IETF Fix For SSL Protocol Complete
Security extension to the SSL/TLS protocol that protects against man-in-the middle attack is ready for prime time, security experts say
January 12, 2010
The Internet Engineering Task Force (IETF) has completed a security extension to the Secure Sockets Layer (SSL) protocol that fixes a flaw affecting browsers, servers, smart cards, and VPN products, as well as many lower-profile devices, such as Webcams, that contain the protocol embedded in their firmware.
Members of the IETF, the Industry Consortium for the Advancement of Security on the Internet, and several vendors, including Google, Microsoft, and PhoneFactor, have been working on a fix since October for the bug, which is basically a gap in the authentication process that lets an attacker execute a man-in-the-middle (MITM) attack and inject his own text into the encrypted SSL connection. The gap occurs in the renegotiation process of the session, when some applications require the encryption process be refreshed at a certain point.
Marsh Ray, a senior software development engineer for PhoneFactor who first discovered the SSL bug in August, says the IETF's extension to SSL, which is the Transport Layer Security (TLS) protocol in IETF parlance, secures the renegotiation process.
"This is a short extension to the handshake protocol of TLS," Ray says. "Some identifiers from the previous session are carried over to the handshake in the subsequent session."
The IETF has not yet formally published the official request for comments (RFC) document for the protocol extension, formally called the Transport Layer Security (TLS) Renegotiation Indication Extension, but Ray says it's stable now, and several open-source groups and vendors are working on implementing it.
So far, OpenSSL, GNU TLS, and NSS have been integrating and testing the new SSL/TLS security extension into their code, says Steve Dispensa, CTO at PhoneFactor.
"Some vendors are working with experimental code now in-house, as well," PhoneFactor's Ray says.
Actual deployment of the patch or protocol extension will take time, however, especially for commercial products that include SSL. "Any security vulnerability is going to be traumatic to a vendor. And this vulnerability, in particular, requires a lot of interoperability testing before vendors can ship it," he says. "That adds another layer of testing, and [quality assurance] has to be done before vendors are ready to ship it."
But, he says, vendors have had information about the flaw since September, so "they are not starting from zero now."
SSL's weaknesses have been under the microscope for the past year, first with the MITM hack by researcher Moxie Marlinspike that dupes a user into thinking he's in an HTTPS session when he has actually been redirected by the attacker elsewhere. Then came Dan Kaminsky's research, which exposed critical bugs in the X.509 digital certificate technology used in SSL.
But PhoneFactor's Marsh says this is the most "severe" attack against TLS. "There were some earlier attacks against SSL Version 2," he says. "This is probably the most significant attack against the core protocol itself."
Not all security experts agree it's a game-stopper, but the flaw definitely affects a wide range of vendor products. And the trouble is, not all of them will get patched. "Most of the major hardware vendors are aware of this, and many have been working with us," Marsh says. "But, realistically, some devices will probably never be [patched]," such as low-end printers and Webcams.
"Hopefully, they will not be handling anything critical enough so a man-in-the-middle attack is something serious they need to worry about," he says.
Meanwhile, the draft specification, which the IETF and the Internet Engineering Steering Group have deemed stable save for some copy editing, can be found 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.
About the Author
You May Also Like