Articles / Maximum Linux Security

Maximum Linux Security

Maximum Linux Security's author is clearly ignorant of cryptographer Bruce Schneier's claim that "Security is a process, not a product." At its best, this book is a catalogue of useful security tools. However, very little context is provided for these tools. There is no discussion of particular vulnerabilities and how they are exploited, of network architecture and the difficulties inherent in TCP/IP networking, or of application-level problems.

Part one, "Linux Security Basics", covers some introductory material, physical security, Linux installation, and basic Linux system administration. Part two, "Linux User Security", covers password security and cryptography. Part three, "Linux Network Security", covers trojans and viruses, sniffers, scanners, and spoofing. Part four, "Linux Internet Security", covers various servers that can run on Linux boxes, including FTP, mail, Web, and firewalls, along with logging and disaster recovery. Part five, the appendices, are little more than shovelware listings of programs and vulnerabilities.

The book's intended audience is unclear. Too much basic system administration is covered to make the book palatable to all but the rawest novice administrators, but without detailed discussions of system architecture or security issues, it cannot catapult these readers into the realm of security-aware readers. Much of the book reads like a laundry list of past exploits; as these have all by definition been discovered and repaired, they are of limited utility. Had these exploits been used to illustrate security concepts or modes of attack, they might have proved useful.

Though the anonymous author is billed as an experienced hacker, methodology of attacks is rarely discussed. The exception is the coverage of password attacks which, while still important, are less important than they used to be. When biometrics are discussed, weaknesses of these systems are not mentioned, nor are possible attacks against them. Given the success of simple techniques in thwarting these systems, some consideration of them would have been apropos. Similarly, social engineering, one of the most fruitful of hacking techniques, does not even appear in the index.

One can certainly believe that the author has broken into a number of systems; this would explain the morbid focus on exploitable packages and the scattershot nature of his understanding. (Though the author's name is not given, his biographic blurb definitely indicates that he is male.) Cryptographic applications such as Gnu Privacy Guard are covered, but the essentials of public-key cryptography are ignored, as are issues of key distribution and certificate hierarchies.

Even his hacking knowledge, however, is out-of-date at best. For example, ethernet switches are recommended as a security measure to prevent sniffing traffic. In fact, for quite a number of years, software has been available that tricks switches into becoming hubs. The existence of these MAC flooders is not acknowledged, nor are even theoretical weaknesses suggested.

The chapter on network scanners assumes reader knowledge of the TCP protocol, including the three-way handshake and the various flags. It is not clear, however, from the laundry list of scanning tools, that the author understands the underlying mechanisms by which any of these tools operate. The information in this list is sadly out-of-date. It seems to indicate that the Jakal scanner's stealth techniques will prevent discovery of such scans. In fact, on most modern networks (and even with many tools that predate Jakal), the illegal flag combinations set on these packets are much more obvious than non-stealth scanning techniques.

Other information is equally out-of-date. The section on basic system administration relies on Linuxconf. Linuxconf has had quite a long security history and is no longer included in most Linux distributions. Even Red Hat, long one of Linuxconf's proponents, no longer ships Linuxconf. Should a user be able to download and install Linuxconf from scratch, the user will find the system administration tutorial to be an insult to his or her intelligence. The official Red Hat documentation does a much better job in any case.

Appendix A, the "Linux Security Command Reference", appears to be useful at first. It lists many security applications and provides descriptions. However, some items (such as "exports") are files, not commands, and for commands that are noted as "add-on" applications, no source is provided. The failure to list any means of acquiring these packages is unfortunate, and negates most of the usefulness of this appendix.

There is information in the book that would help a system administrator secure his or her system. However, that information is not consolidated, and is never expressed as a principle. Had the book discussed the principle of least privilege, the principle of minimalism, or the importance of promptly installing security updates, it might have proven useful. Instead, this information is scattered over more than eight hundred pages.

Had the book served to describe the methodologies of system intrusion or defense, advanced system administrators might have gleaned some useful information from the text. Instead, however, concentrating on outdated exploits and half-understood hacking tools, the book accomplishes nothing of note. It is not recommended, except as a curiosity for the library of a completist.

Recent comments

16 Apr 2003 06:43 Avatar wickedx

Re: Would you take advice from a moron?
I consider myself one of those people who was happy with Windows... but then found Linux. I recently migrated to Linux and have never looked back. A friend of mine, a guru as far as I'm concerned, gave me the book and said "this will get you started". True the book does not mention a lot of basic things new users should know... of course, who wants to read through a couple of chapters about 'this is the mouse and this is how you use the mouse'. Samething goes for a lot of the stuff that 'Should probably have been ATLEAST mentioned' but it was clear to me why it wasn't. The book was much like a laundry list, I admit, however! ..I did not finish the book and say to myself 'Ok, I really don't want to read anything else on Linux'. Quite the contrary, I wanted to read more on the areas, that where left out. Maybe the name of the book is the problem and not the content inside. Maybe it should have been titled: 'Beginners Guide to Maximum Linux Security' or 'The Road to Maximum Linux Security'. It's obviouse that there is a namming scheme, his other books, and the author may have been misleading the reader. In conclution though, I think the book is more usefull then not if you are beginner or an Intermediate to Linux, and will push you in the right direction, give you the questions to research. Yeah! What about backing up data? Hmmm Seems, like you should think.. MAXIMUM SECURITY before you go try to implement it....

08 Feb 2003 10:43 Avatar synr9

Re: Would you take advice from a moron?
Maximum Linux security I think is geard toward youngsters interested in becoming a uber linux hax0r, who just recently switched from Windows to gnu/linux. I come to this point of view because many Linux security administrators due to the needed technical understanding of running a gnu/linux server are already familiar with, or should be with the tools and concepts stated in the Linux Security Maximum book. The people interested in this book are usually not interested in how or why the utilites work the way they do rather they are interested in hacking/defacing websites easily. For these users this book delivers.


I have absolutly no credentials.

11 Nov 2002 12:41 Avatar jonlasser

Re: Would you take advice from a moron?

> His home page states:
>
> I've lost my old pgp keys
> (2047/0xDED5B791 and 1024/0xEC001E4D)
> due to a hard drive crash and a lack of
> backups -- new keys (...
>
>
> This guy is a system administrator?!?
> Trainer?!? Consultant?!?


Please note that the replacement keys are dated from 1998. ;-)

10 Nov 2002 10:26 Avatar linuxwolf1

Re: Would you take advice from a moron?

> No offense, but you are quite the
> dumbshit today. Since you seem to think
> your are a security god, tell me that
> you know what the Bastille project is.
> If not, why don't you look it up. Then
> look to see who the project co-ordinator
> is.

I have to agree with you, the reviewer is more then qualified to write what he did. Also if the starter of this thread had taken the time to read the reviewers resume, they would never have opened their mouth. Though I am some what curious as to why he decided to review a book thats 2 yrs old.
This 2nd edition as they call it was published in June of 2001, meaning it was submitted most likely in Dec of 2000/Jan 2001. I have the original book from about 3 yrs ago, and all this 2nd edition appears to be is a reprint of the original with a few added pages. So I have to agree with the reviewer, the book is lacking in many areas. Probably explains why LinuxConf is still covered in the book as when it was written redhat and others did include it.

09 Nov 2002 20:41 Avatar binford2k

Re: Would you take advice from a moron?
No offense, but you are quite the dumbshit today. Since you seem to think your are a security god, tell me that you know what the Bastille project is. If not, why don't you look it up. Then look to see who the project co-ordinator is.

Screenshot

Project Spotlight

Kigo Video Converter Ultimate for Mac

A tool for converting and editing videos.

Screenshot

Project Spotlight

Kid3

An efficient tagger for MP3, Ogg/Vorbis, and FLAC files.