Vulnerabilities / Threats
11/5/2009
01:36 PM
Connect Directly
Google+
Twitter
RSS
E-Mail
50%
50%

Little-Known Hole Lets Attacker Hit Main Website Domain Via Its Subdomains

Proof-of-concept demonstrates how exploits on Google, Expedia, Chase Manhattan subdomains could lead to compromises of their main sites

Turns out an exploit on a Website's subdomain can be used to attack the main domain: A researcher has released a proof-of-concept showing how cookies can be abused to execute such an insidious attack.

Michael Bailey, senior researcher for Foreground Security, published a paper this week that demonstrates how an exploit in a subdomain, such as mail.google.com, could be used to hack the main production domain, google.com, all because of the way browsers handle cookies.

"There's no specific vulnerability here, but it's widening the attack surface for any large organization that has more than one [Web] server set up. A [vulnerability] in any one of those servers can affect all the rest," Bailey says.

Most Web developers aren't aware that a vulnerability in a subdomain could be used to target the main domain. "We're trying to get the message out that now you have to treat everything [in the domain] as though someone can compromise your crown jewels," says Michael Murray, CSO for Foreground. "You have to realize that every vulnerability, every attack vector in those subdomains, can be used to compromise [other areas of the domain]," he says.

It all boils down to the browsers themselves. Within the DNS architecture, the main domain -- fortune500company.com, for instance -- has control over its subdomains, such as development.fortune500company.com. Development.fortune500company.com has no authority to change anything on the main fortune500company.com site.

But browsers do the reverse, Murray says. Development.fortune500company.com can set cookies for fortune500company.com, the main domain. That leaves the door open for cookie-tampering, he says, when the subdomain has an exploitable vulnerability, such as cross-site scripting (XSS) or cross-site request forgery (CSRF).

"Most developers don't care if there's an XSS bug in the development site. But now all of those vulnerabilities are as important as an XSS in the main site," Murray says.

There's no actual patch to protect against the attack -- it's more about how Web developers set up their site domains. And when the site uses third-party content, it's even more at risk, according to the researchers.

In his paper (PDF), Foreground's Bailey showed PoCs with Google, Expedia, and Chase Manhattan's Websites. "The Chase online banking app could be compromised from one of the marketing companies they partnered with," Bailey says. His example in the PoC showed how images.chase.com, images.capitalone.com, and online.chasevisasignature.com all point to content from images.bigfootinteractive.com, an advertising agency's server. That server is actually vulnerable to XSS, he found, so it can be used to compromise a legitimate Chase user's banking session.

The answer, he says, is to use a completely different subdomain from your main one when pointing to content from a third party.

Bailey says so far he hasn't seen any actual attacks exploiting this. "But I have seen mentions of similar research on black hat forums. I know this isn't completely new to black hats," he says. He recommends users clean up their cookies regularly and use separate Web browsers or browser sessions for sensitive browsing activities.

Meanwhile, any flaw on a subdomain is putting all users on the main domain at risk. "They've got to treat a development site the same as the online banking site," Foreground's Murray says. "The implications are huge here."

Have a comment on this story? Please click "Discuss" below. If you'd like to contact Dark Reading's editors directly, send us a message. Kelly Jackson Higgins is Senior Editor at DarkReading.com. 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 Magazine, ... View Full Bio

Comment  | 
Print  | 
More Insights
Register for Dark Reading Newsletters
White Papers
Flash Poll
Current Issue
Cartoon
Video
Slideshows
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2014-0972
Published: 2014-08-01
The kgsl graphics driver for the Linux kernel 3.x, as used in Qualcomm Innovation Center (QuIC) Android contributions for MSM devices and other products, does not properly prevent write access to IOMMU context registers, which allows local users to select a custom page table, and consequently write ...

CVE-2014-2627
Published: 2014-08-01
Unspecified vulnerability in HP NonStop NetBatch G06.14 through G06.32.01, H06 through H06.28, and J06 through J06.17.01 allows remote authenticated users to gain privileges for NetBatch job execution via unknown vectors.

CVE-2014-3009
Published: 2014-08-01
The GDS component in IBM InfoSphere Master Data Management - Collaborative Edition 10.0 through 11.0 and InfoSphere Master Data Management Server for Product Information Management 9.0 and 9.1 does not properly handle FRAME elements, which makes it easier for remote authenticated users to conduct ph...

CVE-2014-3302
Published: 2014-08-01
user.php in Cisco WebEx Meetings Server 1.5(.1.131) and earlier does not properly implement the token timer for authenticated encryption, which allows remote attackers to obtain sensitive information via a crafted URL, aka Bug ID CSCuj81708.

CVE-2014-3534
Published: 2014-08-01
arch/s390/kernel/ptrace.c in the Linux kernel before 3.15.8 on the s390 platform does not properly restrict address-space control operations in PTRACE_POKEUSR_AREA requests, which allows local users to obtain read and write access to kernel memory locations, and consequently gain privileges, via a c...

Best of the Web
Dark Reading Radio