FTP: How Does It Work?

FTP (File Transfer Protocol) is a way to transfer files on a network such as the internet. FTP is extremely efficient, far more efficient that web browsing, and is the preferred method for transferring large amounts of data. It is also extremely reliable.

One of the primary design goals of FTP was to hide the details of the file system from users. Thus, users do not need to be concerned with whether their files are on Windows, Unix, Linux or any other system. All they need to be concerned with is the business of moving around and transferring data.

When you browse the web, your browser and the web server communicate through a single connection or port. This means that commands to the web server (please send me a web page, for example) occur over the same connection as the web pages, graphics and other data. The direct implication is the connection can only be doing one thing at a time (not totally true, but accurate enough for our purposes here).

For example, your web browser tells the web server, "send me a page". The web server returns the page to the browser, which then asks for the next item, perhaps a graphic. The web server returns that to the browser, which then asks for the next item. As you can see, it is a back and forth transfer, first one then the other.

FTP is much better designed than this, in that it uses TWO connections between the FTP client and the FT server. One connection is used to communicate to the web server using standard Telnet commands. The other is used to transfer data.

This means that FTP is fast and efficient, since it can send extremely large amounts of data and still receive commands, handle aborts and do error correction. Web servers do not have this luxury, since they must wait for sends and receives to cease.