IBM DS6000 Series Server User Manual


 
Appendix A. Operating systems specifics 303
Example: A-3 SAR Sample Output
# sar -u 2 5
AIX aixtest 3 4 001750154C00 2/5/03
17:58:15 %usr %sys %wio %idle
17:58:17 43 9 1 46
17:58:19 35 17 3 45
17:58:21 36 22 20 23
17:58:23 21 17 0 63
17:58:25 85 12 3 0
Average 44 15 5 35
As a general rule of thumb, a server with over 40 percent waiting on I/O is spending too much
time waiting for I/O. However, you also have to take the type of workload into account. If you
are running a video file server, serving I/O will be the primary activity of the machine and you
will expect high %wio values.
A system with very busy CPUs can mask I/O wait. The definition of %wio is: Idle with some
processes waiting for I/O (only block I/O, raw I/O, or VM pageins/swapins indicated). If the
system is CPU busy and also is waiting for I/O, the accounting will increment the CPU busy
values, but not the %wio column.
The other column headings in the example indicate:
%usr: Time system spent executing application code
%sys: Time system spent executing operating system calls
%idle: Time the system was idle with no outstanding I/O requests
The implementation of the
sar command is different for the various UNIX variants. However,
the output of
sar -u is the same for all.
There are other modes to use
sar, which we will not discuss further:
Ongoing system activity accounting via cron
Display previously captured data
sar offers many different options and parameters. Refer to your system documentation and
the
sar man page for more information.
VMSTAT
The vmstat utility is a useful tool for taking a quick snapshot or overview of the system
performance. It is easy to see what is happening with regard to the CPUs, paging, swapping,
interrupts, I/O wait, and much more. There are several reports that
vmstat can provide. They
vary slightly between the different versions of UNIX. Refer to your system documentation and
the
vmstat man page for more information.
IBM AIX
This section covers items specific to the IBM AIX operating system. It is not intended to
repeat the information that is contained in other publications. We focus on topics that are not
covered in the well known literature or are important enough to be repeated here.