Assessing Network Security

Anton Chuvakin writes "I've read some pretty bad books on penetration testing; till now, nobody seemed to get this fun subject right! Good news - this time somebody did. Assessing Network Security comes to us direct from the bunkers of Redmond. Written by three Microsoft security researchers, the book provides a great overview as well as an in-depth coverage of assessing security via penetration testing ('pentesting'), scanning, IT audit and other means." Read on for the rest of Chuvakin's review of the book.

Assessing Network Security
author	Ben Smith, David LeBlanc, Kevin Lam
pages	592
publisher	Microsoft Press
rating	8/10
reviewer	Anton Chuvakin
ISBN	0735620334
summary	Great pentesting book

Assessing Network Security starts with a nice overview of key principles of security (definitely not news for industry practitioners, but nice anyway), and then goes on to defines vulnerability assessment, penetration testing and security audit. A critically important section on reporting the findings is also nicely written, and shows that the authors are knowledgeable, and interested in showing a complete security process rather than just the looking-for-leaks part.

The authors then go into developing and maintaining pentesting skills, including advice on choosing training and resources (nice for those starting in the field). The actual pentesting process is split into non-intrusive (combining the usual "intelligence gathering" with port scans, sweeps and various host queries) and intrusive tests (such as running a vulnerability scanner, brute-forcing passwords, DoS testing and others). Some entries seem to belong in both categories (such as sniffing) but are placed into the intrusive section, for whatever reason. Up-to-date content (wireless, Bluetooth and web assessment, for instance) is well represented.

The authors also include a fairly insightful social engineering testing section (touching on dumpster diving and other non-network assessment methods). My favorite chapter was the one presenting various case studies - examples of specific threats/tests against Web, email, VPN and domain controller systems.

Among other features that I liked in Assessing Network Security were 'notes from the field' sidebars with fun stories related by authors, and FAQs at the end of each section. On the down side, the book is somewhat Windows-focused (although it is amazingly vendor-neutral in most respects, considering the source). The book is also somewhat dry, although the sidebars provide some needed relief when the text gets too process-oriented at times.

Assessing Network Security is largely about methodology, but I'd have preferred to see a bit more technical content, since it is a 600-page volume. I think the checklists present in the Appendix are a great step in that direction.

Overall, I enjoyed the book and think it is both a great guide and a reference for most security professionals, especially for those starting to be involved with penetration testing.

Anton Chuvakin, Ph.D., GCIA, GCIH is a Security Strategist with a security information management company and maintains the security portal info-secure.org. He wrote Security Warrior and contributed to Know Your Enemy, 2nd Edition . You can purchase Assessing Network Security from bn.com. Slashdot welcomes readers' book reviews. To see your own review here, carefully read the book review guidelines, then visit the submission page.

FP - obilgatory NS at redmund?

They work at MS?!?!? what's the world coming to????

What about...

I've read some pretty bad books on penetration testing; till now, nobody seemed to get this fun subject right!

What about Kama Sutra?

Seriously though, this book is written by three Microsoft security researchers, I guess that said enough.

Is this a case of do as we say, not as we do.

Re:What about... Linux code?

Seriously though, this book is written by three Microsoft security researchers, I guess that said enough.

Assumption is the mother of all fuck-ups. You consider the security researchers incompetent because they are (or were) part of the Microsoft team?
So, because some Linux kernel coders make mistakes which lead to 'r00t3d' boxes, all Linux kernel coders are incompetent?
I think you're thinking a little bit simplistic here.

Re:What about... Linux code?

I think you're thinking a little bit simplistic here.

No, I think he's being a lot simplistic here, but that's just part of the larger mindset of Slashdot. "Linux GOOD! Microsoft BAD!" It's become the sheep's favorite thing to say during intense meetings on this Animal Farm we call Slashdot. You can lead a zealot to the truth, but you cannot make him think.

Re:What about... Linux code?

terrorists ...as they have no thought to any other side to a story than what they've been told to believe ... without letting any facts or alternative ideas get in the way ....

The story of The Man in the Tinfoil Hat [trilobyte-mag.com] is a poignant one here... The relevant quote is (emphasis mine)...

"I'm going to give you the best advice you'll ever get for survival in this field. Here it is: Never ask the lunatic if he's crazy. I repeat, never ask the lunatic if he's crazy. Of course he's going to say 'NO.' What do you expect? You're always going to get a self-serving, agenda-driven answer.

People are 'crazy' because they can't, and don't, see the lunacy that drives their lives. Never ask the lunatic if his illusions are real. Of course he's going to say 'YES.' If his problems aren't real, then he's crazy. But since he's not crazy, the delusions are real. In a nutshell (no pun intended): the more absurd the belief, the more deeply it must be held, the more aggressively it must be promoted and angrily defended if the patient is to see himself as right and sane. Get it?"

As the author of that article puts it further down:

"If MS (and all its staff) is not evil and incompetent, then the zealots are crazy."

I am a Linux user and advocate, but I still find these assertions silly...

Now, if they'd only...

put this kind of effort into securing their software.

Karma be damned...

"I've read some pretty bad books on penetration testing [...] Assessing Network Security comes to us direct from the bunkers of Redmond."

Nah, too easy.

"Microsoft security"

What is an oxymoron, Alex.

remember now

"...via penetration testing..."

remember guys, often times computers are like women.

this is not one of them

Re:remember now

My floppy is...uhh...stuck.

@microsoft.com

Kevin Lam [msdn.com],
Ben Smith [winnetmag.com],
David LeBlanc [winnetmag.com]

Windows

"...On the down side, the book is somewhat Windows-focused..."

Why am I not surprised? :)

A book on security from microsoft... ?

flamebait
Wouldn't that be sort of like George Bush writing an english book?
/flamebait

'somewhat dry'

is that a euphamism for 'the most boring book to hit the world since Inside OLE2?'

I never got more than a quarter of the way through that, i fell asleep every time i tried to pick it up.

i saw internal MS OLE training i saw Kraig Brockshmit did back in about 95 - jesus it was boring. we are talking boredom in an entirely new area of bordem till then undiscovered by man

When did M$ become a network device?

Maybe I'm missing something, but when did M$ OS become a network OS? I haven't seen any winblows routers or switches? The book should be Windows Penetration testing, and to fix that problem, just press the off button. I get sick and tired of hearing M@ included in the same sentence as network. It's a stinking OS, not a router. M@ couldn't compete with Cisco, Juniper, etc...

Re:When did M$ become a network device?

And it wasn't. Apperantly the drive crapped out a long time ago, and everything was running from memory. Btw the time I had to reboot it, the system wouldn't come back up. Finally got the company to cough up the money to get a Catalyst 3550 to replace it.

This is old news

A beta version of the book was leaked to the internet a year ago.

From the Author:

"Written by three Microsoft security researchers, the book provides a great overview as well as an in-depth coverage of assessing security via penetration testing."

From the referenced BN page [barnesandnoble.com]:

"I have been fascinated by leadership dynamics throughout my working career. [...] A concern is that we often get to hear the same leadership issues over and over again, yet leaders continue to lead with mediocrity and passiveness".

His Billness will not take this lightly! These guys can kiss their jobs at Microsoft Research goodbye! :-)

No grain of salt here

Being relitvely novice at network security I only have an extremely humble opinion but at the same time I must say that Mr. Chavaukin strikes me to be an extremeley adept man on this subject. Having just finished the Security Warrior I have learned a lot and I find his (and his co-author Mr. Pekari) insights and information to be extremely astute.
No, I will take no grain of salt regarding his comments about the book in question, untill I have achieved a decent status in the matter I will refer to Mr. Chavaukin's comments eagerly!

Re:No grain of salt here

As you say you are a security novice, I would suggest you take a look at the Common Criteria for information security evaluation [nist.gov].
This is what most security evaluations are assessed against. Threat Analysis/Risk Assessmeny (TARA) consultants are in high demand and can earn a lot of $$ these days.

Meanwhile, back at the MS penetration testing labs

Stewie: "Does anyone else smell astroglide?"

Can it be, O brothers

that from the city of the fallen, True believers come ?

Creation of the book

First, one takes all relevant MS security bulletins.
Next one invent some text around it.
Publish it and makes people pay for it.
So you make money from you're mistakes.

Funny:
I doubt most hackers who contributed to this masterpiece will ever see a paycheck. Probably they are mention in the hall of shame. (As ought to be)

Seriously, this might be a good book.

Great Source

M$ is the best source ever on security. After all they've made all the mistakes and created all the holes. Who else would know it better?

Security From MS Press

With all these folks pointing out the funny irony of all this, I'm here with what I think are valid questions --

What's up with creating an inherently insecure system and selling a book on security? Shouldn't they use that same advice to create better products? Almost like the conspiracy theory of making someone ill and then selling them the cure.

Maybe the book brings up interesting points and great ideas...but it's like asking me to believe everything Baghdad Bob said.

more to security researchers than lab work

Ben Smith (one of the authors) is also actively involved in Microsoft's private trainer newsgroups, and has always been a good source of information for security related questions that are way, way out of what "the theory" is normally limited to.
The utility of the book comes from not just spreading the word about security, but having to do so in forums and formats that require it to be relevant, useable and correct.
As a security consultant and trainer myself I can attest to the gap between theory and practice and the need to put security issues in to terms that are able to be applied in the real world.
Comments above that assume that just because someone works for Microsoft, they don't know how things work in reality are generalisations made out of ignorance or jealousy. This book is a good example that the truth about Microsoft employees, like security, is often misunderstood.

oh christ

someone posted story w/ the words "penetration" and "testing" following each other on /., that's like saying it in front a bunch of 15 year old boys.

Wait a minute...

Not all Microserfs are dolts

Most are actually quite intelligent, and like the money and perks you get when working for the best funded company on the planet..

I bet a lot of them do great work FOR the company, but its caught up and diluted by the much larger 'machine' that makes Microsoft go..

port scans? not in sp2

it's really too bad that microsoft classified port scanning tools (nmap) as "attack tools."
http://seclists.org/lists/nmap-hackers/2004/Jul-Se p/0002.html [seclists.org]

Open source tools?

This reminds of a question I've been pondering lately, which I believe would be on topic.

I have a box on a public IP -- speaking as a person who cannot devote 24/7 to security, are there any good automated tools to verify its "openness" in terms of security vulnerabilities?

I'm not talking about just potential root exploits and the like, but also about things like file permissions, which I find are hard to get exactly right on Unix (read: Linux with no special ACL stuff installed), where the file system does not support inheritance of security attributes.

Many Linux distros come with a script that's run nightly to report potential vulnerabilities, changed files etc. There are also tools like Snort and Tripwire. I also use Munin and check it daily for signs of DOS attacks and other suspicious activity (eg., a sudden increase in the number of listening ports).

What other automated tools do people here recommend?

Re:Oxymoron

No it makes perfect sense! Microsoft is showing us what not to do with our networks. It must have been a typo.

Nice amazon referrer link,

You money grubbing whore.

Here's a whore free link [amazon.com] and some healthy capitalist competition to boot [pcprotection.ca].