As the hub of global Internet communications, there are only 13 root servers in the world, of which 10 are in the United States alone. If the United States shuts down the root servers one day, will our Internet really be cut off?
The so-called root server does not actually have 13 physical devices but refers to 13 top-level IP addresses or 13 root domain names. Many domain names can be derived from the 13 root domain names, and the descending level is the first level. The domain name, second-level domain name, and third-level domain name. In 2004, the United States and Libya diverged on the management and ownership of the (.LY) first-level domain name, and eventually, the United States stopped providing Libya with the domain name resolution service. As a result, Libya evaporated from the global Internet for three days. In addition, the United States has also stopped the resolution of the first-level domain name of Iraq (.iq), which prevented Iraqi netizens from accessing all URLs with a (.iq) suffix.
We often talk about air supremacy and sea supremacy, and these events show that there is still “net supremacy” on the Internet.
How does the root server work, and why does the United States have a unique “net power”?
Is a specific process? First, when we enter the URL, the browser will check whether it has a corresponding cached IP address. If there is, the website can be opened directly. If not, the browser will continue to ask the computer’s built-in local server. If there is still no IP address, the browser will be Ask the authoritative server on the Internet, or there is no, continue to request the top-level domain name server, can’t find the answer? Even if there is no cache at all levels of servers, the root server will eventually intervene to resolve the domain name and return the IP address corresponding to the URL to the terminal computer, and then the computer can successfully access the website.
If you compare the Internet to a book, the root server is equivalent to a directory of books. When we visit a website on the computer, the process of entering the URL is generally like the process of searching the directory, and the website address is the number of pages corresponding to the directory. Only when you enter the correct address and find the correct page Count, you can see all the content on this page.
In fact, it rarely happens that the computer browser directly accesses the root server, because all levels of servers from the computer to the Internet will have caches. If there is no cache, we finally visit the root server.
There are a total of 13 root servers in the world, also known as 13 roots or 13 root DNSs. Their names are root domain names, and their names are from A.root-servers.net to M.root-servers.net, which is from the letters ABCD All the way to M, the ones starting with A are called primary roots, and the remaining 12 are called secondary roots. Ten of these 13 root servers are located in the United States, and Japan, the United Kingdom, and Sweden each get one. There are many physical servers running behind each root.
Well, when the website we visit can’t find the IP address, must it be resolved by the American root server?
The answer is not necessarily. Since 2003, my country has been introducing root mirror servers, including Hong Kong, Macao, and Taiwan. At present, my country actually has 28 root mirror servers. These mirror servers are like clone servers and can be the same as 12 auxiliary roots. Synchronizing data from the main root to itself, and finally sharing 13 IP addresses with the main root, so when we want to request a root server, we don’t have to visit the American root server, just go through the mirror server in our country.
At present, the 13 roots are basically managed by independent institutions or companies in various countries. The US Information Administration and other departments have delegated management power to these organizations. The beginning and end of the domain name resolution in Iraq and Libya mentioned at the beginning, one is the temporary suspension of service from the replacement of domain name managers, and the other is the result of differences in domain name management between institutions, which are not serious in nature.
However, based on the above theory, the data content of the primary root may still be tampered with, which will cause errors in the data synchronized by the mirror server. Although the Internet is open to the world, the possibility of such an incident is extremely low, but what if it happens?
In fact, we also have countermeasures. First of all, in theory, we can respond in time without synchronizing these tampered data. Secondly, these 13 roots are based on the product of the IPV4 Internet era. Due to technical limitations, the number of root servers has been limited to 13, as an important infrastructure of the Internet, in order to break the limit of 13 root servers.
As early as 2015, my country launched the “Snowman Project” in conjunction with Japan’s WIDE and many other institutions. Among them, WIDE is the operator of 13 roots. IPV6 has also become the sixth version of the Internet protocol. The Snowman Project is based on this protocol. As of the end of 2017, a total of 25 root servers have been installed in the world. my country has deployed four, including one main root and three auxiliary servers. root. Although these new root servers are different from the 13 original roots, they are also effective methods for countries around the world to seek root server dilemmas. In short, there are many technical branches involved in the details, and we don’t need to worry about it.