Tracking Hackers

Anton Chuvakin submitted this review of Lance Spitzner's Honeypots: Tracking Hackers. Spitzner has previously contributed to a book and many online documents about the Honeynet Project. Chuvakin starts off, "If you liked "Know Your Enemy" by the Honeynet Project, you will undoubtedly like Lance Spitzner's (the Honeynet Project founder) new book "Tracking Hackers" much more. In fact, even if you did not quite like "Know Your Enemy", you will likely be deeply impressed with the new book on honeypots and their use for tracking hackers." "Honeypots: Tracking Hackers" author Lance Spitzner pages 480 publisher AWL rating 5/5 reviewer Anton Chuvakin, Ph.D, GCIA ISBN 0321108957 summary using honeypots to track electronic intruders

The structure of the book is different from the "Know Your Enemy": Lance starts from the very beginning - namely, his first honeypot penetration experience and then goes on to talk about all aspects of honeypots. In-depth and structured background on honeypot technology is provided. Honeypots are sorted by the level of interaction with attacker they are able to provide.

In addition, the book covers the business benefits of using honeypots. By classifying the value of honeypots into prevention, detection and response (exactly as done in Honeynet Project white papers) Lance Spitzner analyzes the honeypot technology contributions to an overall security posture. Also, the book describes the differences between the research and production honeypots and demonstrates the benefits of both for various deployment scenarios.

A good part of the book is devoted to particular honeypot solutions: 'honeyd' by Niels Provos and several commercial honeypots with detailed explanation of how they work. For example, there is a clear description of ARP spoofing and how it is used by the 'honeyd' honeypot daemon. An interesting chapter on "homegrown" honeypot solutions (such as the ones used to capture popular worms of 2001) sheds some light on the simplest honeypots that can be built for specific purposes, such as to capture a popular attack by means of a simple port listener. Use of UNIX chroot() jail environment for honeypots is also analyzed.

Of course, a special chapter is devoted to honeynets - Project's primary weapon in a war against malicious hackers. The Generation II (GenII) honeynet technology is first introduced in a book. The chapter not only lists honeynet deployment and maintenance suggestions, but also talks about the risks of honeynets.

Another great feature of the book is a chapter on honeypot implementation strategies and methods, such as using NAT to forward traffic to a honeypot and DMZ honeypot installation. The information is then further demonstrated using the two full honeypot case studies, from planning to operation.

What is even more important, maintaining the honeypot architecture is covered in a separate chapter. Honeypots are a challenge to run, mainly since no 'lock it down and maintain state' is possible. One has to constantly build defenses and hide and dodge attacks that cannot be defended against.

"Tracking hackers" also has a "Legal Issues" chapter, written with a lot of feedback from the DoJ official. It dispels some of the misconceptions about the honeypots such as the "entrapment" issue, summarizes wiretap laws and related data capture problems.

The book describes an almost cutting edge of the honeypot research and technology. To truly get the cutting edge and to know about the Honeynet Project latest activities in detail, wait for the second edition of "Know Your Enemy" (coming out next year). In "Tracking Hackers" Lance makes some predictions about honeypots in "Future of Honeypots" chapter. Honeypot-based early warning system and distributed deployments, analysis of new threats and expanding research applications, making honeypots easier to deploy and maintain are all in this chapter.

To conclude, Marcus Ranum's enthusiastic preface is not an overstatement, it is indeed a great book for both security professionals and others interested in this exciting technology. While I was already familiar with most of the information in the book, it was a fascinating read! This is the kind of book you don't want or even cannot put down until the last page is turned.

Anton Chuvakin, Ph.D., GCIA (http://www.chuvakin.org) is a Senior Security Analyst with a major security company.

You can purchase Honeypots: Tracking Hackers from bn.com. Slashdot welcomes readers' book reviews -- to see your own review here, read the book review guidelines, then visit the submission page.

Comments

Use of honeypots

[TrueKonrads]

I don't understand why everybody suddenly wants to have a honeypot? You are not a fed and you won't mail them "Look there is a ssh scanner poking my network, go arrest them, oh brave ones". Stick to snort with decent rulesets and be set. Add kernel patches and protect your butt from exploits. Honeypots seem to be overhyped.

Re:Use of honeypots

[Hayzeus]

Good point. I suppose a honeypot might be a useful tool for organizations like banks and other places that might be targets for organized attempts at breakins-for-profit. In such a situation, a honeypot might provide both a good early warning system and keep the intruders around long enough for the feds or whomever to do something useful.

Or maybe not. Otherwise, I pretty much agree with your post. Not only does this seem like an expensive waste of time, but potentially dangerous if some idiot comes along and misconfigures the honeypot (or if an idiot put it up in the first place). Unless your workplace is idiot-free. Then it would be OK.

Re:Use of honeypots

[Tetsu no Chef]

I don't understand why everybody suddenly wants to have a honeypot?

Perhaps they just sound like really good idea on a lonely Saturday night.

Re:Use of honeypots

[kasperd]

potentially dangerous if some idiot comes along and misconfigures the honeypot

This is all about where you place the honeypot. Place the honeypot where it has no more access to your important network than a random computer on the internet would have had. In that case your network is no more vulnurable with the honeypot than without.

Placing the honeypot in an important network behind your firewall is plain stupid. Placing the honeypot outside the firewall is acceptable. Placing the honeypot in a DMZ zone with nothing else is also acceptable, if the firewall is configured correctly. If the extra leg on the firewall causes more bugs in the setup, two independend firewalls would have been better.

Re:Use of honeypots

[Oztun]

Actually I use several honeypots (RH 7.3, OpenBSD 3.1, and Solaris 9). I patch all known exploits in hopes of finding someone using unknown new ones. So far I have captured a few over the years.

As for using a honeypot with known vulns. I think it is an ok way to learn but a dangerous one. My honeypots only capture the exploit and then lock out the cracker.

How many arrests?

[Animats]

Is this actually leading to arrests? If so, we need more honeypots; if not, it's a waste of time.

There's a tendency in what passes for a computer security community today to focus on the most numerous threats, rather than the most effective one. Maybe there's only one hit a month from the guy who's breaking in and reading your credit card number file, but that's the person you need to find.

Re:How many arrests?

[P!Alexander]

If so, we need more honeypots; if not, it's a waste of time.

Perhaps a honeynet on it's own is not terribly useful to the general population. However, the documentation, case studies and other material provided by this SPECIFIC honeynet project has enormous value. Their whitepapers are a very thorough look at real life hacking situations. I could see university classes formed based upon the research and publishing they have done.

As everybody knows, theory is great but real world examples can be just as, if not more, valuable. And here we have a project that has provided those examples.

Re:How many arrests?

[freuddot]

Is this actually leading to arrests? If so, we need more honeypots; if not, it's a waste of time

This is Soooo Year 2002 thinking.

How 'bout
Is this actually leading to less cracked systems? If so, we need more honeypots; if not, it's a waste of time

Guys, wake up, the whole point is not to arrest people doing bad things. The points is stopping those bad things from happening.

Re:How many arrests?

[Telastyn]

Actually, the question is how many firings this leads to.

More common than not is deploying the honeypot internally as something like "finance server 1" or something similarly obvious. Usually it's not someone breaking in to get the credit card file, it's someone already in who's digruntled, or perhaps a little too curious for their own good.

Re:How many arrests?

[bourne]

Is this actually leading to arrests? If so, we need more honeypots; if not, it's a waste of time.

Honeypots provide a way for people to learn how to deal with an incident on the real servers they protect. Just like emergency personnel hold disaster preparedness drills, just as the US Military stages large red team exercises in the desert, just as medical students use cadavers to learn what they're doing before they cut into a live patient.

Let's face it - when your database server gets compromised, do you want the guy responding to thrash around, destroy evidence and erase tracks, and generally screw up the response? Or do you want him to do a careful, correct job?

Honeypots are where admins learn NOT to run around like a chicken with their head cut off.

"honeypot penetration experience"

[pr0nbot]

honeypot penetration experience

*sigh* I remember mine fondly too...

Silly ole' bear

[Apocalypse111]

...misconceptions about the honeypots such as the "entrapment" issue...

You mean Pooh never really got his head stuck in one of those things? :)

That's CRAKCERS not HACKERS

[Neil Watson]

Get it right. Crackers are the criminals. Hackers are law abiding citizens who are also computer experts.

Re:That's CRAKCERS not HACKERS

[TedCheshireAcad]

Honeypots attract leet script kiddies, not crackers. Experienced crackers do not give a shit about a soon-to-be-owned box sitting on a T-1 line.

-my $.02

Re:Use of honeypots - for educational purposes

[Anonymous Coward]

Honeypots are very interesting for educational purposes. If you're a security professional (or want to become one), then using a honeypot can provide 'real-world' experience with dealing with crackers, analyzing logs, studying behavior etc. This way if you end up with a job in computer security you won't have to look at your boss after a security breach and shrug "I dunno, I've never actually seen this before". There is a difference between reading and doing. You can learn a lot from reading about something, but you'll learn it in a different way when your fingers are actually on the keyboard and it is actually happening.

Academic Value?

[Vagary]

Do honeypots have any value for teaching security experts? Could the study of crackers and cracking techniques ever belong outside the Sociology Dept. at a university?

I can certainly envision course projects surrounding the analysis of a real honeypot or perhaps a system that has been compramised by the teacher. But would this actually help the students or would they be better off learning in a more theoretical fashion? (Because cracking is too variable and changes too quickly for the study of specific techniques to be of value.)

Hit count, not arrest count.

[Anonymous Coward]

Arrests are meaningless.

I want to know how many kneecappings and crushed hands have resulted from skript kiddiez going after the wrong targets.

Teach the ankle-biters to steer clear of networks owned (in all senses of the word) by Guido and "doze computer-talking guys dat be wit us, watchacall'em, geeks or sumthin? fuggetabout'it."

Hit count takes on a whole new meaning in this context,
Capice?

Disclaimer IANAMM (I am not a made man, but my brother-in-law was, at least until they found him stuffed in the trunk of his caddy)

tracking hackers is easy for the slashdot crowd

[Jacer]

just find any of us and hand him or her a mirror

Real world applications? (Was "Academic Value?")

[Nonesuch]

Can the study of the behavior of attackers ever NOT be of value to the defenders?

Detectives study the behavior of criminals, The FBI studies the behavior of terrorists, ROTC students study the behavior of attacking armies, and network security analysts study the behavior of crackers.

Not every cop is a "Criminologist", not every sysadmin needs to be a "Security Analyst".

Got one here

[r_j_prahad]

My employer is hosting an extremely intricate and rather sizeable home-grown honeypot solution.

It was supposed to be our corporate web server, but our sysadmin is a dolt.

Pre-Order Only

[dehex]

You have to place an advance order and wait a month and a bit till it comes out.

Amazon.com has a cheaper price ($31.49) and an early release date (Sept 20th) than Barnes&Noble.com ($35.99, release Sept 27th).

Looking forward on reading it :-)

Entrapment Issue?

[The Wing Lover]

What entrapment issue? Entrapment only happens when a person is convinced by a law enforcement official to do something that they wouldn't ordinarily do.

If someone gets stuck in a honeypot, he ordinarily would've been attempting to scan my system...