Archive

Posts Tagged ‘HitB2010AMS’

HitB2010Ams – XProbe-NG: Building efficient Network Discovery Tools

July 2nd, 2010 No comments

By Fyodor Yarachkin

To clear up a common misunderstanding, this Fyodor is not the same Fyodor as the author of Nmap.

XProbe-NG was written to discover a rouge server in a network of the major Taiwanese internet provider. It turned out that XProbe was not sufficient to handle all the application level stuff that was going on in this case.

However doing level 7 probes introduced two problems:

  • Bandwidth – Having to send far more data
  • Time – Making sure you finish in time

Other motivations for XProbe-NG include:

  • Scanning other protocols then IP only
  • Bulk scanning
  • Probing “en-route” systems
  • Migration to IPv6
  • Honeypots/nets
  • Improving precision

Read more…

HitB2010Ams – Ten Crazy Ideas That Might Actually Change the State of Information Security

July 2nd, 2010 No comments

By Mark Curphey

Mark starts of by giving a very funny overview of his very impressive career. He currently has a non-security security job at Microsoft running the MSDN subscription services department. Being away from security has given him room to think about information security more.

His talk is about 10 crazy ideas that might change the state of information security. These ideas all cost little money, but may have a big impact.

#1 – Adopt Chinese Medicine Business Model

In China the doctor gets paid to keep you healthy, not to cure you. There are currently actually two companies that are experimenting with this business model.

#2 – Stop Human Pattern Matching

Humans seen things they expect so see. The brain is wired to see what it is expecting to see. This is why optical illusions work, which was demonstrated to the audience with two illusions. Security people do his all the time. I have XSS, this is going to happen, this vulnerability will cause this worm.

#3 – Community Driven Statistical modelling

An example of this is http://freerisk.org. It allows people to input and consume financial modelling data. In the security world there is no data that will give us some predictable model of how security behaves. Wine quality can actually be captured in a formula: Wine Quality = 12.145 + 0.00117 * winter rainfall + 0.0614 average growing season – 0.00386 harverst rainfall. Where is the equivalent of security? Rubbish you say? Well, the formula for wine quality is actually used in the field now

Read more…