Will we see any more Linux System Administration books in the future?

A reliable source says the days of Linux sysadmin books have ended. If you can barely sell 5,000 copies, then why bother? Programming books continue to sell at a fast pace, so you can guess where job demand in IT has gone.

"Linux System Administration" LSA (ISBN-10: 0596009526) written by me and Bill Lubanovic took a monumental effort. I signed the first contract in the fall of 2004 and O'Reilly released it in March 2007. Granted it’s a high level admin book and not another command line collection of man pages. With the growing number of Linux power users wanting to make the leap to administering servers, shouldn't the demand translate to sales?

Fortunately, LSA isn't another book jammed full of man pages and the reviews are good. My reliable source thinks we might have published one of the last commercial books of its kind.

If you remember "The Whole Internet Users Guide" published in 1992, then you know it was one of the largest selling technical books of all time. Its appeal has also waned. System administration books may suffer the same fate. Who wants to buy a book when people believe they can find everything they need by doing a search on the Internet?

The end of sysadmin books lies in the myth that you can get everything you want by searching the Internet. From an author's experience, I can say it's not true. If you could put Linux servers together and create an enterprise then why did it take Bill and I so long?

Linux sysadmin documentation is incomplete in all areas. If you wish to argue against that, then you can write your own sysadmin book and see if you can get what you need to build out a Linux infrastructure.

When O'Reilly finally released LSA, I needed to return to the real job market. I did the usual thing people do when they start to look for a job. I posted my resume on Monster and Dice. I put up a sysadmin resume and just for curiosity, I put up a technical writer's vitae. I got four hits on the sysadmin resume and 83 on the tech writer resume.

Guess what I'm doing today. I'm a tech writer/business analyst for a division of a global Fortune 1000 company. I thought writing the book on Linux Administration would have caught someone's eye. Instead, the need for documentation caught more eyes.

Now, if you wonder why my reliable source says the days of sysadmin books has ended, my experience says maybe it has. Build servers and learn to use and administer them, please. But remember the importance of good and timely documentation.


Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

We think this is an

Home Refurbish Course's picture

We think this is an inspiring article.

Books Vrs Internet

ADAC's picture

I don't think good overview books will ever go out of style. When ever I hit a new subject I start by buying a book on it. I read some sections, skim others, but this gives me a good start on how to use the software, os, etc.

Later, I go on the internet for specific problems and topics.


ord's picture

As long as networks and standards changes there is a need for new books. Books with complete overviews are good for beginners. An expirienced admin is more interested in a small ebook or only chapters about his special problem.

Quality books will sell

Computer Help Forum's picture

I think a quality book on Linux System Administration still worth to be purchased.
When I need to solve an unknown tech problem, first I check my books, then if I don't find a quick and current solution, I search the specific topic on Internet.
I find more practical the use of paper book over ebooks or electronic documentations.

Foundational information

Mark Dean's picture

The Sys Admin books are great foundational information-especially to those new. The best ones I have used (besides the requisite Essential System Administration) are the Linux System Administration (as has already been mentioned) and its Unix version cousin. Those provide a great framework and are recent enough to still be relevant. Especially what was helpful was the detailed information on boot strapping, rc scripts, and networking. For someone who came from the Novell world, it was really informative. Much of what is going on in my world of Linux, HP-UX and VMware ESX administration and engineering are really the same as has always been done. I use the Internet for quick refreshers but my books are always an arm's length away. I agree with the post about the abysmal man pages-with no standard so you get some who simply repeat the --help switch and others who write a tomb on a 20k executable utility.

I guess what I'm really saying, is another admin book really necessary? It's like the Unix/Linux Bibles-once you have one, do you really need to pony up $40-$60 for a new one just because Red Hat as moved from 4 to 5? And cover all those same *nix subjects that have changed little if at all in the last 10 years? Maybe, but what is better is a diff version. Show me what has changed and I'll buy that one.

Good books are hard to write and publishers like profit

Adam Trickett's picture

The problem is that a good book takes a lot of effort to write. Effort means it takes time and costs money to do.

Publishers face high manufacturing and distribution costs - paper isn't cheap to buy or more around.

There is a pyramid structure of buyers, many are happy with lowest grade/tier of book, by the time you get to high quality advanced books the target audience is getting awfully small.

Open source is also a fast moving field, the cutting edge moves daily, who wants a book about a two-version old version of Ubuntu? customers don't and neither do book sellers or publishers.

Google is your friend...

I have to agree that the odds are against you.

Publishers seem to be very keen on pushing high volume "parrot fashion" books that beginners often buy and because they age very quickly they are interested in only the highest volume topics. The world is awash with idiot books for PHP, JavaScript, Windows, Linux - very few if any are any good.

There are lots of good highly technical books out there but I'd have to agree with you that a topic like Linux Admin isn't one that there are a lot of options for.

Sysadmin Books

Les's picture

Hi Tom!
I bought your book, and while reading through the thread, there are some things I agree with. Specifically, the "medium" comment - I use the O'Reilly Safari system to access their entire catalog because at times it's much more convenient to copy/paste code samples this way.

As far as the materials out there being dated, I would disagree - I recently saw the "Red Hat 5 Unleashed" titles, and "Beginning Ubuntu" now in it's 2nd edition.

I think the problem with a Linux System Administration title is what distribution do you focus on? Do you try to broadly cover them all as the other sysadmin title did? Personally, if a book titled "Ubuntu Server Administration" was written, I would be the first in line.

Smaller Task Specific Books

Anonymous's picture

Maybe the last book on general sysadmin/IT stuff has been written, I doubt it. Perhaps the market should shift to smaller task specific books. A good example of this would be the Pragmatic Fridays series of "ebooks". The three books listed are 90 pages or less and cost only $8.50.

These books focus on a specific task. They're shorter and only available in pdf format. Instead of a general treatment of sysadmin topics, smaller books that focus on something like "Using LDAP for Account Mgmt on RHEL 5 and Ubuntu" may be more useful. Another could be "Network Printing Using LPD and CUPS"

Speaking from experience

John Lewis's picture

I am a self-employed "sysadmin" and I must admit that in a way I feel guilty for not purchasing these kinds of books, but I have spent the last 7 years learning Linux, trawling the internet for information and learning why things have to be done a particular way. Some times it is better to do things the hard way.

When I was a Windows admin I had to constantly buy books and get certification, I got sick of it.

Now between the man pages, wikis, forums and "Rute" (which I found a great resource on how to script) I have all the information I need to build a Linux server that will do anything a small business needs.

I appreciate that you are putting all of this information into one place that is easy to digest. However I think the average Linux admin is motivated enough to find salient information themselves.

I can predict that if the uptake of Linux becomes stronger and all the "doze" admins start to desert the sinking ship, you can be sure that sales of books like yours will go through the roof.

Until then though you are trying to sell to an already well informed and motivated minority.

Best of luck with the books in the future and don't give up!

Linux users tend to be smart...

Jim Smith's picture

...and they are usually able to figure things out on their own. The entire culture of free software is one of rugged individualism that encourages people to pick themselves up by their own bootstraps. Not everyone needs a book to figure things out.

It seems the marketplace has spoken on this issue: the world has enough $30 books that explain how to use free software.

Outdated subject matter

rj's picture

It seems the marketplace has spoken on this issue: the world has
enough $30 books that explain how to use free software.
And which are basically outdated by the time they hit the store shelves. For instance, how many books were there that were hyping Redhat 6 as the hot new thing when people were running Redhat 7,8 and 9?

Not necessarily, I've learned

Anonymous's picture

You raise a good point. However, there is another side to this.

Eight years ago (that'd be 1999), I read the book "UNIX System V Release 4: An Introduction", vintage 1990. That was my first UNIX book. Turns out that nearly everything in that book was directly applicable to the (modern in 1999) Red Hat Linux 5.2 and Caldera Linux 1.3 that I was learning at the time. That "old" UNIX book from 1990 helped me *greatly* to understand GNU/Linux. Same with the O'Reilly book "Practical UNIX Security" (yes, the first version from waaaay back in the 1990's), which I read in 2002. It was still applicable to thge modern GNU/Linux of that day (Slackware 8.x, Red Hat 7.3).

So, I'd say that you're right; the development pace on Free Software can be very fast indeed. But the fundamentals don't change. It's all still applicable. I find that I still reference--yes--that now-17-year-old book about UNIX System V...and not infrequently, either.

It's the medium, not the content

Roy Schestowitz's picture


The problem can be generalised. It's like newspapers. The problem is not the content, but the change in terms of content delivery methods, which may or may not involve interaction (corrections, updates, etc). I have seen this as I worked on reading journals, attending conferences, writing papers etc. The world changes. Information is exchanged in other ways.

Best wishes,


Technical Books

Jim Steichen's picture

The problem with books on technological subjects is that they're obsolete by the time they're published. Also, some Linux books only focus on the more popular distributions. If you don't use Debian, Red Hat or SUSE, chances are that the book you're looking for does not exist. That being said, the Linux Administration Handbook, 2nd Ed., by Evi Nemeth, et. al. (Prentice Hall, 2007, ISBN: 0-13-148004-9) that I have is still useful. It's still sometimes quicker to look for info in a book rather than the Huge haystack that the internet has become. Books as a source of information are not going away anytime soon as long as information is difficult to find on the internet.

wiki is just better for this ad hoc answers

Anonymous's picture

I dunno, I always find what I need in either man pages or online. Wiki is much better than a static book for this kind of content. I find it nice to get multiple opinions on an approach also; bookmark them in delicious, and have them there for later (my own index if you will). The comments in blogs/wikis are useful for seeing if the approach actually worked. No way I'll buy a static book for ad hoc issues, it's not rich enough medium. Nice way to get an overview, but that's about it.

OpenBSD's situation may provide a hint

Dan Farrell's picture

OpenBSD (among other things) is routinely praised for it's thorough and well-written documentation in the form of it's man pages (which make linux man pages look like a joke) and the published faq. For standard situations, this and the mailing list handle 90% of what most learning and mid-level admins need.

So there are less than even a handful of books directly related to OpenBSD, and not that many more (recent) related to BSD in general.

If the documentation is that good for everyday situations and common issues... why do you need a book?

It's true, however, that for enterprise design issues and large projects, no man page or faq is going to be appropriate... but how many books are written for that audience?

Maybe the publishing world will wake up and realize that even a decent "how to be a good admin" approach isn't enough anymore for a book- you need to tackle topics like comprehensive security design, or enterprise architectural considerations... IMHO, people will still find value in books of that nature. All of those burgeoning admins that are trolling the Internet for quick fixes will appreciate the 'bigger picture' books (not picture books lol) that will 'bring them to the next level' of system administration (read- take them to the next position in the corporate ladder.)

i think it`s more about too much information

Anonymous's picture

Usually if i have a problem (regarding linux, open source, programming, or anything on that matter) i check the internet for available information on the subject, and most if not all the times i end up adding 25 bookmarks to my browser without actually solving the problem

I realize the books are expensive, not very practical (for space and waight) and just not user friendly.. but if i`m certain that the book is specially writen to attack at not-so-common problems.. i foud it to be a good investment

As long as it is not one of those *nix for dummies type of books, i have the internet to find reference like that.. i would like to see very specific (and abundant) information when i go to a book

But than again.. that`s just me.. greetings

So, are you saying that we need somethng like Microsoft's MSDN ?

Computer Keyboard Reviewer - Alex Smith's picture

A centralized place for discussion and documentation is the best idea, but it requires a lot of initiates and collaboration of Linux experts. I truly hope this will become true.