I have not been  successful  in finding  any resolution for my Unraid  cpu_iowait issues.  There are others with this issue that have not found any resolution.    Please share any good successful comments with me on this issue.  

Anyone having these issues?

 

IO Wait isn't an "issue" the way that I *think* you're thinking that it is.

 

It's a metric that reflects that process(es) are waiting for I/O to complete from the drives / ssd's.  This could be because of multiple reasons -> the drives simply can't keep up with the request, too many simultaneous requests happening, failing drives, drives continually dropping off and reconnecting, exceeding the bandwidth by a fair margin of what the PCIe bus can handle etc.

 

There unfortunately is no one size fits all solution to what boils down to be a rather generic term.

 

As netdata words it's comments on the metric:

Keep an eye on iowait  (0.17%). If it is constantly high, your disks are a bottleneck and they slow your system down.

 

All that being said, from the screen shot you've included, the process Plex Transcoder is writing to disk3.  Assuming you have a parity drive, that's going to be a major bottleneck, as writes to a parity protected array are inherently slow, and the storage of the temp files for transcoding belongs on an SSD.

 

Or, put another way: Nothing is stopping you from using a DVD-RAM drive (I'm one of the few people who owned one) as a cache drive.  You're I/O wait is going to be sky-high.  But nothing is actually wrong because it's just a slow device.

Can you tell me how you determined what process was causing the iowait and how you determined the disk it was using? I have the same issue I'd like to figure out where the bottleneck is and I'm struggling to find what you see in the screenshots.