Optimizing Linux Performance: A Hands-on Guide to Linux Performance ToolsThis book provides readers with a quick method to understand the basics of the Linux performance tools and allows the tools to be used immediately. It covers the most recent (and powerful) of the Linux performance tools. For each tool, this book provides information about the tool's purpose, the tool's options, and an example of how to use the tool. It provides readers with a general method for tracking down Linux performance problems while requiring a minimal knowledge of the underlying system, allowing the reader to be more productive and less intimidated by the performance problem at hand. Finally, and most importantly, the reader is shown case studies that use performance tools and methodologies to diagnose and fix a performance problem on example open source applications. These case studies will save readers a large amount of time, allowing them to simply adapt a technique that already works rather than inventing their own. |
Common terms and phrases
1sof allocated amount of memory application application's backtrace bash binary buffers bytes cache cachegrind chapter Command-Line Options Option configuration context switches CPU usage debugging determine disk I/O display eth0 etherape Ethernet executable ezolt txt ezolt@wintermute fgrep figure filter following command line function g_type_check_instance_is_a GIMP gkrellm gnumeric go to Section Goto gprof ifconfig invoked iostat iptraf Linux kernel Linux performance tools memory usage memprof monitor mpstat nautilus netstat network device network traffic objdump opreport oprofile optimization output packets particular percent performance hunt performance investigation performance problem performance statistics performance tools Performance-Related Options physical memory pixel pop-up menu prelink processor procinfo provides quick mode resident set size root@localhost samples script shared libraries shared memory shell script shown in Listing shows sockets spent strace subsystem swap system calls system-wide Table total number track valgrind vmstat