Getting hard drives to properly spin down in FreeNAS (especially with ZFS) requires understanding multiple system interactions. The immediate spin-up behavior you're observing typically occurs because:

• Background ZFS processes (scrubs, resilvers, snapshots)
• System monitoring tools (smartd, sysctl)
• Misconfigured APM/SATA settings

First, verify your current disk states:

smartctl -i /dev/ada0 | grep -i 'APM Level'
sysctl dev.ada.0 | grep 'apm'
zpool status

Then implement these critical adjustments:

# Disable unnecessary disk access
sysctl vfs.zfs.scrub_delay=3600
sysctl vfs.zfs.resilver_delay=3600
sysctl vfs.zfs.vdev.async_write_active_min=1
sysctl vfs.zfs.vdev.async_write_active_max=1

Edit /usr/local/etc/smartd.conf to reduce polling frequency:

DEVICESCAN -a -o on -S on -n standby -s (S/../.././02|L/../../7/03)

Create a post-init script (/usr/local/etc/rc.d/spindown):

#!/bin/sh
. /etc/rc.subr

name="spindown"
start_cmd="${name}_start"

spindown_start() {
    for disk in $(sysctl -n kern.disks); do
        camcontrol cmd ${disk} -a "EF 03 00 00 00 00 00 00 00 00 00 00" -r -
        camcontrol apm ${disk} -l 127
        camcontrol standby ${disk} -t 1200
    done
}

load_rc_config $name
run_rc_command "$1"

Monitor disk states with:

while true; do
    printf "%s - %s\n" "$(date)" "$(smartctl -n standby /dev/ada0 | grep 'mode')"
    sleep 60
done

Expect output like:

Mon Jul 10 14:00:01 UTC 2023 - Device is in STANDBY mode
Mon Jul 10 14:30:45 UTC 2023 - Device is in ACTIVE mode (brief access)
Mon Jul 10 14:31:01 UTC 2023 - Device is in STANDBY mode

---

When I set up my FreeNAS 11.2 system with a USB boot device and 4-disk ZFS pool, I expected the HDD Standby (20 minutes) and APM Level 127 settings to keep disks spun down during inactivity. However, the disks kept waking up almost immediately after spindown - a common frustration for energy-conscious sysadmins.

To identify what's preventing persistent spindown, SSH into your FreeNAS box and run:

grep -E 'sd[a-d]' /var/log/messages | grep -i -E 'start|stop'

This reveals disk activity patterns. Common culprits include:

• ZFS background scrubs (even when idle)
• SMART monitoring services
• System logging to disks
• Periodic snapshot operations

Add these to your /etc/local/sysctl.conf:

# Reduce ZFS background activity
vfs.zfs.scrub_delay=86400
vfs.zfs.resilver_delay=86400
vfs.zfs.vdev.async_write_max_active=1
vfs.zfs.txg.timeout=60

Edit /etc/periodic.conf to minimize disk activity:

daily_status_zfs_enable="NO"
daily_scrub_zfs_enable="NO"
weekly_scrub_zfs_enable="NO"

Then disable SMART monitoring for non-critical disks via the FreeNAS GUI at Services → S.M.A.R.T.

Create /usr/local/etc/rc.d/spindown with:

#!/bin/sh
. /etc/rc.subr

name="spindown"
start_cmd="${name}_start"

spindown_start() {
    for disk in $(sysctl -n kern.disks); do
        if [ "$(camcontrol identify $disk | grep -c 'ZFS')" -eq 0 ]; then
            camcontrol apm $disk -l 127
            camcontrol standby $disk -t 1200
        fi
    done
}

load_rc_config $name
run_rc_command "$1"

Make executable and enable:

chmod +x /usr/local/etc/rc.d/spindown
echo 'spindown_enable="YES"' >> /etc/rc.conf.local

Check actual spindown status with:

for disk in $(sysctl -n kern.disks); do
    echo -n "$disk: "
    camcontrol cmd $disk -a -E "5C 00 00 00 00 00 00 00 00 00 00 00" -r -
done

For more control, create a cron job (crontab -e):

*/10 * * * * /usr/local/bin/spindown_check

With /usr/local/bin/spindown_check containing:

#!/bin/sh
for disk in $(sysctl -n kern.disks); do
    if ! lsof +D /mnt | grep -q "$disk"; then
        camcontrol standby $disk
    fi
done