Customer Reviews

UNIX Internals: The New Frontiers

5 star:		(15)
4 star:		(2)
3 star:		(1)
2 star:		(0)
1 star:		(0)

 

 

I am very surprised that this book is largely get overlooked by Unix society. I read this book for over 2 years, I'm still frequently revisit it. The beautities of the book are 1. The information is up-to-date, a lot of information is summarized from latest Usenix proceedings 2. Topics involved are moderate to advanced, assuming you have enough basic knowledge on Unix Internals, such as those from BACH's book. So it focused on SMP, threaded kernel, virtual file system/journal FS. VM/Fs integration. It fillin the gap for those reader who need to know the progress of current Unixes of current Unixes since mid 80's 3. Solutions and comparisions. The author detailed different solutions on flavour of Unixes, including BSD, SVR4, Solaris (although derived from SVR4), OSF, and the legend MACH. This is the gem of the book, it outlined pro and con of each solution that brodened user's mind. 4. Well balanced contents. The author keep well balance between mechanism and implementaion detail, so that it is detail enough for reader to grasp the essence algorithm but not embroil into topic-irrevalant mess that may distract reader's attention.

This is a fasinating Unix internal book on market that can take the reader into the Unixes forest, in the meantime, keep readers a clear view about the forest evolution and landscape.

The book is not for Unix beginner.

A certain small and select set of books can be found in any serious programmer's library. "UNIX Internals" is one of those books.

I originally encountered "UNIX Internals" in my undergraduate operating systems class. At the time, I liked it, but I didn't really appreciate its full beauty until I re-read it with a few years of operating systems experience under my belt. I work as an operating systems programmer for a living, and whenever my knowledge needs brushing up, I go back to this book.

Uresh Vahalia does an excellent job of comparing and analyzing the approaches taken by different operating systems, rather than merely describing them. His deep understanding of the topic is what really sets this book apart. In addition, it is well written, conveniently organized, and thoroughly indexed. If you really want to understand operating systems, this is the perfect book for you.

As others have noted, this book is not for the beginner. You should probably have a minimum of three years of computer science experience before picking this book up.

If you want to know how the Unix kernel works, and doesn't work, you need this book. Published in 1996 about "The New Frontiers", most of the "new" ideas are in current (1999) production systems. Vahalia is an excellent writer. The book is well organized and clearly written. Covers Sys V, BSD, Mach kernels and design tradeoffs. Threads, multiprocessors, IPC, lightweight processes, etc, it's all here. Excellent bibliography and references.

I found this book to be extremely helpful in studying for my OS PhD prelims. It really lays out the issues and presents solutions taken by SVR4, BSD, Mach, and Solaris in a clear fashion. Excellent detail. The Mach stuff is great, but it looks like it was almost cut and paste from the research papers. This book isnt for novices. I recommend using the BSD4.4 book as well. Both books dont look at Windows at all (for obvious reasons), so a look at the Dinosaur book (Operating System Concepts) is helpful as well for the case studies. You should have a
clear understanding of OS after finishing this book.

As someone whose occupation is providing support for developers on the UNIX OS, this book is worth it's weight in gold. It explains in sufficient detail all of the major features of a modern UNIX kernel, including kernel threads, user-space threads, signal generation and handling, scheduling, IPC, filesystems, etc. I've read just about every popular book printed on UNIX internals and this is my favorite. It reads easily and the description is great. Importantly, is organized so you can skip around and get the information you need.

Congratulations to Dr Vahalia! I think you've written the seminal work for UNIX students for the next few years.

I have to admit I couldn't finish reading this book because a lot of content is beyond my knowledge. But I read quite a few chapters and know this is something I'll come back to in the future, once I finish David Curry's Systems Programming for SVR4. I wish Mr. (or Dr.) Vahalia gave us more examples in the book; something like "Experiment" in David Solomon's Inside Windows NT would be the best. Exercises with answers or hints would also be very good. (Humans are not computers in that we have to practice to learn.) In a nutshell, this book is a masterpiece every word of which comes out precisely.

This book is a great read. All covered topics have been dealt at in depth. The chapters on filesytems and memory management stand out! Altogether a great book but takes some doing to read from beginning to end.

Having read the first edition am eagerly awaiting the 2nd edition...

I saw this book on some of my coworker's bookshelves. So I bought it and have been very happy reading it. It covers a lot of newer Unix technologies. It's not for beginners but if you have some Unix experience and wants to know more the design, don't miss this book.

Uresh Vahalia explains complicated Unix concepts in a very comprensive way. This book is easy to read and goes deep into unix internals. Vahalia explains very well memory managment, scheduling, ipcs, LWP, file systems. If you develop under unix, this book will give you the knowledage to take adventage of the underlying system.

Initially I didn't like this book, until I really had to use it. Then I ended up reading a lot of it, and realised that it really does do a better job than most things. At first glance for me I didn't like it but now I feel it gives a very good technical overview of the topics covered, and I find the chapters on IPC and Synchronisation are hard to live without, unless you can find better. Now, this book covers a great many topics, but it covers them WELL. You surely can and will find books that deal with some of the same topics, the entire time, however this is a good broad overview, and as you get deeper and deeper into Unixoid kernels, you'll find more and more things covered by Uresh that are covered shortly and sweetly, and occasionally, have very good long explainations. He also shows multiple methods as used in different Unixoid kernels for a lot of the topics, which is really really nice, especially once you start thinking about interoperability. What I've found I like the most though, is that he doesn't hide behind opinions and rhetoric, he gets right to the point.