http://www.nwfusion.com/news/2004/0607confuse.html

By Paul Roberts
IDG News Service
06/07/04

A report last week about a security hole in a wireless broadband
router made by Cisco's Linksys division overstated the severity of the
vulnerability, according to the man who first warned of the problem.

Independent technology consultant Alan Rateliff said Monday that
Cisco's Linksys WRT54G wireless routers are not, by default,
vulnerable to remote takeover from a malicious hacker. However, a
vulnerability in the software that runs on those devices could still
allow a malicious hacker to access administrative features for the
router and take control of the device.

Rateliff first posted a warning about the WRT54G on the Bugtraq
discussion list on May 31. Based on testing with a sample Cisco
router, Rateliff concluded that the routers were shipped with a
configuration that would allow remote attackers to access the
Web-based administration interface for the devices over two common
communications ports, 80 and 443. The WRT54G, like other wireless
routers, enables multiple computers to share a broadband Internet
connection using wireless networking equipment

The Bugtraq post prompted numerous responses that contested Rateliff's
findings. After testing additional WRT54G devices, Rateliff said he
found that the devices were not vulnerable in their default
configuration, but could still be compromised remotely given the right
circumstances.

In particular, Rateliff discovered that a firewall feature in the
routers is enabled, rather than disabled, by default, which prevents
compromise on new systems.

On versions of the router using firmware versions 2.02.2 and 2.02.7,
malicious hackers can access the router's administrator interface and
change the configuration of the router if the firewall feature is
disabled and if the router's owner does not change the default
administrator's password. The devices could be compromised regardless
of whether a feature that provides remote, Web-based access to the
routers was enabled or disabled, he said.

Cisco has since released a test, or "beta" version of software for the
device that fixes the remote access problem, he said.

Rateliff posted a message to Bugtraq on June 2 and acknowledged that
he made an error in his initial warning about the problem, but said he
was just posting his findings based on a test of the Cisco hardware,
standard practice in the Bugtraq forum. Rateliff did not expect the
immense response to his post, which spawned stories in a number of
online news outlets and prompted multiple responses on Bugtraq.

"The exposure on this is not as bad as the (discussion) on Bugtraq
made it seem. I can't account for the results of the first test, but
at this point that's irrelevant. What's relevant now is that 'out of
box' home users are safe," Rateliff said.



_________________________________________
ISN mailing list
Sponsored by: OSVDB.org

Reply via email to