IP,IPv6 Routing Protocols, Internet protocols version six, IPv6

Source :

Deploying IPv6 Networks

By Ciprian Popoviciu, Eric Levy-Abegnoli, Patrick Grossetete

...............................................

Publisher: Cisco Press

Pub Date: February 10, 2006

IPv6 Routing Protocols

Numerous IPv4 routing protocols (RPs) are available for finding routes between networks, and almost every one of them has an IPv6 correspondent or extension: Routing Information Protocol next-generation (RIPng), Open Shortest Path First version 3 (OSPFv3), Intermediate System-to-Intermediate System (IS-IS), Enhanced Interior Gateway Routing Protocol (EIGRP), and Border Gateway Protocol (BGP). So far, IPv6 has brought few innovations to the IP routing paradigm. There are still interior gateway protocols (IGPs) and exterior gateway protocols (EGPs), distance vectorbased and link-state-based routing protocol algorithms, and so on.

The concept of the autonomous system, defined as a set of networks controlled by a common administrative entity, remains unchanged with the introduction of IPv6 RPs. The same autonomous system (and autonomous system number [ASN]) will route both IPv4 and IPv6. IPv6 IGPs, used to exchange routes within the autonomous system, are namely RIPng, OSPFv3, IS-IS for IPv6, and EIGRP for IPv6. Only BGP4 is available to exchange IPv6 routes between autonomous systems. Multiprotocol extensions provide support in BGP4 for IPv6 routing.

The requirements for IGPs and EGPs are quite different, in terms of routing table size, number of supported routers, convergence time, security, routing policy, and so forth. For that reason, they use different algorithms and mechanisms, which also affect the type of information they exchange and store. IGPs use distance vector and link state, whereas BGP uses the path vector RP algorithm. The following table represents RP taxonomy, and highlights their IPv6 correspondent. For more details on how to choose the RP, refer to Top-Down Network Design, Second Edition, by Priscilla Oppenheimer.

Table 4-1. Taxonomy of Routing Protocols

Deployment Domain	Algorithm	RP	Scalability	Convergence Time	Metric	IPv6 Version
Interior gateway protocol (IGP)	Distance vector	RIP	15 hops	Slow	Hop count	RIPng
		EIGRP	1000s routers	Quick (via DUAL algorithm)	Bandwidth, delay, reliability, load	EIGRP for IPv6
	Link state	OSPF	1000s routers (100s/area)	Quick (via LSAs and HELLO)	Cost (function of bandwidth on Cisco routers)	OSPFv3
		IS-IS	1000s routers (100s/area)	Quick (via LSPs)	Configured host, delay, expense	IS-IS for IPv6
Exterior gateway protocol (EGP)	Distance vector	EGP			Integer <=255
	Path vector	BGP AF=IPv4	1000s routers	Slow (via UPDATE)	Function of path attributes and other configurable factors	BGP AF=IPv6

The rest of this section briefly reviews existing unicast RP technologies. The next section reviews each of the available IPv6 RPs and provides configuration examples. Then the last two sections cover the topic of multihoming and deployment aspects, respectively.

IP Mobility, IP, IPv6, IP version 6, Internet protocol

Source :

Deploying IPv6 Networks

By Ciprian Popoviciu, Eric Levy-Abegnoli, Patrick Grossetete

...............................................

Publisher: Cisco Press

Pub Date: February 10, 2006

IP Mobility

The Internet has become so pervasive that no matter where you are, you can plug your computer into a wall, or attach to a wireless LAN, and, after a while, you will be able to communicate. Is not this mobility? Well, not quite.

That type of "mobility" is achieved by getting a new IP address within the network of attachment and losing all sessions bound to the previous IP address. This might be acceptable for corporate users moving from work to home, but can be much more cumbersome for road warriors, and it can be a showstopper for IP telephony.

Mobile IP provides a network layer for hosts that enables them to maintain the same IP address no matter where they are in the Internet, and keep receiving traffic as they move.

"Advanced ServicesIPv6 Mobility," MIPv6 is compared to MIPv4. Even though MIPv4 is a mature and deployable technology, it faces limitations because of the nature of IPv4. At the same time, IPv6 mobility is considered as one potential enabler for IPv6. The number of IP-enabled devices and the need for any-to-any communications among them is driving requirements that IPv4 cannot easily satisfy, and it is opening opportunities for IPv6. By integrating functionalities designed for Mobile IPv4 into standard IPv6 protocols, and by leveraging existing IPv6 capabilities, MIPv6 has built up a MIP model that is much more compelling than its IPv4 counterpart.

It must be noted that enhancements to mobility are largely taking place in IPv6 related working groups, even though a fraction gets retrofitted into the IPv4 standards. Although MIPv6 has benefited greatly from its MIPv4 parent, it is now the driver of the evolution of IP mobility, and it is widely expected to be a foremost steering force for IPv6 deployments.

In terms of deployment, it must be considered that IP mobility enables new flows, which impact the wireless infrastructure: Telephony over IP demands a higher level of coverage, latency, and QoS enforcement, whereas peer to peer imposes always-on reachability and multimedia capabilities.

The application of the MIP and NEtwork MObility (NEMO) standards is not limited to hosts and routers that actually roam around the Internet as a usual behavior. Sales of consumer routers are plummeting. At the moment, they are related to IPv4 NAT operations. With IPv6, it can be expected that people will deploy unmanaged yet globally addressable networks at home. NEMO support by the home gateways would enable a service provider to deploy preprovisioned devices, and could save hundreds of thousands of network-renumbering operations per year as customers move from one home to the next.

At the core, MIP builds dynamic tunnels, and NEMO exchanges routes over those tunnels. In a way, this is a revamping of the traditional model of the core where BGP routers exchange the bulk of the Internet routes over peering tunnels. But whereas the model of the Internet is designed for fixed, aggregated routes that are locally injected and slowly distributed throughout its fabric, MIP and NEMO techniques enable a new model where routes are projected where and when they are needed, on-demand; this opens to a new level of hierarchy for the fine-grained mobile routes, and a new order of scalability for the Internet.

But the Internet of today is not fully ready for IP mobility. Even if IPv6 can exist over an IPv4 fabric as a transitional method, a significant number of improvements must be made to cope with the latency of the protocol and enable multimedia interactive applications such as voice calls and video.