Analyze the Domain Name System in the IPv6 Era

IPv6 is the next generation network protocol that replaces IPv4. It has many new features and functions. Domain Name System (DNS) is the basic architecture of the Internet. The new features of IPv6 also require DNS support. Therefore, DNS must be upgraded to meet IPv6 requirements. This article analyzes the architecture of IPv6 DNS, IPv6 address resolution, IPv6 address automatic configuration and plug-and-play, and IPv4 to IPv6 transition. research.
I. IPv6 advantages
Domain Name System (DNS) is mainly used to precisely locate network resources through the ing between Domain names and IP addresses, that is, querying IP addresses based on Domain names, and vice versa. DNS is the basic architecture of today's Internet, and many network services are built on the basis of the DNS system. "Only by understanding DNS can we truly understand the Internet," said an industry leader ." IPv6 is an Internet protocol used to replace IPv4. Compared with IPv4, IPv6 has many advantages. First, it provides a huge address space. Second, IPv6 address structure and Address Allocation adopt a strict hierarchy to facilitate Address Aggregation, in this way, the route table scale in the router is greatly reduced. Once again, IPv6 supports automatic configuration of network node addresses to implement the plug-and-play function. In addition, IPv6 has good support for host mobility and is suitable for more and more internet mobile applications. IPv6 has more advantages than IPv4 in terms of security and support for Multimedia Streams.
DNS in IPv6 networks is very important. Some new IPv6 features are inseparable from DNS support. This article analyzes the architecture of IPv6 DNS, IPv6 address resolution, IPv6 address automatic configuration, plug-and-play, and IPv4 to IPv6 transition. research.
Ii. Architecture of IPv6 Domain Name System
The DNS in the IPv6 network is consistent with the IPv4 DNS in the architecture, as shown in Domain Name Space 1 in the tree structure ). The difference between the IPv4 protocol and the IPv6 protocol does not mean that the IPv4 DNS System and the IPv6 DNS system must be applied separately. On the contrary, their DNS systems and domain name space must be consistent, that is, IPv4 and IPv6 share the same domain name space. During the transition from IPv4 to IPv6, the domain name can correspond to multiple IPv4 and IPv6 addresses at the same time. With the popularization of IPv6 networks, IPv6 addresses will gradually replace IPv4 addresses.

At the top of Figure 1 is the only Root in the DNS tree structure. The next Level of the root is called Top Level Domain (TLD), also known as Top Level Domain. The lower Level of the top-Level Domain is the Second Level Domain (SLD), and the lower Level of the Second Level Domain is the third Level Domain. Each Domain is a subdomain of its parent Domain. For example, ".net.cn" is ". cn, while cnnic.net.cn is both a subdomain of net.cn and a subdomain of net.cn. the subdomain of cn.
Each node on the DNS tree has a Label. The root node ID is "null", that is, the length is 0), and the ID length of other nodes is 1 ~ Between 63 bytes. The domain name of a node is composed of all the identifiers in the path from this node to the root node, which are arranged from left to right and separated. For example, http://www.ccw.net.cn/
The entire DNS domain name space is divided into several zones), see the elliptical mark in. Each zone has a Domain Name Server including the primary server and secondary server), and stores domain name information in the form of Resource Record. Resource records include host name domains), IP address ing, and subdomain server authorization. When using the DNS service, you do not have to carefully understand the tree structure of the DNS domain name space. You only need to specify a DNS server when setting the network or use dynamic hosts to configure DHCP) and other related technologies. Your applications can access the DNS system through the Resolver embedded in the operating system to query network resource information related to domain names.
Iii. IPv6 address and its representation
The IPv6 address is 128 bits in length and can be divided into three types based on its transmission type:
1) Unicast Address (Unicast Address): used to identify a single network interface. The packet whose destination address is the unicast address will be sent to the network interface identified by this address.
2) Anycast Address: identifies a group of network interfaces that generally belong to different nodes ). The destination address is the IP address of the network interface closest to the route.
(3) Multicast Address (Multicast Address): used to identify a group of network interfaces that generally belong to different nodes ). Packets sent to multicast addresses are sent to all network interfaces in this group. In IPv6, there is no Broadcast Address) and it is replaced by multicast Address. Unicast Addresses are classified into aggregated Global Unicast Addresses Aggregatable Global Unicast Addresses according to the transmission range of Addresses), NSAP Addresses, IPX hierarchical Addresses, and Site-Local Addresses) and link-Local Address. All network interfaces must have at least one local link address. At the same time, they can have multiple addresses, including unicast addresses, anycast addresses, and multicast addresses ).
During the expression and writing of IPv6 addresses, 128 bits are divided into eight 16 bits by a colon, each of which contains 4-digit hexadecimal numbers. For example:
1080: 0000: 0000: 0000: 0008: 0800: 200C: 123A
In the hexadecimal number of each 4-digit group, if its high value is 0, it can be omitted. For example, write 0800 as and as 0. So 1080: 0000: 0000: 0000: 0008: 0800: 200C: 123A can be abbreviated as 1080: 0: 0: 0: 8: 800: 200C: 123A. To further simplify the process, the rule of overlapping colons is imported in the specification, that is, the consecutive 16-Bit 0 in the address is replaced by overlapping colons. For example, after three consecutive 0 replicas in the preceding example, they can be expressed as follows: 1080: 8: 800: 200C: 123A. Rules that overlap colons can only be used once in an address. For example, address 0: 0: 0: BA98: 7654: 0: 0: 0 can be abbreviated as: BA98: 7654: 0: 0: 0 or 0: 0: 0: BA98: 7654:, but cannot be recorded as: BA98: 7654 ::.
In addition, you can use "IPv6 address/prefix length" to indicate the address prefix. This representation is similar to the IPv4 address prefix representation in CIDR. Here, the IPv6 address is the IPv6 address in any of the preceding notation. The prefix length is a decimal value, which specifies the leftmost bits in the address to form the prefix. For example, for 32-bit prefix 10800000 hexadecimal), it can be expressed as follows:
1080: 8: 800: 200C: 123A/32
1080:/32
Iv. DNS support for IPv6 address hierarchy
The global unicast addresses that can be aggregated by IPv6 are addresses that are used globally. Layers and Address Aggregation are required. Its hierarchy is shown in the following table:
Where:
FP001): Format Prefix for aggregating global unicast addresses. FP: Format Prefix) 3 bits );
Tla id: Top-Level Aggregation Identifier );
RES: reserved for future use;
Nla id: Next-Level Aggregation Identifier );
Sla id: Site-Level Aggregation Identifier );
Interface id: INTERFACE identifier.
The allocation of IPv6 global unicast addresses is as follows: the top-level Address Aggregation organization TLA is a large ISP or address management organization) obtains a large address and is responsible for distributing it to the secondary Address Aggregation agency NLA small-and medium-sized ISP) assign an address. NLA assigns an address to the SLA subnet of the site-level Address Aggregation Organization) and a network user. The IPv6 address hierarchy can be well supported by the address chain technology in DNS. The following is an analysis of forward and reverse DNS resolution.
1. Forward Parsing
The resource record for IPv4 address forward resolution is A record. There are two types of resource records for forward resolution of IPv6 addresses: "AAAA" and "A6. "AAAA" mentioned earlier that it is A simple extension of "A" records. As the IP address is increased from 32 to 128 bits, the number of resource records is increased from "A" to four "". "AAAA" is used to indicate the ing between a domain name and an IPv6 address, and does not support address hierarchy.
"A6" proposed in RFC2874 that it establishes a connection between an IPv6 address and multiple "A6" records. Each "A6" record only contains part of the IPv6 address, after integration, it is assembled into a complete IPv6 address. The "A6" record supports some new features not available in "AAAA", such as Address Aggregation and address change Renumber.
First, the "A6" record method splits 128-bit IPv6 addresses into several levels of address prefixes and address suffixes Based on the allocation levels of TLA, NLA, and SLA, forming an address chain. Each address prefix and suffix are a part of the address chain. A complete address chain forms an IPv6 address. This idea conforms to the IPv6 address hierarchy and supports Address Aggregation.
Second, when you change the ISP, you need to change the IPv6 address of the ISP as the ISP changes. It is very cumbersome to manually modify all the addresses registered in the DNS in the user subnet. In the address chain represented by the "A6" record, you only need to change the ISP name corresponding to the address prefix, which greatly reduces the modification of resource records in DNS, in addition, the Address Allocation layer is closer to the underlying layer, and less changes are required.
2. Reverse resolution
The record of IPv6 reverse resolution is the same as that of IPv4. It is a "PTR", but there are two address representations. One is a half-byte hex digital Format separated by ".". The low-byte address is in front, the high-byte address is in front, and the domain suffix is "IP6.INT .". The other is the binary string Bit-string format, which starts with "\ [". The hexadecimal address has no separator, and the high position is in front and the low position is in the back, add "]" after the address, and the domain suffix is "IP6.ARPA. ". The half-byte hexadecimal numeric format corresponds to "AAAA" and is a simple extension of IPv4. The binary string format corresponds to the "A6" record, and the address is similar to "A6". It can be divided into multi-level address chains for representation. "DNAME" is used for each level of authorization. Like A6, the binary string format also supports the address hierarchy feature.
V. plug-and-play and DNS in IPv6
IPv6 supports automatic address configuration, which is a plug-and-play mechanism. IPv6 supports automatic configuration of stateless addresses and automatic configuration of stateful addresses.
In the automatic configuration mode of stateless addresses, the network interface that needs to configure the addresses first uses the neighbor discovery mechanism to obtain a local link address. After the network interface obtains the local address of the link, it accepts the address prefix advertised by the router and obtains a global address based on the Interface ID. The automatic configuration of stateful addresses, such as DHCP (Dynamic Host Configuration Protocol), requires a DHCP server to Obtain Address Configuration Information from the DHCP server in Client/Server mode.
The IPv6 node automatically obtains the IPv6 address and gateway address through address configuration. However, the automatic address configuration does not include the automatic configuration of the DNS server. How to automatically discover the DNS server that provides the Resolution Service is also a problem to be solved. The automatic discovery solution of the DNS server being studied can be divided into stateless and stateful.
In stateless mode, you need to configure any multicast addresses within the site range for the DNS servers in the subnet. The node to be automatically configured sends a server Discovery request with the target IP address as the stream play address, and queries DNS information such as the DNS server address, domain name, and search path. This request is sent to the nearest DNS server. Based on the request, the server answers DNS information such as the unicast address, domain name, and search path of the DNS server. The node configures the local DNS information based on the server's response. Later DNS requests will be sent directly to the DNS server using the unicast address.
When a stateful DNS server discovers a domain name, it uses a server like DHCP to notify the node of DNS information such as the DNS server address, domain name, and search path. Of course, this requires additional servers.
Vi. IPv6 transition phase and DNS
During the transition from IPv4 to IPv6, the DNS service as the Internet infrastructure must also support the upgrade and conversion of this network protocol. IPv4 and IPv6 DNS record formats are different, in order to achieve the IPv4 network and IPv6 network between the DNS query and response, you can use the Application Layer Gateway DNS-ALG combined with NAT-PT method, it serves as a translation between IPv4 and IPv6 networks. For example, IPv4 address domain name ing uses the "A" record, while IPv6 uses the "AAAA" or "A6" record. Then, IPv4 nodes to IPv6 network DNS query request is "A" record, the DNS-ALG will "A" into "AAAA", and sent to the IPv6 network DNS server. When the server's answer arrives at the DNS-ALG, The DNS-ALG modifies the answer, change "AAAA" to "A", change the IPv6 address to the IPv4 translation address in the DNS-ALG address pool, inform the NAT-PT of the ing between the IPv4 translation address and IPv6 address and return the IPv4 host as the resolution result. The IPv4 host uses this IPv4 translation address as the destination address to communicate with the actual IPv6 host through the NAT-PT. Example. For A dual-protocol stack transition method, "A" record and "AAAA" or "A6" record exist simultaneously on the DNS server. Since nodes can handle both IPv4 and IPv6 protocols, there is no need to switch devices like DNS-ALG. No matter whether the DNS server answers the "A" record or the "AAAA" record, it can communicate.
VII. Summary
With the development of Internet technology, IPv6 is getting closer and closer to us. As a basic network service in the IPv4 era, DNS plays an important role in the Internet. In the upcoming IPv6 era, new protocols and functions require that DNS not only provide simple resource positioning in the traditional sense, but also provide basic functions similar to IPv4 DNS, on the other hand, the new features of IPv6 are combined with other protocols to provide new functions, making network configuration, maintenance, and use easier and more convenient, let users feel the new experience brought by new technologies.