I've been dealing with IO Wait issues for months now. I've searched through the forum and tried all kinds of solutions. My most recent effort was changing the cache drive from btrfs to XFS. Now it seems like the issue is worse than ever. It took me 2 hours to even be able to login to the server. Everything is locked up and incredibly slow. I've added my logs and some relevant pictures below. I can see that the disk sda is busy, but I think that is my usb stick, so I'm not really sure why that would be the case.  I've also updated my bios to the most recent version. It's weird because my dockers appear to be working fine, it's just the unraid UI that is slow and becomes inaccessible. I've tried pinning cpu cores to every docker container to keep some available to the server but that didn't really make a difference. 

 

Looking for any help to get my server to be usable again.

 

 

unraid-diagnostics-20230918-1611.zip

Try rebooting and start the containers one by one to see if it gets worse after a specific one.

On 9/19/2023 at 3:22 AM, JorgeB said:

Try rebooting and start the containers one by one to see if it gets worse after a specific one.

It appears to be Photoprism that is having the biggest impact. Is there any reason why it would be reading from the flash drive? 

Not that I'm aware, you can try asking in the container support thread/discord, assuming one exists.

 

It's happening again now during a sabnzbd download and it seems like the usb stick is causing the issue. I don't really understand what is going on there, because it's not even mapped to anything in the docker template. 

Where do you see that the flash drive is the problem? There should not be any major reads or writes to the flash drive

On netdata it looks like the sda drive is having the most issues. And on the unraid dashboard that is the flash drive.