Comparison of IPV4 and IPV6
- The shortage of addresses is the main motivation for upgrading to IPV6.
- IPV4 has only 32Bits address length and only about 4 billion available addresses.
- The main reason for insufficient addresses is that the allocation of addresses is extremely unreasonable, and relatively few are allocated to developing countries such as China.
- As of November 2019 in China (data source network, doubtful), there are only about 2502 IPV4 addresses, less than 200 class B address pools, and none of class A addresses.
In order to adapt to the complex network type at that time (there are many types of protocols), the IPV4 introduced in the 5.80s has a complex header design, and a total of 12 fields are normally required.
- There is an Option option in the packet header, which causes the length to be not fixed. The router needs to analyze the entire packet header before forwarding the data packet, and then determine the corresponding processing method.
- The IPV4 address can be configured with a second backup address on the interface, but when establishing a dynamic routing protocol, only the primary address can be used as an update source.
If addresses are to be re-planned in the future, it is difficult to achieve a transition without cutting off network usage.
IPV6 can configure two IP addresses under the interface, and both can be used as update sources.
- For various reasons, it is also difficult for IPV4 addresses to be reclaimed and reassigned by IANA.
- The address allocation is unreasonable, discontinuous, and has no hierarchy, which makes route summarization difficult.
- The routing protocol needs to transmit a large number of routes, which increases the burden on the router itself and causes the routing table to be too large.
- Because IPV4 is involved early, more consideration is to enable the network to be connected, and data security, multicast applications and mobile ip applications are not considered, resulting in insufficient support for related technologies–Ipsec was given to the network in the early days. It was designed for IPV6 and was later ported to IPV4.
- For security, all users use private network addresses, and it is impossible to accurately capture hackers.
- The multicast of IPV4 is divided into – broadcast and multicast. Broadcast will generate broadcast storms in the second-layer network, while multicast provides group addresses. Class D addresses are theoretically only 2 to the 28th power ——- —–IPV6 has no concept of broadcast, multicast = multicast and ARP of IpV4 is replaced by NDP.
- Although IPV4 can also provide the corresponding Anycast technology, there is no dedicated Anycast address.
- Multicast and Anycast traffic usually appear in the form of streams. IPv4 has a complicated way of processing streams. IPv4 does not have a specific identification of quintuple, so it is relatively difficult to determine a group of data streams, while IPv6 has a separate method for a group of data streams. The quintuple is used as the field of identification, which makes it more convenient to do QOS later.
- IPV4 addresses are not ideal either as identifiers or as locators.
The identifier is used to define the source end, and the locator is used to define the destination end.
The requirements for identifiers and locators are clearly defined in RFC2101, unique and permanent.
The existence of private network addresses leads to a large number of identical addresses in the network, so they are not unique and are not ideal identifiers.
Because the address space is not enough, after the host obtains the address through DHCP, it will change once the address is released.
Therefore, it is not permanent and is not an ideal locator.
- The existence of NAT technology greatly delays the implementation of IPV6 and alleviates the problem of insufficient address space.
Although NAT technology can delay the problem of address space exhaustion, but because NAT generates Translaton entries during translation.
In addition, each translation in the table entry needs to occupy 65KBytes of memory, consumes a lot of device resources, and each entry still has a lifetime problem, and once it is converted by NAT, the return route will also have problems.
- NAT will destroy the end-to-end IP application because the IP address will change.
- The VLSM subnetting technology proposed by IP seems to save space, but it is actually a waste in disguise, because each subnet has two addresses that are not available. The more subnets are divided, the more addresses are wasted.
- Although CIDR provides a method of summarizing routes and reducing routing entries, it is difficult to apply the problem of disorganized addresses allocated for historical reasons.
- DHCP can solve the problem of host address configuration, but DHCP server deployment is very troublesome.
22.IpV5 exists, but it died at the beginning of the design because of the design direction problem. And the draft of IpV7-V8-V9 has been proposed.
The advantages of IPv6
- It provides 128Bit, which is about 3.4*1038 addresses, which is theoretically 100 times the number of atoms on the earth’s surface. Every grain of sand on the earth can have an IP address.
- Not relying on NAT technology.
- The address distribution is orderly, making it easier to implement route summarization and simplify the routing table.
- An interface can be configured with multiple IPV6 addresses, so it is easier to re-modify the IP address.
- The host can automatically generate an ip address.
- The packet header is simple, 40Bytes, which makes the data forwarding more efficient.
- The flow label is provided in the packet header to better realize QOS.
- Expand Baotou to achieve new technical expansion, and should not be the original Baotou.
- The data packet is only fragmented at the source, and the interphase is not fragmented. The PMTUD technology is used to detect the minimum MTU value on the path at the source.
- There is no broadcast data, all are replaced by multicast data.
- There is no checksum, and the integrity is determined by Layer 2 and Layer 4 detection mechanisms.
- IpSec and Mobile ip are embedded in IPV6, which has better support.
- Multi-protocol migration method, dual-protocol routing protocol (support IpV4 and IPV6 at the same time), 6to4 Tunnel.
NAT-P conversion and other technologies are performed before IPV4 and IpV6 addresses.
Comparison of packet headers between IPV4 and IPV6
Compared with the IPV4 header, the IPV6 header removes the IHL, Identification, flags, Fragment, Offset, Header, Checksum, OPtion and padding fields.
The original Type of Service was changed to Traffic Class.
The original Time to live was changed to Hop Limit.
The original Total Length is changed to Payload Length.
The original Protocol is changed to Next Header.
IHL in IPv4 is the length of the IP header.
Total Langth in IPV4 is the total length – minus IHL is the length of the payload. The length of payload Length in IPV6 is the length of the payload.
Type of service = TOS in IpV4 is the field for QOS. The Trffic class in IPV6 is used for QOS.
New in IPV6 –Flow label Flow label – can label data flow.
Time TO LIVE =TTL in IpV4. Hop limit = TTL in IPV6.
Protocol in IPV4 is used to identify the ip protocol number. NEXT Header in IPV6 to identify the ip protocol number.
IPV6 has no checksum and fragmentation fields.
ThinkMo CCNA Dump exam information exchange group：
CCNA/CCNP/CCIE telegram study group：https://t.me/ccie_ei_lab
ThinkMo CCNA 200-301 Tutorial VIP Exclusive：
The complete EVE_NG file, free learning PDF and PPT that can be used directly, as well as video explaining the technical points are all here!