[Content Navigation] Page 1: International Top Domain Name Overview Page 2: Hierarchical Structure of Domain Page 3: Domain Name Server Type Page 4: Domain Name Server Configuration Required File Page 5: UNIX DNS Server Bind Software
DNS (domain name system) is a domain name system. It is a service for domain name and IP address. The domain name is one of the most important identities on the Internet and one of the necessary elements of establishing a corporate website. According to the international top type domain name, it can be divided into 8 types:
The top type of my country's domain name can be divided into 6 types:
DNS is the same as most TCP / IP services, also using client / server mode. Use DNS to resolve hostnames into IP addresses to use a global, hierarchical distributed database system. The database contains all domain names on the Internet corresponding to its IP address. The hierarchy of the database allows the domain name space to be partially managed, called Domain. Database distributed features allow all different parts of the database to different networks of domain name servers, so that domain name servers can achieve independent management. Name Server contains partial information in the distributed database system, and uses this information to serve a client called a resolver. The domain name server runs a server process (which is generally a NAMED process in the UNIX system), which is parsed to the IP address and provides some permanent host information. DNS enables each name to control part of the information in the entire database and ensure that this information can be accessed throughout the network.
1. Hierarchical structure in the domain
DNS is a distributed hierarchy system that contains host information. Under DNS, the central database of all Internet information is not stored, which distributes on several gatenamers in a hierarchy, which is very similar to the hierarchies in the UNIX file system. Their structure is organized into a reversed tree, the root of the top of the structure. In the UNIX file system, the root is "/"; in the DNS, the name of the root is ".", And is called the root domain. The following figure shows a typical example of this structure.
The next level of the root field is called a top field, and these domains can be further divided into subdomains, and the subdomains can further divide into subdomains - the division directory and subdirectory in the UNIX file system are very similar. In the UNIX file system, the directory can also include both subdirectories or files, likewise, domains may include hosts or subdomains. The domain name in the DNS is generally equivalent to the absolute path name in the file system, which is a list of names to the root domain, "separated, such as"
Www.ggg.com ". Of course, one of the DNS can specify the name relative to its parent domain, similar to the relative path name in the file system. It is necessary to point out that the domain name relative to the root domain is a domain name ( FQDN, in DNS, a whole domain records a tag sequence, starting from the target domain name, ending with root domain name. For example: cs.gg.org is a domain name of the subdomain CS.
2, the type of domain name server
In the domain name service, the configuration of the domain name server is the most core content. If you configure the domain name server using BIND software, you can divide the domain name server into three types: primary domain server, auxiliary domain server, and cache server. The following is a simple comparison of these three types of servers:
(1) The primary domain server (PRMLARY NAME Server) The primary domain server is authoritative for domain names of one or more zones. It loads information directly from the local disk file built from the system administrator, and the primary domain server will also periodically transfer the data to the auxiliary domain name server in the same area. The primary domain server is an authoritative server that answers all the queries in the domain in absolute authority. When the primary domain server answers the customer request, you have a time period (specified by TTL) to specify the validity period of the data for the data. When you select a primary server, you should choose a computer that is relatively large memory and disk space. At the same time, a computer can act as a primary server in a zone, and can simultaneously act as an auxiliary domain server for several other domains. Configuring the primary domain server requires a complete set of files, the primary domain server will confirm that the area you are served and the data saved in which zone files are kept by booting file /etc/named.boot or named.conf. (2) Secondary Name Server Auxiliary Domain Name Server You can get a set of domain information from the primary domain server. All information is obtained from the primary domain server and backed up these data in a local disk file. The auxiliary domain server will play a backup in the case where the primary server does not work properly, thereby improving the reliability of the entire system. It will regularly update data on the server (that is, the transfer of zone files), thereby maintaining consistency with the primary domain server. On the auxiliary domain server, due to the full information of the domain, it is also possible to answer the query in the domain client. Configuring a secondary domain server does not need to create a local area file, you can download it from the primary domain server. But the cache file and the return file are still needed.
(3) Caching-Only Server A cache server can run a daemon of a domain name server, but does not have a local domain name server database file. It acquires the results of each query from the remote domain server, and saves these results to a local cache file, and it can be directly obtained from the cache when you need to query the same information. The query result saves the time in the cache to specify the "time-to-limited" (TTL) included by the data. After the expiration, the system will automatically delete this result. Cache servers can use forwarders (Forwarders). The forwarding server is a server connected to an external network, combining the data of these servers constitutes a large DNS database. In this mode, the cache server first puts a query request to a forwarding server. If necessary, the latter will go to query other servers to obtain data. The configuration of the cache server is relatively simple, only a cache file is required. But the most common configuration also includes a local return file. The files required for the above three domain name servers are required when configuring the following tables.
The main functions of the above documents are as follows:
Named.boot ---- This is the first file that the DNS daemon name is at startup; the file sets a general NAMED parameter, pointing the source of domain database information used by the server. Such sources can be a local disk file or a remote server. Named uses this file to determine the database file name and its location on this host and remote host. Named.ca ---- This file points to the root domain name server, contains the names and addresses of the root domain server of the Internet. With this file maintained in this file, a domain name server can contact the root domain server to parse the name inquiry. Named.hosts ---- This file defines the domain domain name server management, mainly responsible for mapping the host name as an IP address. It is a domain data generation and modification and propagation to other servers on the network or on the Internet. Named.REV ---- This file defines the reverse domain in-ADDR.ARPA of the management name, which is mainly responsible for mapping the IP address to the host name. Named.local ---- Used to transfer the return address locally, responsible for analyzing localhost into address 127.0.0.1. When the Named daemon starts, read the named.boot file, obtain the type of the server, and initialize data information. This file holds the basic parameters of setting named and pointing to the source of domain databases used by the server. Bind (Berkeley Internet Name Domain), Burkeley Internet Name System, is an open source package of Internet Software Federation (ISC), Bind provides a parser and name server software, the parser doing actual query work and the name server Then provide a response, the DNS server currently used by UNIX is mainly implemented by the BIND software. Bind has three major versions: BIND4, BIND8 and BIND9.
Bind4 is in the late 1980s, Bind 8 was released in 1997, and Bind9 was posted in the mid-2000. Without Bind 5, 6 and 7, Bind 8 is a major update version of such meaning, so that everyone feels that it is twice the version number using an old version number, but in fact, it is not true, Bind 8 and 4.4 BSD It is released together, so all version numbers are raised to 8; the latest version of Bind is bind9.3.0.
Bind 8 is integrated into many advanced technologies capable of improving efficiency, stability, and security. Bind9 is even farther, with multiprocessor support, thread security operation, real-time security (public key encryption), IPv6 support, incremental area transfer, and some other features.
Bind9 is completely redesigned and re-real, it is independent of the specific part of the operating system, so it can be more easily ported to the non-Unix system. The internal mechanism of BIND9 has been significantly different, but its configuration step remains unchanged. The BIND system consists of three parts: answering the query called NAMED daemon; contact the DNS distributed database server resolution host query library routine; DNS command line interface: nslookup, DIG, and Host. In terms of DNS, similar to a daemon (or running it) is called "name server", the client program is called "parser".
The format of the Named configuration file has a thorough change between Bind4 and Bind 8, and the new format is similar to the format of Gated.conf. The configuration file name also changed: in Bind4 is /etc/named.boot, and in Bind 8 and 9 /etc/named.conf, the format of the cache file and the data file remains unchanged.
