Iostat – How does %util affect %idle and %iowait

iostat

Looking at iostat, if the system needs to issue a request for io, but has not yet been able to do so due to %util being pegged, would that wait be counted as %idle or %iowait?

Best Answer

The meaning of three concepts：

%util：The previous answer has clearly described the meaning of idle.

%idle：The system is idle, that is, there is no running process.

%iowait： For a given CPU, the I/O wait time is the time during which that CPU was idle (i.e. didn’t execute any tasks) and there was at least one outstanding disk I/O operation requested by a task scheduled on that CPU (at the time it generated that I/O request).

High %util means that IO is busy, and the rate of the IO system is much lower than the processing rate of the CPU, so:

1) If the system has some CPU-intensive tasks running, the CPU will run other tasks while the IO is waiting, the CPU usage is high, and the %idle is low. At the same time %iowait is low (iowait indicates the proportion of IO transmission when IDLE)

2) If the system has no tasks to run, the CPU is idle when IO waits, this time the CPU usage is low, %idle is high, %iowait is high.

Related Solutions

Mysql – What iowait values are ok

iowait, as a measure of system load, is only a problem inasmuch as it's consuming CPU time that could be reasonably used by other processes. If you've still got idle CPU time, then high iowait (in and of itself) is not a problem.

On the other hand, a high iowait can be a symptom of a problem; however, you really should be profiling your application to see if time spent in disk I/O is a significant percentage of the total time taken to complete a useful task. You cannot do this by looking at CPU stats, top, htop, iotop, or any other system-level tool that people will try to recommend. You need to look at your application, and the way it works, to determine if changing disk I/O subsystem would be of any benefit in your application.

Linux – very specific iostat output needed in realtime

I'm not sure how to get the iostat formatting you're looking for. The tool I use for this type of monitoring, though, is customizable to the point where you could display the relevant fields. It's also excellent for generating data for replay and graphing purposes.

I use Collectl monitoring (available in CentOS via yum), and the disk check flags should give you what you need:

collectl -sD --dskfilt sda

Sample output (you can change the units and columns printed):

collectl -sD --dskfilt sdb
waiting for 1 second sample...

# DISK STATISTICS (/sec)
#          <---------reads---------><---------writes---------><--------averages--------> Pct
#Name       KBytes Merged  IOs Size  KBytes Merged  IOs Size  RWSize  QLen  Wait SvcTim Util
sdb              0      0    0    0     336      0   52    6       6     0     0      0    1
sdb             16      0    1   16     389      1   52    7       7     0     0      0    3
sdb              0      0    0    0    1236      9  109   11      11     0     0      0    0
sdb             16      0    1   16     676      3  140    5       4     0     1      1   14
sdb              0      0    0    0      64      1   16    4       4     0     2      1    2

Best Answer

Related Solutions

Mysql – What iowait values are ok

Linux – very specific iostat output needed in realtime

Related Topic