Ludwig Isaac Lim <[EMAIL PROTECTED]> writes: > Hi: > > I like to ask a few question regarding the "top" > command. Suppossing I am monitoring the performance of 2 > programs namely "trigger" and "db_server" running on a > single processor machine, and using top command gives me > the following output (Some imptertinent stuff removed). > load average : 0.55, 0.51, 0.47 > > Cpu(s) : 13.6%us > PID %CPU COMMAND > 2768 24 db_server > 2773 12 trigger > > > 1) How is 13.6% related to 24% and 12%? Is the actual CPU > utilization of db_server equal to 24% x 13.6%? How do I > interpret the %CPU column(sometime there are entries in the > CPU column that exceed 100%)
The 13.6% measure above refers to the amount of time spent by the CPU(s) in executing user space code. In brief, there are three summary CPU numbers: %user, %kernel, and %idle; which represent time spent by the CPU in userspace code, kernelspace code, and in doing nothing (the idle loop). The 24%/12% per-process CPU refers to how much CPU each process is using in relation to other processes. > 2) How do I interpret the load average entry found in the > upper right side of a top output? The values usually mean the number of processes in the kernel runqueue (i.e. waiting to be run) as measured in 1-, 5-, and 15- minute averages. In other words, the first number is the average number of processes in the kernel runqueue as measured in one minute intervals. > 3) Is vmstat a better tool (i.e. more accurate) as opposed > to top for monitoring overall performance ? vmstat is more useful for measuring I/O performance and memory access performance, IIRC. -- JM Ibanez Software Architect Orange & Bronze Software Labs, Ltd. Co. [EMAIL PROTECTED] http://software.orangeandbronze.com/ _________________________________________________ Philippine Linux Users' Group (PLUG) Mailing List [email protected] (#PLUG @ irc.free.net.ph) Read the Guidelines: http://linux.org.ph/lists Searchable Archives: http://archives.free.net.ph
