Problem: Persistent 15-Minute “Pokes” Preventing HDD Spindown (TrueNAS SCALE 25.10.2)
System Environment:
-
OS: TrueNAS-SCALE-25.10.2 (Goldeye)
-
Hardware: 9-Drive HDD Array (MainPool), separate SSD Boot-Pool, separate NVMe App-Pool.
-
Services: System Dataset is confirmed on the boot-pool (SSD). All Apps (Jellyfin, etc.) are on NVMe. No SMB/NFS activity during the test window.
The Issue: Despite a 120-minute spindown timer being set, the drives never enter standby. Power consumption stays at 100W (active) instead of dropping to the expected 67W (idle/standby).
The Evidence (Logs): Looking at /var/log/syslog, I am seeing a “heartbeat” from truenas_audit_handler.py almost exactly every 15–16 minutes (e.g., 03:14, 03:30, 03:45). This periodic audit/polling is resetting the internal ZFS/HDD idle timers before they can ever hit the 120-minute threshold.
Roadblocks in 25.10.2:
-
Missing UI Options: The “Enable S.M.A.R.T.” toggle and “S.M.A.R.T. Extra Options” boxes are no longer present in the Disk Edit UI for this version.
-
Middleware API Changes: Attempting to mute the service via CLI (
midclt call smart.update '{"powermode": "STANDBY"}') returns “Method does not exist,” indicating the legacy S.M.A.R.T. API has been removed or replaced. -
Kernel Polling: It appears the new
drivetempkernel module or Netdata reporting is polling the drives for temperature/stats every 15 minutes, acting as a “forced wake-up” or timer reset.
The Question for the Forum: In 25.10.2, how do we globally force the middleware/audit-handler to respect standby states or increase the polling interval of drivetemp so that a 120-minute spindown timer can actually complete?