, you define an NET and enable CLNS on the Ethernet 0 and Ethernet 1 interfaces. You must then define an ES neighbor and define a static route with the clns route global configuration command, as shown. In this situation, there is an ES on Ethernet 1 that does not support ES-IS. Figure 19-3 illustrates this network.
Figure 19-4 and the configurations that follow demonstrate how to use static routing inside of a domain. Imagine a company with branch offices in Detroit and Chicago, connected with an X.25 link. These offices are both in the domain named Sales.
The following example shows one way to configure the router in Chicago:
This configuration brings up an X.25 virtual circuit between the router in Chicago and the router in Detroit. Routing updates will be sent across this link. This implies that the virtual circuit could be up continuously.
If the Chicago office should grow to contain multiple routers, it would be appropriate for each of those routers to know how to get to Detroit. Add the following command to redistribute information between routers in Chicago:
Figure 19-5 and the example configurations that follow illustrate how to configure two routers that distribute information across domains. In this example, Router A (in domain Orion) and Router B (in domain Pleiades) communicate across a serial link.
The following configuration shows how to configure Router A for static interdomain routing:
The following configuration shows how to configure Router B for static interdomain routing:
CLNS routing updates will not be sent on the serial link; however, CLNS packets will be sent and received over the serial link.
Figure 19-6 and the example configuration that follows illustrate how to configure dynamic routing within a routing domain. The router can exist in one or more areas within the domain. The router named Router A exists in a single area.
Figure 19-7 and the example configuration that follows illustrate how to configure a router named Router A that exists in two areas.
The example that follows illustrates how to configure a router with overlapping areas:
Figure 19-8 and the configurations that follow illustrate how to configure three domains that are to be transparently connected.
The following configuration shows how to configure Router Chicago for dynamic interdomain routing:
The following configuration shows how to configure Router Detroit for dynamic interdomain routing. Comment lines have been eliminated from this example to avoid redundancy.
Chicago injects a prefix route for domain A into domain B. Domain B injects this prefix route and a prefix route for domain B into domain C.
You also can configure a border router between domain A and domain C.
The examples that follow illustrate the basic syntax and configuration command sequence for IS-IS routing.
The following example illustrates using the IS-IS protocol to configure a single area address for Level 1 and Level 2 routing:
The following example illustrates a similar configuration, featuring a single area address being used for specification of Level 1 and Level 2 routing. However, in this case, interface serial interface 0 is configured for Level 2 routing only. Most comment lines have been eliminated from this example to avoid redundancy.
The following example illustrates an OSI configuration example. In this example, IS-IS runs with two area addresses, metrics tailored, and different circuit types specified for each interface. Most comment lines have been eliminated from this example to avoid redundancy.
The following example illustrates route redistribution between IS-IS and ISO-IGRP domains. In this case, the IS-IS domain is on Ethernet interface 0; the ISO-IGRP domain is on serial interface 0. The IS-IS routing process is assigned a null tag; the ISO-IGRP routing process is assigned a tag of remote-domain. Most comment lines have been eliminated from this example to avoid redundancy.
The following two examples show how to configure a router in two areas. The first example configures ISO-IGRP; the second configures IS-IS.
In the following example, the router is in domain 49.0001 and has a system ID of aaaa.aaaa.aaaa. The router is in two areas: 31 and 40 (decimal). Figure 19-9 illustrates this configuration.
To run IS-IS instead of ISO-IGRP, use this configuration. The illustration in Figure 19-9 still applies. Ethernet interface 2 is configured for IS-IS routing and is assigned the tag of test-proc2.
To allow CLNS packets only to blindly pass through an interface without routing updates, you could use a simple configuration. The following example shows such a configuration:
In the following example, serial interface 1 on Router A acts as a DTE for X.25. It permits broadcasts to pass through. Router B is an IS, which has a CLNS address of 49.0001.bbbb.bbbb.bbbb.00 and an X.121 address of 31102. Router A has a CLNS address of 49.0001.aaaa.aaaa.aaaa.00 and an address of 31101. Figure 19-10 illustrates this configuration.
The following example shows how to set ES hello packet (ESH) and IS hello packet (ISH) parameters in a simple ISO-IGRP configuration, as well as the MTU for a serial interface:
The following example enables cluster aliasing for CLNS:
The following example redistributes two types of routes into the integrated IS-IS routing table (supporting both IP and CLNS). The first routes are OSPF external IP routes with tag 5, and these are inserted into level-2 IS-IS LSPs with a metric of 5. The second routes are ISO-IGRP derived CLNS prefix routes that match CLNS filter expression "osifilter." These are redistributed into IS-IS as level-2 LSPs with a metric of 30.
Given the following configuration, a RIP learned route for network 160.89.0.0 and an ISO-IGRP learned route with prefix 49.0001.0002 will be redistributed into an IS-IS level-2 LSP with a metric of 5.
The following example returns a permit action if an address starts with either 47.0005 or 47.0023. It returns an implicit deny action on any other address.
The following example returns a deny action if an address starts with 39.840F, but returns a permit action for any other address:
The following example builds a filter that accepts end system adjacencies with only two systems, based only on their system IDs:
Posted: Wed Jul 2 23:32:33 PDT 2003
All contents are Copyright © 1992--2003 Cisco Systems, Inc. All rights reserved.
Important Notices and Privacy Statement.