Wrong data reported in vfs.dev.read.await[sdb]

[alex_1983]

Hello,

I'am using Zabbix 5.0.6 for monitoring Arch Linux host. The problem is I am getting wrong data for the item vfs.dev.read.await[sdb].

The disk /dev/sdb is dedicated completely to a VM, the problem is, when the VM is running I get correct values, but when the VM is shut down I get extremely high values, like 100-150 ms triggering an alarm.



I checked with iostat, and. in fact IO equals 0. Meanwhile, I have more disks on that host and the r/w await time is reported correctly.



What should I be looking at?

I attach screenshots.

[temask]

Hello,

alex_1983 whether you managed to solve this problem?
I have a similar situation.
There is a zfs draid pool.
All disks in the pool swear at await while the storage is idle. As soon as you start reading something, the correct data starts to be read.



time ranges are highlighted in yellow when reading from storage was started