WP CI FourElement of is-Is

8/11/2019 WP CI FourElement of is-Is

1/24

Expert Reference Series of White Papers

The Four Elementsof IS-IS

1-800-COURSES www.globalknowledge.com
http://www.globalknowledge.com/http://www.globalknowledge.com/http://www.globalknowledge.com/


2/24

The Four Elements of IS-ISRaymond B. Dooley, CCSI, Global Knowledge Course Director

IntroductionAll advanced IP routing protocols (OSPF, EIGRP, IS-IS) have several elements that are always present. This isbecause the function of all routing protocols is the same, which is to find the best path to an IP destinationaddress.

There are four common elements:

1. Operations, processes, and rules for the exchange of packets to accomplish the routing protocolobjective.

2. Construction and maintenance of the routing protocol database (link state databaseLevel 1 and Level 2LSDB in this case).

3. Construction and use of the data structures created by elements one and two.

4. Configuration and verification of the entire process.

The purpose of this paper is to examine these elements for the implementation of Intermediate System Intermediate System (IS-IS).

Operation of IS-IS with Shortest-Path-First (SPF)

AlgorithmIS-IS was developed by the International Organization for Standardization (ISO). Virtually all vendors of routingand switching equipment support IS-IS. Everything for a routing protocol starts with the initial exchange ofpackets. As opposed to EIGRP, OSPF, and BGP, IS-IS is unique in the way routing protocol packets (Link StatePackets [LSPs]) are exchanged.

Where OSPF and EIGRP transmit packets encapsulated in IP, IS-IS sends packets encapsulated at Layer 2 overeither Ethernet or point-to-point links. This gives IS-IS tremendous flexibility to not only build tables for IP routes,

but also things like switch number and MAC, IP address and MAC, to support such Cisco protocols as Fabric Pathand Overlay Transport Virtualization in the data center. That feature of IS-IS will not be covered, and IP routingwill be the focus of this paper.

Copyright 2014 Global Knowledge Training LLC. All rights reserved. 2


3/24

The graphic shows the IS-IS Protocol Data Units (PDUs). IS-IS was originally designed to carry ISO datagrams(Layer 3 of ISO Connectionless Network Protocol [CLNP], which is the equivalent of IP for ISO).

One packet type is used to support CLNP, which is not relevant if routing only IP.

The middle packet type is to provide hellos to ISO End Systems (ESs), which are the ISO equivalent of an IP host. Ifonly IP routing is supported, then that packet is also irrelevant since there are no ISO workstations.

So, the only relevant PDU is the one going from router to router, which is the top one in the graphic.

Notice that the contents of the packet are type, length, and value (TLV) fields, which help give IS-IS its flexibility.For IP routing, the contents of the PDU will be IP routing information.

OSI PDUs

PDUs between peers:Network PDU = datagram, packet

Data-link PDU = frame

Examples:



4/24

The equivalent of an OSPF Link State Advertisement (LSA) in IS-IS is the Link State Packet (LSP). During theloading process that occurs immediately after an adjacency is established, the Partial Sequence Number PDU

(PSNP) is used to send summaries of known routes and ask for details of specific routes similar to the function ofthe Data Base Descriptor (DBD) and the Link State Request (LSR) in OSPF. The Complete Sequence Number PDU(CSNP) is similar to the Link State Update (LSU) in OSPF and provides all details of a route.

IS-IS updates are reliable and acknowledged by PSNPs.

The exchange of packets at Layer 2 (data link layer) allows the following to occur:

A neighbor adjacency Creation of a database of reachable networks

The End System Hello (ESH) and Intermediate System Hello (ISH) are not relevant for Integrated IS-IS (IP routing).The Integrated IS-IS adjacencies use the IIH (Intermediate System Intermediate System) hello only.

IS-IS PDUs

IS-IS PDUs are encapsulated directly into a data-

link frame. There is no CLNP or IP header in a PDU.

IS-IS PDUs are as follows:

Hello (ESH, ISH, IIH)

LSP

PSNP (partial sequence number PDU)

CSNP (complete sequence number PDU)



5/24

When only IP routes are carried in the TLV values in the IS-IS control plane packets, it is called Integrated IS-IS.However, for all forms of IS-IS, the Layer 3 adjacency and the Dijkstra SPF tree are based on ISO CLNP addressing,

the general format of which is shown in the graphic. For decoding purposes (even though Cisco does not usethem all as defined by ISO), here is a helpful list:

AFI: Authority and Format Identifier (allows public and private address space) IDI: Individual Domain Identifier for hierarchical addressing (not used for Integrated IS-IS) DSP: Domain Specific Part used by Cisco to identify IS-IS area System I

: Roughly equivalent to host or node address

NSEL: Network Selector byte is used to define Layer 3 protocol in ISO, and is set to 00 for Integrated IS-IS

Integrated IS-IS NSAP Address Structure

The Cisco implementation of Integrated IS-IS distinguishes only thefollowing three fields in the NSAP address:

Area address: Variable-length field (1 to 13 octets)composed of the higher-order NSAP octets, excluding system IDand NSEL.

System ID: ES or IS identifier in an area; fixed lengthof six octets in Cisco IOS software.

NSEL: One octet NSAP selector, service identifier.

Total length of NSAP is from 8 (minimum) to 20 octets (maximum).



6/24

As shown in the graphic, Integrated IS-IS as implemented by Cisco uses its own version of the CLNP ISO addressformat. This is possible since the tree is established for each router in IS-IS, using a special CLNP address on every

router. Not all of the ISO-defined fields are needed for this address.

OSI Addressing: NET Addresses

NSAP address includes NSEL field (process or portnumber)

NET: NSAP with a NSEL field of 0

Refers to the device itself (equivalent to the Layer 3 OSI

address of the device)

Used in routers because they implement the network

layer only (base for SPF calculation)

This address is called the Network Entity Title (NET). Each IS-IS router must have a unique NET and Layer 3adjacencies are established between routers using these addresses.

With this arrangement, the IP routes being carried in the IS-IS packets are leaf objects in the tree. This allowshigh scalability with limited routing traffic. Most IP topology changes will not affect the Connectionless Network

Typical NSAP Address Structure

The simplest NSAP format used by mostcompanies running IS-IS as their IGP is as follows:

Area address (must be at least 1 byte): AFI set to 49

Locally administered; thus you can assign your own addresses.

Area IDThe octets of the area address after the AFI.

System ID Cisco routers require a 6-byte system ID.

NSEL Always set to 0 for a router.



7/24

Service (CLNS) tree.

Link-State Protocol Data Structures

Link-state routers recognize more information

about the network than their distance vector

counterparts.

Neighbor table: also known as the adjacency database

Topology table: referred as the LSDB

Routing table: also known as the forwarding database

Each router has a full picture of the topology

Link-state routers tend to make more accurate

decisions

As with OSPF, IS-IS is a link state protocol and similar data structures are the result. The details of the routingtable creation will be covered later.

Are Adjacencies Established?R2# show clns neighbors

System Id Interface SNPA State Holdtime Type Protocol

R3 Se0/0/1 *HDLC* Up 28 L2 IS-ISR1 Fa0/0 0016.4610.fdb0 Up 23 L1 IS-IS

R2#show clns interface s0/0/1

Serial0/0/1 is up, line protocol is up

Checksums enabled, MTU 1500, Encapsulation HDLC

ERPDUs enabled, min. interval 10 msec.

CLNS fast switching enabled

CLNS SSE switching disabled

DEC compatibility mode OFF for this interface

Next ESH/ISH in 45 seconds

Routing Protocol: IS-IS

Circuit Type: level-2

Interface number 0x1, local circuit ID 0x100

Neighbor System-ID: R3

Level-2 Metric: 35, Priority: 64, Circuit ID: R2.00

Level-2 IPv6 Metric: 10

Number of active level-2 adjacencies: 1

Next IS-IS Hello in 5 secondsif state UP

Notice in the graphic that the neighbors are CLNS neighbors not IP neighbors. The same thing is true of theinterfaces. The IP part comes later.

The adjacency is established over the link to either a Level 1 or Level 2 router (defined later). It is over either anEthernet or point-to-point link.



8/24

The interface details are related to CLNP (CLNS) addresses and attributes rather than IP addresses and attributes.

IS-IS Metric

i s i s met r i c metric [ delay-metric [expense-metric [ error-metric] ] ] {l evel -1 | l evel - 2}

Rout er( conf i g- i f ) #

Configure the metric for an interface; the default is 10.

Metric value is from 1 to 63.

Rout er ( conf i g- r out er) #

met r i c default-value {l evel - 1 | l evel - 2}

Alternately, configure the metric globally for all interfaces.

The primary purpose of any IP routing protocol is to provide the best path to an IP destination address. Bestis expressed as a number called a metric.

The metric is calculated in a different way by each routing protocol. In IS-IS with the Cisco implementation, the IS-IS cost metric is configured manually on each interface. A more complex methodology is shown in the upper partof the graphic but is only used by MPLS providers in their core networks to support Traffic Engineering tunnels.

Integrated IS-IS vs. OSPF:

Area Design

OSPF is based on a central backbone with all otherareas attached to it.

In OSPF the border is inside routers (ABRs).

Each link belongs to one area.



9/24

The graphic depicts the typical OSPF area design, with the backbone area (0) with Area Border Routers (ABRs)connecting other areas of various types. With this arrangement, a LSD is constructed with every OSPF routersending LSAs for internal (intra-area), inter-area, and external routes.

The IS-IS area design is quite different. The area boundary is on the link between routers. The key to this idea isthe definition of Level 1 routers which only build a database within the area and Level 2 routers, which only builda database between Level 2 routers. The equivalent of an OSPF ABR is the IS-IS Level 1/2 routers, which haveboth databases. The end result is much smaller databases and less routing traffic.

There is no layer 0 in IS-IS. The backbone is simply a mesh of links between Level 2 and 1/2 routers.

Integrated IS-IS vs. OSPF: Area Design

(Cont.)

In IS-IS the area borders lie on links:

Each IS-IS router belongs to exactly one area.IS-IS is more flexible when extending the backbone.



10/24

Where OSPF sends five or more LSA types, IS-IS has only Level 1 LSPs and Level 2 LSPs. In each LSP, additionalfunctionality is identified in the LSP header.

The end result is that IS-IS is much more scalable than OSPF with up to 1,000 routers per area, but fewer features.

The graphic depicts the location and role of the Level 1, Level 2, and Level 1-2 routers. The type of router on eachend of a link will determine the type of circuit (Level 1, Level 2) used as well as adjacencies and metric calculationfor best route. (The best route is the one with the lowest sum of all the link metrics in the end-to-end path.)The IS-IS metric must be manually configured on each link.

Advantages of Integrated IS-IS

Supports CLNP and IPMore extensible through TLV design

Routers identified as Level 1, Level 2, or Level 1-2:

Level 1 routers use LSPs to build topology for local area.

Level 2 routers use LSPs to build topology betweendifferent areas.

Level 1-2 routers act as border routers between Level 1 and Level2 routing domains.

IS-IS Link-State Operation



11/24

As mentioned earlier, IS-IS specifications only define broadcast (Ethernet) and point-to-point links. Cisco hasproprietary configurations for other types of WAN designs such as Frame Relay and ATM clouds, which will not

be covered here.

However, the issue of a large number of logical adjacencies when there are multiple IS-IS routers on a singleEthernet still exists. Where OSPF uses the Designated Router (DR) and Backup Designated Router (BDR) to solvethis problem, IS-IS uses a different approach.

With IS-IS, a Designated IS (DIS or Pseudonode) is elected and all routing adjacencies from the other routers aremade to the DIS and it generates LSPs for the group similar to OSPF. There is no need for a backup because anew DIS can be elected in a few seconds if the DIS fails.

Construction of the IS-IS Databases

LSP Representing Routers: LAN

Representation



12/24

LSP Flooding

Single procedure for flooding, aging, and updating ofLSPs.

Level 1 LSPs are flooded within an area.

Level 2 LSPs are flooded throughout the Level 2backbone.

Large PDUs are divided into fragments that areindependently flooded.

Each PDU is assigned an LSP fragment number, starting at 0and incrementing by one.

Separate LSDBs are maintained for Level 1 and Level 2LSPs.

The IS-IS Level 1 and Level 2 databases are constructed during a loading process which is the exchange LSPs. Theprocess is LSP flooding.

Hello PDUs establish the adjacency at Level 2.

LSPs establish the adjacency at Level 3.

Neighbors exchange their databases with PSNP (Partial Sequence Number PDU) which not only provide asummary of reachable networks, but also acknowledge reliable packets.

The CSNP (Complete Sequence Number PDU) is used to provide detailed information about routes during theloading process and also for routing updates during normal operations.



13/24

The graphic shows how each LSP is constructed. It includes information that identifies the type of LSP, ISneighbors (there are no ES neighbors in Integrated IS-IS), and routing peer authentication data.

The graphic shows an example of an IS-IS LSP for Integrated IS-IS. The area number Number of neighbors Number and type of IP internal routes (leaf objects)metric, prefix, and mask Number and type of IP external routes (leaf objects)metric, prefix, and mask

A Link-State Packet Represents A

Router

Router describes itself with

an LSP LSP header contents

include: PDU type, length, LSP

ID, sequence number,remaining lifetime

TLV variable-length fields: IS neighbors

ES neighbors

Authentication information

....

LSP TLV Examples

B = byte

Each set of information, called a tuple, includes a type code, length

field, and value.



14/24

The graphic represents a simple network with area 1 and area 2.

Level 1 and Level 2 Topology Table

R1#show isis topology

IS-IS paths to level-1 routers

System Id Metric Next-Hop Interface SNPA

R1 --

R2 10 R2 Fa0/0 0016.4650.c470

R2#show isis topology



R1 10 R1 Fa0/0 0016.4610.fdb0

R2 --



R1 **

R2 --

R3 35 R3 Se0/0/1 *HDLC*

After the loading process, the IS-IS topology for each router is shown in the graphic. This is the SPF tree for thissmall network. Once Integrated, IS-IS is configured and an IP address and mask are configured on each interface,the IP leaf objects will appear in the routing table.

Routing in a Two-Level Area Structure

Example: OSI Intra-Area and Interarea

Routing



15/24

Level 1, Level 2, and Level 1-2 Routers

Level 1 (like OSPF internal nonbackbone routers): Intra-area routing enables ESs to communicate.Level 1 area is a collection of Level 1 and Level 1-2 routers.Level 1 IS keeps copy of the Level 1 area LSDB.

Level 1-2 (like OSPF ABR): Intra-area and interarea routing.Level 1-2 IS keeps separate Level 1 and Level 2 LSDBs and

advertises default route to Level 1 routers.

Level 2 (like OSPF backbone routers): Interarea routing.Level 2 (backbone) area is a contiguous set of Level 1-2

and Level 2 routers.

Level 2 IS keeps a copy of the Level 2 area LSDB.

The graphic is a review of the IS-IS router level functionality.

To complete the description of IS-IS adjacency and topology construction, the graphic shows connections that willform adjacencies and some which will not.

A Level 1 router may only connect to another Level 1 or Level1/2 router in the same area. To forward packets out of an area, a router must be connected to a Level 1/2 router in the same area.

The backbone area consists of connected Level 2 and Level 1/2 routers.

Example: IS-IS Adjacencies



16/24

IS-IS tracks information using these datastructures:

Interface table: Active interfaces

Neighbor table: Discovered neighbors

Link-state database: Network topologyinformation Level 1 and Level 2

Routing information base: Results of the SPFcalculation (After integrated IS-IS is configured)

IS-IS Data Structures

IS-IS Neighbors

So far, the data structures shown in the graphic have been established. Once Integrated, IS-IS and IP addressingare present on every interface, the Routing Information Base (RIB) will show up.

The Configuration and Verification of Integrated IS-ISThe configuration of IS-IS on Cisco devices does not require in-depth understanding of all the underlyingmathematical process that make IS-IS work, so the commands are simple. Verification and troubleshooting can bea bit more challenging, however. The skill is in understanding the output of verification commands.



17/24

Building an IP Routing Table

Partial Route Calculation (PRC) is run tocalculate IP reachability.

Since IP and ES are represented as leaf objects,

they do not participate in SPF.

Best paths are placed in the IP routing table

following IP preferential rules.

They appear as Level 1 or Level 2 IP routes.

The graphic shows that the IP routing table (RIB) is constructed by a Partial Route Calculation (PRC), which meansa change in IP topology does not affect the CLNS created IS-IS tree and area structure.

IP routes to IS-IS are like IP host routes (/32) in IP. They are either there or they are not. If one fails, the rest ofthe routers in the area need to know and how far the change will be propagated, which depends on proper IPsummarization techniques, just like all other routing protocols.

Routes will appear as Level 1 (intra-area), Level 2 (inter-area), or external.



18/24

Integrated IS-IS Configuration Steps

1. Define areas, prepare addressing plan (NETs) for routers, and determine

interfaces.

2. Enable IS-IS on the router.

3. Configure the NET.

4. Enable Integrated IS-IS on the appropriate interfaces. Do not forget

interfaces to stub IP networks, such as loopback interfaces (although

there are no CLNS neighbors there). The addresses must still be

advertised

The steps to configure Integrated IS-IS are shown in the graphic. There is no network command in IS-IS and IS-IS isimplemented on each interface with an interface command.

Step 1: Define Area and Addressing

Area determined by NET prefix:

Assign to support two-level hierarchy.

Addressing:

IP: Plan to support summarization.

CLNS: Prefix denotes area. System ID must be unique.

As mentioned earlier, each IS-IS router must have a CLNS NET address. IP address summarization should beconfigured when possible at area boundaries (Level 2 and Level 1/2) to minimize routing traffic.



19/24

Step 2: Enable IS-IS on the Router

router i s i s [ area-tag]

rout er( conf i g) #

Enable the IS-IS routing protocol.

area-tagname for a process

When routing of CLNS packets is also needed, use the clns routingcommand.

CLNS routing is disabled on the Cisco IOS by default. It is not needed if only IP is being routed. There is also noneed for the IS-IS area-tag argument with Integrated IS-IS since there are not typically multiple IS-IS processes.

Step 3: Configure the NET

net network-entity-titleRout er( conf i g- rout er ) #

Configure an IS-IS NET address for the routing process.

The graphic depicts the IOS syntax to configure the NET.



20/24

Step 4: Enable Integrated IS-IS

i p router i s i s [area-tag]

router(conf i g- i f ) #

Includes an interface in an IS-IS routing process

Once an interface has an IP address and mask, it can be configured to use IS-IS for IP routing purposes asdescribed in the graphic.

Simple Integrated IS-IS Example

interface FastEthernet0/0

ip address 10.1.1.2 255.255.255.0

ip router isis

!

interface Serial 0/0/1

ip address 10.2.2.2 255.255.255.0

ip router isis

!

router isis

net 49.0001.0000.0000.0002.00

The configured router acts as an IP-only Level 1-2 router.

A simple Integrated IS-IS network is shown in the graphic. Note that this router is in area 1 and no IS-IS area-tagargument is used. Unless otherwise designated, this router acts as a Level1/2 router by default.

Routers can be configured as Level-1, Level -1-2, and Level-2 only.

Links between routers may be configured as IS-IS circuit-type Level-1, Level-1-2, or Level-2 only.



21/24

A more complex configuration is depicted in the graphic.

Router level designation Circuit level designation Interface metric

On Level 2 and 1/2 routers IP address summarization is configured as a router sub-command, not on theinterface.

Example: Is Integrated IS-IS Running?

R2#show ip protocols

Routing Protocol is "isis"

Invalid after 0 seconds, hold down 0, flushed after 0

Outgoing update filter list for all interfaces is not set

Incoming update filter list for all interfaces is not set

Redistributing: isis

Address Summarization:

None

Maximum path: 4

Routing for Networks:

FastEthernet0/0

Loopback0

Serial0/0/1

Routing Information Sources:

Gateway Distance Last Update

10.10.10.10 115 00:00:02

10.30.30.30 115 00:00:03

Distance: (default is 115)

R2#

Displays the parameters and current state of the active routingprotocol processes

Router#

show ip protocols

Tuning IS-IS Configuration

Change router type on

R1 and R3

Change interface

levels on R2

Change metric on

S0/0/1



22/24

The configuration can be verified with the show ip protocolscommand. Now it is starting to look more like theIP routing in OSPF and EIGRP. The administrative distance for IS-IS is 115.

Example: Are There Any IP Routes?

R2#show ip route isis

10.0.0.0/24 is subnetted, 5 subnets

i L2 10.30.30.0 [115/45] via 10.2.2.3, Serial0/0/1

i L1 10.10.10.0 [115/20] via 10.1.1.1, FastEthernet0/0

R2#

Displays the current state of the routing table

show ip route [address [mask]] | [protocol [process-id]]

router#

Show ip routeprovides additional verification by the presence of IP routes in the routing table.



23/24

Completing the Remaining IS-IS Data Structures

IS-IS: DATA STRUCTURES

Control Plane

Routing Protocol

IP Routing Table (RIB)

IP Forwarding Table (FIB)

Exchange of

Routing Information

CEFData Plane

Stored in fast

memory cache or

hardware

The RIB is maintained as part of the routing process memory by incremental routing updates after the initialloading process. This exchange of information necessary to maintain this table is called the control plane. SpecificLSPs cause routes to appear in the IS-IS routing table as follows:

i Level 1

i Level 2

i - External

Once the tables are converged and stable, it is necessary to deal with changes in network topology (link failures,device failures), very quickly. With all of the real-time protocols in modern networks (voice, video, multi-cast),repair times are measured in milliseconds. Over the years, Cisco developed multiple features in the Cisco IOS toaddress this issue.

The latest in this evolution is Cisco Express Forwarding (CEF), which copies all of the contents of the RIB (routingtable) to a fast cache in the fastest memory location possible, such as a line module in a 6500 or Nexus multi-layer switch. This table is called the Forwarding Information Base (FIB) and is in the data plane. It is the tableactually consulted by the router when user data packets arrive on an interface and a routing decision is required.

The FIB is refreshed by the RIB regularly. The only time the RIB is consulted is if there is a failure of the FIB toprovide a routing entry.



24/24

The following is a list of additional IS-IS troubleshooting and verification commands:

Show clns Show clns protocol Show clns interface

Show clns neighbors

Show isis route Show clns route

Show isis database Show isis topology

Additional IS-IS Features Not Described in Detail in this

Paper

IS-IS authentication

Passive interfaces Design limitations Issues with NBMA and multipoint WANs IS-IS over MPLS

Route Redistribution Route filtering Troubleshooting

Route Summarization IS-IS design

Learn MoreLearn more about how you can improve productivity, enhance efficiency, and sharpen your competitive edge

through training.

ROUTE - Implementing Cisco IP Routing v1.0

ARCH - Designing Cisco Network Service Architectures v2.1

SWITCH - Implementing Cisco IP Switched Networks v1.0

TSHOOT - Troubleshooting and Maintaining Cisco IP Networks v1.0

Visitwww.globalknowledge.comor call 1-800-COURSES (1-800-268-7737)to speak with a Global Knowledgetraining advisor.

About the AuthorRay Dooley, BS, MBA, CCSI, CCNA, CCNP, CCDA, CCDP, SE, FE, has been a network professional in several

capacities for over 30 years. He is the Global Knowledge Course Director for CCDA, ARCH, SWITCH, ROUTE,TSHOOT, and ICMI. He has done course development for Global Knowledge, Cisco Systems, and GE.
http://www.globalknowledge.com/training/course.asp?pageid=9&courseid=12859&country=United+Stateshttp://www.globalknowledge.com/training/course.asp?pageid=9&courseid=17254&country=United+Stateshttp://www.globalknowledge.com/training/course.asp?pageid=9&courseid=12863&catid=206&country=United+Stateshttp://www.globalknowledge.com/training/course.asp?pageid=9&courseid=12861&catid=206&country=United+Stateshttp://www.globalknowledge.com/http://www.globalknowledge.com/http://www.globalknowledge.com/http://www.globalknowledge.com/http://www.globalknowledge.com/training/course.asp?pageid=9&courseid=12861&catid=206&country=United+Stateshttp://www.globalknowledge.com/training/course.asp?pageid=9&courseid=12863&catid=206&country=United+Stateshttp://www.globalknowledge.com/training/course.asp?pageid=9&courseid=17254&country=United+Stateshttp://www.globalknowledge.com/training/course.asp?pageid=9&courseid=12859&country=United+States

Documents

WP CI FourElement of is-Is