Identifying Systems and Establishing IS-IS Adjacencies


As you now know, an IS can be part of only one Level 2 area (as shown in Figure 6.2). It is the area ID and address that specifically identifies the area and system ID of each network node. Therefore, every IS in a particular area must use the same identifying area address (just as nodes in an IP subnet use the same subnetwork address). An adjacency is created between a host (end system) and a Level 1 router (IS) if they both have an identical area address. ESs create adjacencies only with ISs that have the same area addresses and with other ESs on their same subnetwork. Routing performed at Level 1 is based on SysID values ( generally MAC addresses or an IP loopback address coded into the ID); therefore ES and IS devices must have unique SysIDs (as well as the uniform length) within the area. This is a mandatory 6 bytes when you're running the Cisco IOS.

Because all L2 ISs must be aware of the other ISs in the backbone, they need unique SysIDs as well (refer to Figure 6.3). As a matter of practice, you should ensure that the SysIDs are unique on a domainwide basis to prevent the possibility of conflicts. Referring back to Table 6.2, the SPNA number is the subnetwork point of attachment or Layer 2 mapping to the NET or NSAP address. This is something like a MAC address or DLCI number. A circuit ID is also used to identify the individual interfaces. This could be an 8-byte integer number on a WAN link such as 15, or an identifier attached to the end of the SysID of the DIS, such as 1820.5700.1003.15 where 15 represents the circuit ID number.

Figure 6.3. NSAP addressing on a physical IS-IS network.

As you can see in Figure 6.3, all the ISs are in the same IS domain with the ID number of 40. Each IS also has a unique 3-byte area ID that corresponds to the area in which they are (Area 50, Area 51, and Area 52) and is shared by the other ISs within their respective areas. Third, each IS has a 6-byte SysID that is unique throughout the domain ( 1111.1111.1111 to 8888.8888.8888 ). Finally, each IS has an NSEL value of 00, which specifies this as a NETs.

Table 6.2 also defined the protocol data unit or PDU as the unit of data implemented by OSI. IS-IS and ES-IS data packets are encapsulated by a Data Link (Layer 2) PDU without an IP or CLNP header. Each IS-IS and ES-IS PDU contains the Type, Length, and Value fields (also referred to as TLV). The CLNP data packets, however, have a complete CLNP header placed between the Layer 2 header and the upper-layer CLNS routing information. Table 6.5 shows the four existing L1 and L2 packet types.

Table 6.5. IS-IS L1 and L2 Packet Types

PDU Packet Type

Description

LSP

The link-state packet used to transport link-state information. LSP can be Broadcast (that is, LANs) or Nonbroadcast (that is, WAN point-to-point and multipoint ”IS-IS does not understand NBMA networking topologies). LSP contains LSP header and variable-length TLV fields.

Hello

These Hello PDU packets are used to generate and manage adjacencies. ESHs hello packets are sent by an ES to an IS (ES to IS Hello), ISH hello packets are sent by an IS to an ES (IS to ES Hello), and IIH hello packets are used between two ISs (IS to IS Hello).

CSNP

This is the complete sequence number PDU that is utilized to propagate the IS's complete link-state topology database.

PSNP

This is the partial sequence number PDU that is utilized for link-state acknowledgements and requests .

On LANs, IS-IS protocol data units are forwarded to the following well-known MAC layer broadcast addresses:

AllL1ISs at 01-80-C2-00-00-14

AllL2ISs at 01-80-C2-00-00-15

AllIntermediateSystems at 09-00-2B-00-00-05

AllEndSystems at 09-00-2B-00-00-04


IS-IS routers (ISs) use LSP sequencing to make sure that routers are getting the most up-to-date information for calculating routes. The LSP also has a Remaining Lifetime field to make sure that invalid or old LSPs get dropped from the database after a default interval of 20 minutes (1200 seconds). LSPs also store precise information about the ISs' attached peers. For example, an LSP knows the links to neighbor ISs and the interface metrics, as well as the links to end system peers.

IS-IS Metrics

One required metric and three optional metrics are associated with IS-IS. Regardless of the ones that you implement, they are all applied to the outbound interface of the peer IS. The default, required IS-IS metric value is cost. An interface cost is a 6-bit value between 1 and 63 and has a default value of 10. The total cost (total path metric) to a destination router is the sum of all the costs on all outbound interfaces on a precise path with the least cost being preferred. The maximum value is 1023, although Cisco supports a "wide metric" of 16,777,215 (2 24 “1) with a total path value of one less than 2 32 .

IS-IS LAN Adjacencies

You were also introduced to the term designated intermediate system (DIS) in Table 6.2. All routers on an IS-IS LAN create adjacencies through the elected DIS router. The DIS is also called a pseudonode because it creates a logical link to all the other nodes on the LAN. If the DIS fails, another election takes place because there is no backup DIS such as the OSPF BDR. The election is based on router priority with the MAC address as the last resort tiebreaker. The DIS creates and floods a new pseudonode LSP when a new neighbor is added or dropped or when the refresh interval timer expires. Each IS (router), on the other hand, generates and floods a new non-pseudonode LSP (non-DIS) when a new neighbor comes up or disappears, when new IP prefixes are added or removed, when the metric of a link changes, or when the refresh interval timer expires . This DIS pseudonode process works in concert with the IS LSPs to lower the number of adjacencies and the amount of flooding over LAN networks.

Separate adjacencies are created for Level 1 and Level 2 ISs and are stored in their own particular tables. On a LAN, these adjacencies are created with L1 and L2 IIH hello packets. The hello packet default interval is 10 seconds and the down timer is 30 seconds. Both of these timers are configurable. These separate IIH packets advertise L1 and L2 peers and establish adjacencies based on the inbound IIH and router type (L1L2 or L2). ISs within an area accept L1 IIH packets and establish adjacencies only within their own area ISs. Level 2 ISs or L1L2 routers accept Level 2 IIH packets and establish Level 2 adjacencies only.

IS-IS WAN Adjacencies

On a point-to-point connection, common IIH packets are used to advertise the level type and area ID information. For instance, two Level 1 routers connected by a WAN link in the same area would exchange IIH packets that designate Level 1 (level type) and establish a Level 1 adjacency. Level 2 ISs in the same area connected by a point-to-point link would designate a Level 2 type and establish a Level 2 adjacency. If two L1L2 ISs were in the same area, they would use a common IIH that would designate both types and adjacencies. In a nutshell , Level 1 ISs establish only Level 1 adjacencies, Level 2 ISs establish only Level 2 adjacencies between areas, and L1L2 ISs establish both.

If two routers were physically connected but in different areas, the IIH PDUs would be mutually ignored and no adjacencies would be established because the area IDs are inconsistent.




Cisco BSCI Exam Cram 2 (Exam Cram 642-801)
CCNP BSCI Exam Cram 2 (Exam Cram 642-801)
ISBN: 0789730170
EAN: 2147483647
Year: 2003
Pages: 170

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net