Finding which process is reading from disk constantly on FreeBSD

Question

EDIT: This ended up being a scrub process. Run zpool status -v and read the scan status for your pool clearly.

I have a TrueNAS server running FreeBSD and this morning at 12 AM the disks started to be read from constantly. I thought this was some snapshot task, so I disabled my snapshots and restarted the server, but every time I restart the server it continues to read from my storage pool. I have stopped all running jails as well and the disks still show IO (the screenshots below are with jails running, but they look the same with jails stopped).

I can't figure out which process is causing this. I tried using top -m io, but this shows no R/W activity.

According to this resource, top -m io doesn't work on FreeBSD so I can't find out which process is reading from my disks:

top io mode doesn't properly display stats

I have tried checking if there are any running snapshot tasks using ps -aux | grep rsync, but nothing shows except my grep command.

Here's some resources I've checked:

• Identify which process is responsible for heavy disk i/o under FreeBSD (useless, recommends top)
• Determining which process is causing heavy disk I/O? (useless, recommends iotop which is not available on FreeBSD)

Is it possible to get the top command working? I just want to know what process is reading from my disks and kill it.

Answer 1

Running…

top -a -m io

…should do the trick! If there’s no process associated with a heavy I/O - it’s kernel and you can’t really do anything about it. I mean no way to query.

It could be ZFS performs some integrity checks / scrubbing, so it might be a good lead. Take a look here:

https://docs.oracle.com/cd/E36784_01/html/E36835/gammt.html