Improvement of the routing protocol in IPv6

In IPv6, we can experience new performance. Now let's talk about the IPv6 Domain Routing Protocol in detail. First, let's take a look at the routing protocol types in the network. At present, the IP domain routing protocol mainly uses IS-IS and OSPF. The version of OSPF used to exchange IPv4 routing information IS-IS used to exchange IPv6 routing information IS called OSPFv3; IS-IS used to exchange IPv4 routing protocol version IS-IS, the version used to exchange IPv6 routing protocol is called IS-ISv6.

(1) differences between OSPFv3 and OSPFv2

OSPFv3 (RFC2740) and OSPFv2 (RFC2328) have no fundamental difference in principle. OSPFv3 still uses the LSA database in link state and maintains synchronization between adjacent routers. However, due to the change in the address length from IPv4 to IPv6, in order to support the IPv6 address format, OSPFv3 has made many modifications to the OSPFv2 protocol. From the perspective of the routing protocol standardization process, the OSPFv3 protocol is mature and has a fixed RFC2740 protocol. OSPFv3 improves versatility and enables the network to adapt to changing requirements. This simplifies the complex network and takes some enhancement measures to ensure easy upgrade. OSPFv3 is also optimized and the security is also improved.

The main purpose of OSPFv3 is to develop a routing protocol independent of any specific network layer ". For this purpose, C) st'fv3's internal router information is re-designed. Different from previous versions, () SFFv3 does not insert IP-based data to the header at the starting position of the packet and link status announcement (LSA. C) st'fv3 uses computer network protocol-independent information to execute key tasks that require IP header data in the past, such as identifying the LSA that publishes route data.

In addition to modifying header data, OSPFv3 also redefined the role of LSA. In OSPFv3, tasks that publish network topology and IPv6 data are assigned to new and existing LSA instances.

The OSPFv3 Routing Protocol adds multiple optional functions, such as multicast OSPFv3, to achieve versatility. To achieve this goal, OSPFv3 has extended the data domains used by network devices to publish enabling function options. Most OSPFv3 vro information includes the option domain. devices running OSPFv3 can support up to 24 optional functions, whereas earlier versions only support 8 functions.

To simplify the construction of Complex Fault-Tolerant networks, OSPFv3 introduces the InstanceID and R-bit options. As a component of each OSPFv3 packet header, InstanceID no longer relies on the complex authentication scheme or access list required in the past to control the communication between vrouters in the shared physical network and OSPF domain. In addition to InstanceID, OSPFv3 can also use R-bit to make the final system of the server have effective redundancy.

The difference between OSPFv3 and the previous protocol is that it simplifies the message structure by providing non-inherent security. OSPFv3 messages can be authenticated and encrypted by using the Security packet header Integration System of IPv6 packets. In the past, it was necessary to add independent and complex protocols.

The OSPFv3 routing protocol provides more powerful functions and is widely used to easily support new computer network protocols. The new features simplify network devices and operations, and the upgrade will not be so troublesome when OSPFv3 is used.

(2) The difference between IS-ISv6 and IS-ISv4

Another connection status Protocol widely used by carriers is isforipv6. We know that isis is a standard routing protocol (ISO/IEC10589) that was originally used to support dynamic protocols of CLNS networks. Because the design of ISIS is very conducive to the expansion of new functions, it first expands the IPv4 routing protocol (RFC1195). Some people refer to this Protocol, which can serve both CLNS and IP, as IntegratedISIS.

In a similar way, ISIS can also process IPv6 routing information through simple extensions. Draft sow standard.

IPv6. only the TLV related to IPv6 is added, And the ISO10589 and RFCl195 modes are basically maintained in the creation and maintenance of the neighbor database and Topology database.

Therefore, CLNSIFV4 and IPv6 have the same topology. Also, draft-ietf-isis-00006.txt requires that the IPv4 and IPv6 networks completely overlap. We call this implementation an ISISfor IPv6 single topology mode.

Soon, we realized that the single topology mode's requirement for completely overlapping IPv4 and IPv6 networks limits the deployment of IPv6 networks. Obviously, the scale and coverage of IPv6 will be larger than the previous generation of IPv4 protocol networks, and some networks will only have IPv6 attributes. For example, some networks only have IPv6 addresses but not IPv4 addresses.

At this time, the ISISforIPv6 single topology mode may cause some IPV4 data packets to be mistakenly forwarded to this part of the IPv6 network, resulting in routing confusion and trouble, and cannot meet the needs of different IPv4 and IPV6 expansion ranges.

The ISISfor IPv6 multi-topology (draft-ietf-isis-wg-multi-topology) is a solution to this problem, it uses different IPv4 IPv6 topologies to remove the restrictions that must be consistent between the two networks. In the multiple topology modes, different topology databases are created for IPv4 and IPV6 networks, and the spf Shortest Path priority algorithm is calculated respectively to provide an independent routing subsystem for IPv4 and IPv6, this frees IPv6 network construction from IPv4 restrictions, opening up the development space for IPv6 network growth.