6.8.3 Disk X is high on usage then min later normal

[shaunsund]

So this is odd. I have 8 data disks and all except #1 are nearly full so I set my global warning at 93% and critical is 97% usage so that everything is OK utilization wise. Today was the second time I got the below messages:

Quote

Event: Unraid Disk 7 disk utilization
Subject: Warning [FRACTAL] - Disk 7 is high on usage (89%)
Description: TOSHIBA_HDWD130_X7V0SYUAS (sdh)
Importance: warning

and also one for disk 8. 

Yet, a minute later I get the following:

Quote

Event: Unraid Disk 2 message
Subject: Notice [FRACTAL] - Disk 2 returned to normal utilization level
Description: TOSHIBA_HDWE140_18E2KEOQF58D (sdl)
Importance: normal

also for disk 3, 4, 5, 7, and 8

I got the Warning at 5:19pm and then the Notice emails at 5:20pm.

Disks are practically idle during this time. I don't know why they would hiccup these utilization errors.

If I had a disk or cache drive drop out then maybe the totals might change but there is no indication in syslog as to a cause for these warnings.

 

fractal-diagnostics-20201104-1804.zip

[JorgeB]

You're not the first complaining of this, but it's an elusive issue and difficult to replicate.

[shaunsund]

7 hours ago, JorgeB said:

You're not the first complaining of this, but it's an elusive issue and difficult to replicate.

Wow! I should go buy a lottery ticket.

 

[trurl]

16 hours ago, shaunsund said:

If I had a disk or cache drive drop out then maybe the totals might change

Don't understand why you think that might have an effect.

 

16 hours ago, shaunsund said:

global warning

Are you sure you haven't set a different value for specific disks?

[shaunsund]

well if a drive disappeared then the calculation for disk free would change -- although I would expect that to cause a warning before disk usage.

I did check while writing the first post, every disk's utilization warning and critical are equal to the global. Also, even if they were different, why did they alarm for 2 disks and report ok for 6 disks a minute later?

Like JorgeB mentioned it seems to be "an elusive issue and difficult to replicate"

[JorgeB]

Same issue as this:

 

[trurl]

1 minute ago, shaunsund said:

well if a drive disappeared then the calculation for disk free would change -- although I would expect that to cause a warning before disk usage.

If you are referring to total capacity I don't think there is any warning based on that, why would there be? Each disk is independent.

 

If you had an array data disk drop out while being written, it would become disabled and emulated at that point and nothing about its free space would change.

[shaunsund]

I see your points. Still don't know why it did it.

[Keichi]

Same issue for months now....