Skip to content

Read SMART disk errors

SMART is self-monitoring built into each hard drive. TrueNAS shows SMART status so you can replace a dying disk before the pool goes offline.

This guide explains what to look at and when to worry — not how to fix hardware (that is When the pool is degraded).

Where to look in TrueNAS

  1. Open dsm.saxobroko.com.
  2. Go to Storage → Disks (or Devices → Disks).
  3. Each drive shows:
  4. Serial number and slot
  5. Temperature
  6. SMART status — often PASSED or OK
  7. Click a disk row for SMART Test Results and SMART Attributes.

Also check Storage → pool → Status for drives marked ONLINE, DEGRADED, or FAULTED.

Scheduled SMART tests

TrueNAS can run short and long SMART tests automatically:

  1. Data Protection → S.M.A.R.T. Tests (or Tasks → SMART Tests).
  2. Typical schedule:
  3. Short — monthly (minutes per disk)
  4. Long — quarterly (hours per disk)

If no schedule exists, consider adding one — see Backups and pool health.

When to worry

Act soon (plan a replacement)

Signal Meaning
Reallocated sectors (raw value climbing) Drive remapped bad blocks — early failure sign
Current pending sectors > 0 Unstable sectors waiting to remap or fail
Offline uncorrectable > 0 Read failures the drive could not fix
SMART overall status FAILED Drive is reporting failure
Temperature consistently above ~50°C Cooling or airflow problem — check NAS fans and room temp
Pool DEGRADED with one disk FAULTED Replace that disk — When the pool is degraded

Usually fine (monitor only)

Signal Meaning
Power-on hours high Normal on 24/7 NAS drives
Temperature 30–45°C Typical operating range
SMART PASSED, no pending sectors Healthy
One short test aborted once Retry; check logs if it repeats

ZFS scrub errors vs SMART

Source What it means
Scrub checksum errors Data did not match — often points to a specific disk even if SMART still says PASSED
SMART errors without scrub errors Drive is sick; replace before it takes the pool down

If scrub finds errors, treat the disk as suspect even when SMART looks green.

Run a manual SMART test (one disk)

  1. Storage → Disks → select the disk.
  2. Click Run Manual Test → choose SHORT first (2–5 minutes).
  3. Wait for result — PASSED or FAILED.
  4. If SHORT passes but you still suspect the disk, run LONG (can take hours — run overnight).

Do not run LONG tests on multiple disks at once unless you accept heavy I/O load.

What not to do

Do not ignore DEGRADED or FAULTED

RAIDZ1 survives one failed drive. A second failure before rebuild finishes means data loss. Replace the failed drive promptly.

Do not pull a disk to 'test' it

Removing the wrong drive or yanking a disk while the pool is online can destroy the array. Follow When the pool is degraded.

After you identify a bad disk

  1. Buy a replacement — same size or larger (16 TB class in this build).
  2. Follow When the pool is degraded for hot-swap and resilver.
  3. After resilver completes, run Run a pool scrub.