DNS or Domain Name System was created to resolve the relationship between the Internet address (that is specified as a text string) and the IP address. The network only understands binary addresses but users find it difficult to use binary addresses and would like to use alphabetic or alphanumeric addresses. Similarly, all programs used on the Internet use the alphanumeric (or text) address.
Therefore, the Internet must have some method to resolve the two addresses. This is done using the Domain Name System. In the ARPANET network, where there were only a few hundred hosts, the IP addresses of these hosts were retained in a file called hosts(dot)txt.
This system worked fairly efficiently. Unfortunately, this system cannot be applied in case of the Internet because the number of hosts has gone up from a few hundred by several orders of magnitude. There are two problems with retaining a hosts(dot)txt file.
First of all, with the current number of users, such a file would be too big and unwieldy for the purposes of a search.
ADVERTISEMENTS:
Secondly, there may be a conflict in address names, unless there is a central agency looking after the allocation of the Internet address. These problems cannot be sorted out by having a single Domain Name Server. There are likely two be two problems if a single name server is used.
The first problem is the same as that with using the hosts(dot)txt file. A single server will be swamped with the number of hosts being very large. This will make searches very slow and the users will not accept this kind of delay. Secondly, imagine what would happen if this single name server went down for some reason.
The entire Internet would collapse. Since the Internet has reached a status where many users need the Internet for their daily business and other activities, many businesses world wide would collapse. Clearly, this would be a desperate situation and it must be avoided.
When domain name servers, were to be introduced on the Internet, the initial attempt was to replace the hosts(dot)txt file with a single server.
ADVERTISEMENTS:
Fortunately, the opinions of those, who foresaw the rapid growth of the Internet and the consequent problems with a single domain name server, was finally accepted and multiple domain name servers were accepted. As per the procedure adopted and being used currently, the DNS name space is divided into many non-overlapping zones.
Each zone has its own primary name server and several secondary servers. When a request for an IP address resolution is received, the resolver sends the request to the local name server. If the address being sought falls under its jurisdiction, it returns the IP address.
If not, then it sends the request to the name server higher up in the ladder to get an authoritative record.
It is possible that the local name server may have the IP address in its cache, but that may be old and out of date and may have been superseded. Hence the insistence on authoritative records. This process of IP address location becomes both faster and easier by sending the resolver request “up the domain tree” until the IP address is obtained.
ADVERTISEMENTS:
This discussion on the terms used to describe the Internet can be continued to describe the methods adopted on the Internet to address most of the issues (for example, the details of how IP address resolution is performed), that come up in normally using the Internet. However, it is time for us to look at the services available on the Internet, but we will not look at any of the details.
Before we shift our attention to Internet services, we shall go back to the steps that are taken when a user’s browser makes a request for information from the Internet.
Now that we know and understand the special terms that have been used in the verbal description given, we are in a better position to understand this process. Depending on the hyperlinks and the web pages required, more servers may be involved.
