High rpool space utilization and fragmentation

[arturasb]

Hi.

Please help me understand if my situation with root on ZFS is normal
I have double 224 Intel SDDs in the ZFS mirror configuration to store my root of PVE install. I keeps no backups, no ISO files, no VM images on this ZFS pool just PVE install itself. My only VM (Windows Server 2019) is on separate storage (ZFS mirror on 2x2T HDDs). Backups are stored on CIFS storage on the diferent PVE node. In the PVE web UI I see that ~163 GB of 224G are allocated, but the largest files are only 6 Linux Kernel images, Linux firmware and apt cache (checked with 'df' and 'ncdu'). It is hard to believe that it all takes ~163 GB of storage.

Where should I look at to find the reason of such high storage utilization ? Should some ZFS tools be used to free up some space ?
Is ~40+ % of pool fragmentation normal/OK ?

Thanks for your time in advance.

Regards
ArtūrasB.

 

[arturasb]

Update.
More checks
zfs list -r rpool
Gives this:

NAME               USED  AVAIL     REFER  MOUNTPOINT
rpool              163G  51.7G      104K  /rpool
rpool/ROOT         163G  51.7G       96K  /rpool/ROOT
rpool/ROOT/pve-1   163G  51.7G      163G  /
rpool/data          96K  51.7G       96K  /rpool/data

Checked '/' with ncdu, the result:

   1.3 GiB [          ] /usr                                                                    
  743.7 MiB [          ] /var
  235.2 MiB [          ] /boot
   97.2 MiB [          ] /run
   42.7 MiB [          ] /dev
    3.6 MiB [          ] /etc
   69.5 KiB [          ] /root
   28.5 KiB [          ] /tmp
    3.0 KiB [          ] /datapool
    2.0 KiB [          ] /rpool
@ 512.0   B [          ]  libx32
@ 512.0   B [          ]  lib64
@ 512.0   B [          ]  lib32
@ 512.0   B [          ]  sbin
@ 512.0   B [          ]  lib
@ 512.0   B [          ]  bin
e 512.0   B [          ] /srv
e 512.0   B [          ] /opt
e 512.0   B [          ] /media
e 512.0   B [          ] /home
.   0.0   B [          ] /proc
    0.0   B [          ] /sys

This doesn't like it could take more than 160GB of disk space... My thinking goes like this - if zfs reports this high storage utilization on /rpool/ROOT/pve-1, then scanning the root of PVE installation should reveal which directories & files consumes those ~160GB of space.
Anyone has any thoughts what to check next ?