task txg_sync hung task

[pakyrs]

Hi there,

I am starting to see these errors on one of my nodes, I searched the forum but can't find a definite answer.

I have plenty of ram and the drives are CMR, I don't get what can cause this, a weird behaviour is that the system will keep on running but the network will go, making it unreachable unless jumping on IPMI, restarting the networking service won't make a difference, but a full reboot will fix.

Running VM and containers are just using 45G RAM where my pool is 134G

root@vhost1:~# for i in 201 202 301; do qm config $i | grep memory; done
memory: 8192
memory: 4096
memory: 16384
root@vhost1:~# for i in 101 106 110; do pct config $i | grep memory; done
memory: 4096
memory: 8192
memory: 4096

root@vhost1:~# free -h --giga --committed
               total        used        free      shared  buff/cache   available
Mem:            134G         43G         89G         72M        3.4G         91G
Swap:           7.9G          0B        7.9G     
Comm:            75G         42G         32G

root@vhost1:~# for i in $(ls /dev/sd[e-h]); do smartctl -i $i | grep -i 'rotation'; done
Rotation Rate:    7200 rpm
Rotation Rate:    7200 rpm
Rotation Rate:    7200 rpm
Rotation Rate:    7200 rpm

root@vhost1:~# for i in $(ls /dev/sd[e-h]); do hdparm -I $i | grep -i 'Model'; done
        Model Number:       HGST HUS726T4TALE6L4
        Model Number:       HGST HUS726T4TALE6L4
        Model Number:       HGST HUS726T4TALE6L4
        Model Number:       HGST HUS726T4TALE6L4


root@vhost1:~# cat /etc/modprobe.d/zfs.conf
# Min 8Gb / Max 16Gb Limit
options zfs zfs_arc_min=8589934592
options zfs zfs_arc_max=17179869184

options zfs zfs_recover=1