Is-is protocol Basic Concept-1

Technorati Tags: ccie,isp, operator, Isis,is-is

1, area, Area

The OSPF area----0, the backbone area. 1, non-backbone area.

In Isis, the area is no longer determined by number.

--isis allows the entire routing domain to be divided into multiple zones.

-- A router currently has a maximum of 254 area IDs. Configure different zone IDs for smooth zone merging, segmentation, and conversion.

Unlike OSPF, a route must belong to an entire region, not some interfaces belonging to one region, and other interfaces to another. (and BGP a bit similar)

Here is an example of ISIS's typical networking application:

A total of 4 areas, the area 1,2,3,4, each router belongs to one area completely.

ISIS no longer uses zone numbers to differentiate between backbone and non-backbone areas. So there's no area0 to say.

The region division of ISIS is divided between routers.

650) this.width=650; "Style=" background-image:none;border-bottom:0px;border-left:0px;padding-left:0px; padding-right:0px;border-top:0px;border-right:0px;padding-top:0px; "title=" clip_image002 "border=" 0 "alt=" clip_ image002 "src=" http://s3.51cto.com/wyfs02/M00/73/39/wKioL1X4AaazlT6xAAEh5CIu3as754.jpg "height=" 348 "/>

2, node Nodes and the level of (level)

In OSPF, the node names are: DR, BDR, ABR, ASBR, and so on.

In Is-is, the node name is as follows;

Nodes:

Level-1 Routers (L1 router )--- non-backbone zone routers

For the Level1 router, it is located inside the common area.

█L1 routers only form adjacency with Level1 routers within the region (or L1/L2 with L1 functionality).

The █L1 router has only Level1 link databases (L1 LSDB) within the region. Contains routing information for all L1 routers in the region. A bit like Type-1, router LSA.

█ generates a default router pointing to this device as an egress router (the L1 area is similar to an OSPF stub area) by generating the L1/L2 router (ATT bit=1) with its nearest

█ when forwarding, if the destination address in the region, the direct use of L1 LSDB generated route forwarding messages, if the destination address is not in the region, the region's closest L1/L2 router as the egress of the off-zone network, which may lead to suboptimal routing.

LEVEL-2 routers (L2 router)---backbone area routers

The L2 router is a router located in the backbone area.

Can form a neighbor relationship with other L2 (or L1/L2 routers with L2 capabilities)

The L2 router has a LEVEL2 link-state database (L2 LSDB) that contains all the inter-zone routing information.

Receives messages from other L2 routers in the region and forwards the messages to the L2 routers in other regions (or L2 routers that are forwarded to the same region) at the destination address. Receives messages from L2 routers in other regions and forwards the messages at the destination address.

level-1&2 Routers (l1/l2 router ): ABR

Usually located on a border router, similar to an OSPF ABR.

You can form a neighbor relationship with any level of router in the region. You can form L2 neighbor relationships with L2 or L1/L2 routers that are adjacent to other regions.

There may be two levels of link-state databases.

The L1 is used as an intra-zone router. L2 is used as a route between regions

To complete the exchange of routing information between his area and backbone, and to assume the responsibility of L1 L2.

Attention:

If a l1/l2 router forms an adjacency with a router in another region, it will advertise the L1 router in the region. He has a total export of the region.

The specific method is to send the ATT (Attachment) Bit=1[att bit in the message to all other L1 neighbors in the region when generating the L1 LSP (link state Protocol Data unit) of the region.

Area hierarchy:

LEVEL-1 area: Includes intra-region routers, including L1 and L1/l2-router

Level-2 area (Backbone, Backbone): Includes L2 and L1/l2-router.

Here is an ISIS topology diagram.

On the division of the area number:

L1/L2 are regional border routers. The middle green is a backbone area.

Noted in the figure. Routers in the backbone area can also belong to different areas.

So in Isis, it is important that the definition area is no longer an area number, but rather a l1/l2 as the network boundary. is also one of the distinctions with OSPF.

650) this.width=650; "Style=" background-image:none;border-bottom:0px;border-left:0px;padding-left:0px; padding-right:0px;border-top:0px;border-right:0px;padding-top:0px; "title=" clip_image004 "border=" 0 "alt=" clip_ image004 "src=" http://s3.51cto.com/wyfs02/M01/73/39/wKioL1X4AaeSdF6QAAE26hxsDM8242.jpg "height=" 331 "/>

3, the level of ISIS.

The is-is consists of two levels:

▲level2: Backbone Backbone (a collection of continuous Level2 routers); The backbone is made up of all L2 (including L1/L2) routers, and the backbone must be contiguous. (not say there is a L1 in the middle of the two L2 to separate the argument)

Note: Is-is's backbone is not a specific area and must be separated from OSPF .

▲level1: The concept of a single region, consisting of Level-1 routers in the region, will be automatically published to backbone by other router information . L2 is absolutely not going to publish the routing information into L1. L1 all routers have only one egress gateway, which is l1/l2. L1 if you want to access other zones, there is a default route pointing to the L1/L2 router. So why say L1 is like the stub area of OSPF]

Note: A is-is routing domain (routing domain) does not necessarily require two levels, and if only one area is deployed, it may be all L1, or all L2, and L2 is recommended for better scalability.

4, Dis and pseudonodes (pseudo node)

' DIS ' means ' designated is ', similar to the DR in OSPF.

▓dis similar to the DR in OSPF

▓ in broadcast multi-access, a router is elected as dis (DIS and OSPF dr function is the same)

▓ the point-to-point network, like OSPF, does not need to elect a Dr.

▓ different levels have different dis, L1 level broadcast network election L1 level dis. The L2-level broadcast network elects the L2-class dis. The result of the election may not be the same as.

unlike OSPF, ▓ 's elections are preemptive, there is no backup dis in the Is-is, and when a dis is not working, the other one is directly elected.

The process of electing Dr in OSPF is:

-the first OSPF router to be launched will be elected Dr.

--Compare priority, 0-255, default = 1, point to dot = 0, indicate not to participate in Dr Election. The bigger the higher the priority.

--Router Router-id, the larger the interface address, will be elected as Dr.

▓ because there is no concept of bdis, so once a device is elected for DIS, the time interval for the dis to send hello packets is 1/3 of the normal router, so that the dis can be detected quickly when it fails. To achieve the same effect as the BDR in OSPF.

The role of Dis:

--Create and advertise the pseudo-node LSP (link state Protocol data Unit link status data units) to all routers in the broadcast subnet.

--Flood the LSP (link state Protocol data Unit link status Data unit) in the LAN by periodically sending csnp[full database description per 10s.

The election process of dis: (similar to the Dr Election principle of OSPF)

In a LAN (broadcast network), a router must be elected to be dis.

--election based on interface priority (Cisco default is 64)

--If all interface priorities are the same, the router with the largest subnetwork point of attachment (SNPA) will be selected for DIS.

----LAN, SNPA refers to the MAC address.

----in a Frame relay network, SNPA refers to the DLCI number.

The DIS is an election that is preemptive and does not provide for a mechanism for not participating in the election. (meaning that even if you change the priority to 0, he will still participate in DR's election)

This article is from the "hank--me, Focus on Principles" blog, be sure to keep this source http://361531.blog.51cto.com/351531/1695073

Is-is protocol Basic Concept-1