I'm trying to figure out why 1 core alway stuck at 100% . I ran Pihole , Immich, and Krusader docker only. 

Shares is using cache drive. 

Set spindown at 1hr 

I posted on Reddit and got told there is something constantly write/read on my array. 

Htop and top results (below 1/4 of the bar) are different with Processor % on Dashboard UI (between 25-50%). 

 

Can anyone help me to investigate what going on?

 

 

diag.zip

Me too.

In Grafana I'm investigation exactly the same.

But tools like iotop, htop, top or btop seem not to realize that.

 

If the CPU is almost idle, it looks like this.

 

In the moment the CPU gets some preasure this behavior vanishes.

In case of CPU is again on low preasure the behavior comes back.

 

Interestingly the behavior correlates with IO Waits

 

It does not matter if Docker Service (with all the container) and VM Manager is running or not.

It is always the exact same behavior.

 

This is a six hour perspective:

 

My setup:

Unraid system: Unraid server Plus, version 6.12.5

Motherboard: Intel Corporation NUC12WSBi7, Version: M46422-303

Processor: 12th Gen Intel® Core™ i7-1260P @ 3762 MHz

 

All drives are sleeping except my KINGSTON_SKC3000D2048G_50026B768671EA03 (nvme0n1) with ZFS Pool

 

Eigther it is an kernel issue or it is some kind of ghosty "organizer" process.

 

Tried intel_pstate=passive in syslinux.cfg -> did not have any effect on this behavior

Edited December 1, 20232 yr by Unpack5920

after some hours investigation I found the reason for my issue:

I'm using an USB3.2 enclosure IB-1817M-C31 of Icy-Box.

After pulling the plug the waits and the CPU utilization were gone.

I then decided to do a firmware update of the enclosure.

This solved the issue for now.