Dark Reading is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them.Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Attacks/Breaches

2/13/2007
07:45 AM
Connect Directly
Google+
Twitter
RSS
E-Mail
50%
50%

Study: Two Hacks a Minute

Researchers at the University of Maryland show firsthand why weak passwords are a hacker's dream come true

Here's a scary new statistic: A new University of Maryland study reports that the average computer on the Internet is attacked every 39 seconds. That's not even enough time for you to safely go get a cup of coffee.

The study by the university's Clark School Center for Risk and Reliability and Institute for Systems Research tried to do something that isn't typically done: Quantify attack trends, says Michel Cukier, who headed up the study. "Quantification of computer security is something that's really needed and lacking in the field," says Cukier, who is assistant professor of mechanical engineering and a Clark affiliate.

Cukier and his team of researchers studied trends in how brute-force attackers get in, what they look for, and what makes a computer more likely to get hacked. They ran the study over a 24-day period in November and December of last year, setting up four Linux computers with weak passwords, and that was all it took. "We had 269,262 attack attempts," he says. "And 824 were successful. They got the username and password."

Instead of using forensics, where you analyze what an attacker did to your system, the Maryland researchers looked at the hacker behavior more statistically. "We looked for specific things, like how often an attacker checked for a software configuration, how often he would install some program, download a file or run a program, change the password, or check for hardware configuration," he says. "Once you understand what an attacker does, it can help security administrators build better systems."

The Linux machines used in the study were attacked an average of 2,244 times a day. The attacks came from 229 unique IP addresses, Cukier says. He says he expected to attract attackers with the weak passwords -- but he didn't expect such high frequency of attacks.

RSnake, the founder of ha.ckers.org, says the university's numbers sync with what he's seen in the wild. "We were getting over 1,000 php include-type attacks alone per week. It works out to six attacks per hour... one PHP-include attack per 10 minutes."

In the University of Maryland study, many attackers used dictionary scripts to guess the usernames and passwords. "Root" was the top username guess, attempted twelve times as often as "admin," the second-most guessed password. Attackers also looked for other obvious and popular usernames, such as "test," "guest," "info," "adm," "user," "mysql," "oracle," and "administrator."

To crack the password, the attackers would re-enter or attempt variations of the username. About 43 percent of the attempts were re-entering the username, and other combinations included username plus "123."

Behind most of the attacks, of course, were botnet operators casting a wide net to recruit new zombies. In the most common sequence of events, an attacker would check the software configuration of the computer, change the password, then check the hardware and software configuration again, download a file, install the downloaded program, and then run it.

Cukier says the study's machines were purposely loaded with uninteresting content -- no confidential (or fake confidential) data, no large disk space for hosting malware or for other nefarious purposes. It's difficult to tell if the university was a specific target, he says. "You could identify us through the range of IP addresses" we used, he says. "But most of the attackers were looking for a quick [machine to] compromise, rather than specifically for the University of Maryland."

And that fits with the pattern of botnet operators. "The bad guys are trying the low hanging fruit first -- they try the obvious stuff, which is most likely to give them access, rather than trying more obscure stuff, or attempting to get more information or recon," RSnake says. "Essentially they are running this like a business... Primarily it looked like they were looking to create new bots."

The university researchers have submitted a paper on the study to the IEEE International Conference on Dependable Systems and Networks. In the meantime, they plan to launch Unix and Windows machines into the line of fire for further research.

Cukier says the lesson learned by the study is that weak passwords are still a major target of attackers. "If you don't do something about this, you have a high probability of being attacked."

— Kelly Jackson Higgins, Senior Editor, Dark Reading

Kelly Jackson Higgins is the Executive Editor of Dark Reading. She is an award-winning veteran technology and business journalist with more than two decades of experience in reporting and editing for various publications, including Network Computing, Secure Enterprise ... View Full Bio

Comment  | 
Print  | 
More Insights
Comments
Oldest First  |  Newest First  |  Threaded View
HackerOne Drops Mobile Voting App Vendor Voatz
Dark Reading Staff 3/30/2020
Limited-Time Free Offers to Secure the Enterprise Amid COVID-19
Curtis Franklin Jr., Senior Editor at Dark Reading,  3/31/2020
Register for Dark Reading Newsletters
White Papers
Video
Cartoon Contest
Current Issue
6 Emerging Cyber Threats That Enterprises Face in 2020
This Tech Digest gives an in-depth look at six emerging cyber threats that enterprises could face in 2020. Download your copy today!
Flash Poll
State of Cybersecurity Incident Response
State of Cybersecurity Incident Response
Data breaches and regulations have forced organizations to pay closer attention to the security incident response function. However, security leaders may be overestimating their ability to detect and respond to security incidents. Read this report to find out more.
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2020-11565
PUBLISHED: 2020-04-06
An issue was discovered in the Linux kernel through 5.6.2. mpol_parse_str in mm/mempolicy.c has a stack-based out-of-bounds write because an empty nodelist is mishandled during mount option parsing, aka CID-aa9f7d5172fa.
CVE-2020-11558
PUBLISHED: 2020-04-05
An issue was discovered in libgpac.a in GPAC 0.8.0, as demonstrated by MP4Box. audio_sample_entry_Read in isomedia/box_code_base.c does not properly decide when to make gf_isom_box_del calls. This leads to various use-after-free outcomes involving mdia_Read, gf_isom_delete_movie, and gf_isom_parse_m...
CVE-2020-11547
PUBLISHED: 2020-04-05
PRTG Network Monitor before 20.1.57.1745 allows remote unauthenticated attackers to obtain information about probes running or the server itself (CPU usage, memory, Windows version, and internal statistics) via an HTTP request, as demonstrated by type=probes to login.htm or index.htm.
CVE-2020-11548
PUBLISHED: 2020-04-05
The Search Meter plugin through 2.13.2 for WordPress allows user input introduced in the search bar to be any formula. The attacker could achieve remote code execution via CSV injection if a wp-admin/index.php?page=search-meter Export is performed.
CVE-2020-11542
PUBLISHED: 2020-04-04
3xLOGIC Infinias eIDC32 2.213 devices with Web 1.107 allow Authentication Bypass via CMD.HTM?CMD= because authentication depends on the client side's interpretation of the <KEY>MYKEY</KEY> substring.