/ 20 March 1998

Brutes and sneaks, hackers attack

Hackers use a great variety of techniques to break into private computer systems. John Graham-Cumming reports

In a frenzy of announcements over the past few weeks, Nasa, the Pentagon, the United States Navy and a number of universities revealed that their computers were under cyber-attack. Many of the attacks relied on tried and tested hacking techniques, some t echnical and others psychological. They go by such names as social engineering, trashing, denial of service, brute force, sniffing, buffer overflo w, spoofing and Trojan horses.

Hacking into a private system often requires inside knowledge. Trashing – looking through dustbins for old computer manuals, printouts and password lists – is a good starting point. Companies routinely throw away out-of-date information without shredding it.

Hackers also use a technique called “social engineering” which entails telephoning the organisation and claiming to be an information technology worker. The hacker is often able to extract information about the system, even user login names and passwords .

If this fails, the hacker may resort to brute force, which includes password-cracking programs and war dialling. As many users choose easy-to-guess passwords, a program will often yield good results simply by trying every word in a dictionary. The simple defence is to choose a complex password, such as a mixture of letters, numbers and punctuation.

War dialling gets its name from the 1983 hacking film War Games. A simple computer program is directed to dial all the phone numbers in a specific area code or region and note the tones that would identify a computer answering. The best defence against t his are the so-called “call-back systems” which allow the operator of a secure computer system to specify exactly from whom they will accept exter nal telephone connections.

Computer systems connected to networks, such as the Internet, are also vulnerable to sniffing: eavesdropping on communications between computers. Special software called packet sniffers can be connected to a computer network and extract all the packets o f information used to pass data between computers.

One of the most technical attacks is known as a buffer overflow. These were used in the infamous Morris Worm attack that caused widespread disruption to the Internet in 1988 and have recently been found in AOL Internet Messenger and Microsoft Internet Ex plorer. (Patches, the term for solutions to computer problems, can be found at .)

Many computer operating systems store programs and data interleaved in the same part of memory. When one part of the data is a buffer, a location used to store data being transmitted to the computer, it is often possible to send more data than will fit i n the allocated space.

When this happens the data overflows, possibly overwriting the program that is running. A sequence of commands can be sent that gets written into the computer’s memory and allows the hacker to take control of the machine. Some recent attacks on the Inter net were intended to disrupt service to specific machines or regions of the Internet and prevent legitimate users from getting access to their own machines. This is known as a denial of service.

One attack called a “Smurf” allows a hacker to remotely direct a large number of computers to send messages to a specific machine on the Internet, making it slow down or become unusable. A single message with a false or spoofed source address is sent to an Internet service provider.

The Internet service provider’s router connection to the Internet regenerates the message, sending it to each computer connected to the Internet service provider. Each computer then responds to the message by replying to the false source

And finally, having gone to the effort of breaking into a computer system, many hackers leave behind an easy way to get back in. Programs called Trojan horses are left running in the target computers and appear to be performing a normal service, but they often contain backdoors allowing hackers easy access.

The simplest Trojan horse replaces the messages shown when a login is requested. Users think they are logging into the system and unwittingly provide their user names and passwords to a program that simply records the information for use by the hacker.

Other Trojan horses can be nastier – they perform destructive activities, like deleting hard disks, or help hide the hacker’s tracks by automatically deleting auditing information.

Many of the hacking techniques outlined are easily countered through appropriate policies and guidelines for organisations.

The US Department of Energy’s Computer Incidence Advisory Capability at and the Computer Emergency Response Team of Carnegie Mellon University at put out regular bulletins describing recent attacks by hac kers and the various defences used against them.