Integrated Intermediate System to Intermediate System

Integrated Intermediate System
to Intermediate System
Typically when a link-state routing protocol is talked about, OSPF is the routing protocol that
is being referred to. There is another link-state routing protocol, known as
Intermediate System
to Intermediate System (IS-IS)
. IS-IS was developed by Digital Equipment Corporation as
an International Standards Organization (ISO) protocol to route
Connectionless Network
Services (CLNS)
, which is a Network layer protocol of the Open Systems Interconnection
(OSI) suite of protocols.
IS-IS was being developed by ISO at roughly the same time OSPF was being developed by the
Internet Architecture Board (IAB). Many years ago, industry experts believed that the OSI suite
would eventually replace TCP/IP. With this in mind, it was proposed that IS-IS become the
Internet Engineering Task Force (IETF) recommended standard for routing TCP/IP. An extension
was added to IS-IS to allow the simultaneous routing of both IP and CLNS. This extension
became known as Integrated IS-IS. Integrated IS-IS can route in either a CLNS environment, an
IP environment, or an environment made up of both.
Integrated Intermediate System to Intermediate System
207
Numerous battles were fought over whether OSPF or Integrated IS-IS would become the recommended
standard routing protocol for IP. When the war was over and the smoke had cleared,
OSPF had won.
So you may be wondering why you need to learn IS-IS if OSPF became the recommended
standard for IP. IS-IS is still implemented in numerous service provider backbones. With
MPLS starting to make a presence, it is certain that at some point in your career you will
encounter IS-IS. Remember that the only routing protocols that can be used for MPLS traffic
engineering are IS-IS and OSPF.
Because you have a good understanding of OSPF, learning IS-IS shouldn’t be that much of
a challenge. The two routing protocols share many of the same basic concepts:

Both of them are link-state routing protocols.

To maintain their link-state databases, both routing protocols use the Dijkstra SPF
algorithm.

Both, through the use of areas, support a hierarchical network topology.

They both use Hello packets to form adjacencies with their neighbors.

For broadcast multi-access networks, they both elect a designated router (DR).

They both support VLSM and the summarization of areas.

Both allow the use of authentication to ensure a more secure network.

Both allow multiple instances per device, OSPF with the process ID and IS-IS with a tag
after the
router isis
command.
Although IS-IS and OSPF share many common features, they do have quite a few differences:

Whereas OSPF routers can be part of multiple areas, an IS-IS router belongs to only one
area per routing process.

In OSPF, the boundaries of areas are set in the router. The boundaries of areas are on the
network connections between routers for IS-IS, reiterating that each router is in only one
area per routing process.

IS-IS utilizes CLNS protocol data units (PDUs) to send information between routers instead
of using IP packets, like OSPF does.

IS-IS allows for the preempting of DRs, where OSPF does not.

OSPF DROthers do not form adjacencies with other DROthers on broadcast multi-access
networks, while in the same environment, all IS-IS intermediate systems form adjacencies
with one another.

The backbone of an IS-IS network is designated by the type of routers in it instead of being
designated by an area number (0, in the case of OSPF).
Now that you know a little about the history of IS-IS and how IS-IS compares to OSPF, let’s
focus on its operation.