During BlackHat, David Litchfield disclosed a security issue with the Oracle 10g and 11g database platforms. The vulnerability centers on the ability to exploit low security privileges to compromise Oracle's Java implementation, resulting in a total takeover of the database. While the issue appears relatively easy to address, behind the scenes this disclosure has raised a stir in database security circles. The big issue is not the bug or misconfiguration issue, or whatever you want to call it. The issue is ethical disclosure -- a topic over which the security research community remains hotly divided.Many feel "responsible disclosure" means you do not release details of an attack until the vendor has patched the vulnerability. The idea is security researchers are duty-bound to both inform the vendor and patiently wait for the fix. On the other side of the coin, responsible disclosure means going public with details of a vulnerability when a vendor fails to patch in a timely fashion. Disclosure provides both public awareness so IT staff are aware of the problem, and a degree of motivation to push vendors to fix problems. The fundamental premise is attackers are every bit as skilled as their research counterparts and have even odds at discovering and exploiting vulnerabilities before anyone can react.
This particular disclosure has generated an inordinate amount of correspondence from members of the research community, as well as database and security product vendors. They feel dropping this 0-day exploit was premature, and the level of detail provided in research papers (PDF) is unwarranted to promote public awareness. But Oracle has ignored issues like this in the past, which the research community states puts database users at risk, and therefore makes it feel fully justified to disclose issues of a serious nature.
Litchfield, meanwhile, says he contacted Oracle several months ago -- which should be ample opportunity to address -- and therefore would be considered responsible disclosure.
Regardless, Web application firewall vendors, database activity monitoring vendors, and assessment platform providers -- which were not included in the process -- have all been scrambling during the past couple of days to close the gap.
While the damage that can result from this exploit is catastrophic, any seasoned database professional would not allow Oracle to be deployed with the required permissions settings any more than they would leave default DBA passwords. Second, since Litchfield is leaving his present employer, this has the feel of a publicity stunt. Any time "Oracle" and "exploit" are in the same sentence, it's deemed newsworthy. Throw in the longstanding Litchfield/Oracle feud, and it's guaranteed to get attention.
Whether you agree or disagree with the details of the attack being made public, be sure to remove EXECUTE privileges from the PUBLIC role for Java packages, and you should be safe.
Adrian Lane is an analyst/CTO with Securosis LLC, an independent security consulting practice. Special to Dark Reading. Adrian Lane is a Security Strategist and brings over 25 years of industry experience to the Securosis team, much of it at the executive level. Adrian specializes in database security, data security, and secure software development. With experience at Ingres, Oracle, and ... View Full Bio