Massive disk IO at 12 noon each day

[jj_uk]

How can I find out what's causing the massive disk IO that occurs each day at 12-noon ?  It's causing some containers to fall over.

3:30am is my CA backup, 8am my Duplicati backup begins, transferring to a different server, and it finishes at around 10:30am so I expect to see high disk and network usage peek a this time, but I have no idea what the 12-noon disk usage is?

ssdtrim runs every hour, other than that, I have no idea.

 

 

 

 

Edited December 6, 2021 by jj_uk

[Squid]

Plex? Mover? One of the arrs scanning?  Multitude of apps have scheduled tasks.

 

iotop from NerdPack (also requires Python 2.7) may help figure it out.

[jj_uk]

12:00:

 

 

 

12:12: (seems back to normal) :

 

 

Edited December 7, 2021 by jj_uk

[Squid]

Looks like it's trimming (Dynamix SSD Trim Settings under Scheduler).  Makes sense for the containers to all pause etc as no other I/O can happen to the device while a trim is in progress

[jj_uk]

Yes, I have the trim setting set to hourly.  Is manually trimming on a schedule really required? Can the OS just trim when it deletes something instead?

Edited December 7, 2021 by jj_uk

[Squid]

BTRFS will do that.  XFS won't.  Hourly is a bit much to run it though.  Daily or weekly during off hours is more than sufficient

[jj_uk]

My cache drives are using btrfs, so i'm ok to just disable it?