The tools to analyze the performance of applications are varied and have existed in one form or another since the early days of UNIX. It is critical to understand how an application is interacting with the operating system, CPU, and memory system to understand its performance. This chapter will help you understand where the bottleneck in your system is occuring, and how to fix it.
Performance Tools for Optimizing Linux: Process-Specific CPU
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Great article, too bad that you have to purchase this information. Looks to be an interesting read.
oh yeah, first post
This is a great article, thanks for bringing it to my attention. The book will be even more interesting, without doubt.
Its all in the IO. And linux has no tools for measuring that. It also has no tools to delegate different IO to different processes/files etc.
Wake me up when I can actually improve my system performance, and not just minimize the number of cpu instructions.
Well…pardon me for saying this but the new Dynamic Tracing mechanism and tools offered in the Solaris 10 and Open Solaris OS are much beyond the general performance optimization commands or even specialized tools like LTT or DProbes in Linux. Unfortunately DTrace may only work for Solaris 10…as legally CDDL is not ment to cope up with GPL. Though has anyone compiled dtrace to make it work in linux ?
I am hoping that the book covers the use of tools like sysstat and Enhanced System Accounting, becuase the tools mentioned are fine once you determine you have a performance problem. With sysstat and ESA, you would create logs of your system(s) performance and analyze them for problems, then use the advanced tools if necessary.
Many problems do not immediately “crop up” and you might have to collect data for some time before you see the issue come up.
Linux can use CFQ with I/O priorities.