What were you trying to do that didn't work?

●Environment
The virtual environment uses VMware
Storage consisting of multiple disks
The OS disk is created on the storage with thick provisioning
The io status is periodically obtained with iostat -x
Sysstat version: sysstat-11.7.3-9.el8.x86_64
Kernel version: kernel-4.18.0-477.10.1.el8_8(No third-party Multipath software)
●Phenomenon
When compressing 5-7 multiple files close to 1G into a tar.gz file, the %util value obtained with iostat exceeded 100%.

For example:
~~~
> Server Date Time Device rrpm wrpm r w rsec wsec avgrq avgqu await r_await w_await svctm %util
xxxxxx 2024.09.09 6:00 sda 0 2.03 0 25.06 0 0.4 0 16.44 0 0.61 0 0 12.32
xxxxxx 2024.09.09 6:01 sda 0 13.37 0 1408.92 0 0.27 0 1.96 0 0.92 0.01 0 105.41
xxxxxx 2024.09.09 6:02 sda 0 108.35 0 104512.64 0 0.68 0 0.63 0 20.6 2.23 0 964.58
xxxxxx 2024.09.09 6:03 sda 0 7.08 0 4502.65 0 0.47 0 6.18 0 1.76 0.01 0 635.67
xxxxxx 2024.09.09 6:04 sda 0 6.63 0 4646.56 0 0.32 0 4.56 0 1.87 0.01 0 700.49
xxxxxx 2024.09.09 6:05 sda 0 2.1 0 25.75 0 0.33 0 13.7 0 0.47 0 0 12.26
~~~

What is the impact of this issue to you?

This make the customers confused. And they cannot evaluate disk performance.

Please provide the package NVR for which the bug is seen:

sysstat-11.7.3-9.el8.x86_64

How reproducible is this bug?:

Easy.

Steps to reproduce

When compressing 5-7 multiple files close to 1G into a tar.gz file, the %util value obtained with iostat exceeded 100%.

Expected results

The %util should not exceed 100%.

Actual results

The %util value obtained by iostat may exceed 100%.