It should go without saying that pen testing is one of the most important pieces of an IT security shop's vulnerability management program. And yet it's something that was declared a dead art by Fortify Co-founder Brian Chess a couple years ago.
Since Fortify, recently acquired by HP, is all about scanning code for flaws at the development level, Chess' thinking was hardly surprising. The reality, many security practitioners have noted since that article came out, is that pen testing will probably always be needed. Better code security will certainly make the Internet a safer place, but vulnerabilities will always slip through. Meanwhile, some of the bigger weaknesses corporate IT practitioners need to worry about stem from misconfigured devices on the network that open attack vectors the bad guys can exploit even without software flaws.
And so, since pen testing will remain a critical tool in the security arsenal, it's important for security newbies to understand the basics of what goes into stress testing the network.
At last month's SANS Boston 2010 training sessions, SANS Institute President Stephen Northcutt ran through such basics in a talk called "SANS Security Leadership Essentials for Managers with Knowledge Compression."
This is the third article in a series on that talk. The first story dealt with the basic building blocks of vulnerability management and the second focused on tools of the vulnerability management trade.
Pen testing starting points
Northcutt started his students off with the following considerations:
* Pen testing is used to test the security of a network or facility
* It's more economical to conduct a pen test AFTER vulnerability scanning and remediation
* The most common problem pen testers come across is a denial-of-service, but sometimes the pen test team gets too carried away and attacks systems too aggressively. To that end:
* The rules of engagement must be clearly stated.
He noted that pen testing can get expensive in terms of manpower and time. Therefore, it makes more sense to go after the unconventional stuff and led the more obvious things slide.
"There's no point in having a highly trained team find obvious problems," Northcutt said. "First, run a vulnerability scanner and remediate those problems. Then consider an exploitation tool like Core Impact or Canvas Immunisec to fix those problems."
After those things are done, it's time to bring in the pen-testing team, he said.
Pen testing is used to test external perimeter security of a network or security itself, he added. It can take on many forms: the perimeter security of a building, server or network must be tested regularly. Otherwise, the bad guys just might do it for you with more disastrous consequences.
Pen test techniques
From there, the pen tester has a variety of tools at his or her disposal. Northcutt concluded the pen testing discussion by running students through a list of some of the bigger items:
* War dialing: Using a computer to dial thousands of phone numbers in hopes of gaining access to a system;
* War driving: Traveling around a neighborhood with a laptop propped on your lap, testing to see whose Wi-Fi networks you can latch onto, for example;
* Sniffing: Packet sniffing, for example, is used to appropriate valid TCP/IP network addresses by reading packets (units of data). Malicious code is then labeled with the trusted network address and blasted through the network unquestioned;
* Social engineering: One of the oldest techniques in the book, this is where a variety of tricks are used to dupe the user into doing something they shouldn't be doing, like clicking on a malicious link;
* Dumpster diving: Literally, the act of diving into the trash in search of sensitive data;
* Eavesdropping: Another old technique where someone's phone line is tapped or a recording device is hidden in an office.
* Radiation monitoring: This is a complicated process but can be effective in snagging images, data or audio from an unprotected source, Northcutt said. Radiation signal monitoring equipment is small enough to fit in a van. You could park the van outside an office window and actually reconstruct everything on your screen as you view it; and
* Physical security assessments, which run the gamut of testing locks to testing how easy it is for someone to access a restricted area.