In the past few weeks, I’ve received two new titles on Linux performance Tuning; Performance Tuning for Linux Servers from IBM Press and Optimizing Linux Performance: A Hands-On Guide to Linux Performance Tools from Prentice Hall Professional Technical Reference.
I’m quite happy to see books of a more advanced nature coming out about Linux. Beginner’s books are nice, but the real need for documentation is on the more advanced topics, where man pages and HOWTOs aren’t quite sufficient to get it done. Performance tuning, in particular, is of heavy interest to admins who are deploying or thinking about deploying Linux, and they need to get the most bang for their buck.
In the past few weeks, I’ve received two new titles on Linux performance Tuning; Performance Tuning for Linux Servers from IBM Press and Optimizing Linux Performance: A Hands-On Guide to Linux Performance Tools from Prentice Hall Professional Technical Reference.
I’m quite happy to see books of a more advanced nature coming out about Linux. Beginner’s books are nice, but the real need for documentation is on the more advanced topics, where man pages and HOWTOs aren’t quite sufficient to get it done. Performance tuning, in particular, is of heavy interest to admins who are deploying or thinking about deploying Linux, and they need to get the most bang for their buck.
Monday, August 1st, 2005
Share This:
One thing to point out about both books. Despite the HP and IBM branding, respectively, both titles are appropriate for anyone using Linux on any brand of server. The books are both Linux-specific rather than HP or IBM-specific, and that’s a very good thing — so don’t let the branding throw you.
The good thing is that both books spend a lot of time teaching the reader how to use open source tools like sar, OProfile, and others to profile system performance and find out what bottlenecks exist on a system.
Where they’re a bit weaker is in actually instructing the reader how to improve system performance. Obviously, if one finds that disk access is the problem, then the answer is to try to beef the system up to better handle I/O intensive tasks or to try to spread the load out a little more (or both). But, there’s not as much direction on actually tuning a server for better performance as I’d hoped.
I also feel that the IBM title does a much better job of explaining the components that affect performance and giving the reader a more thorough understanding of the Linux subsystems. For example, I really liked Chapter 8 in Performance Tuning for Linux Servers, which goes into heavy detail on the Linux scheduler.
That’s not to say that the HP book is not a good title. It is, but it spends…
Please log in to view this content.
Not Yet a Member?
Register with LinuxMagazine.com and get free access to the entire archive, including:
