Private browser-based analyzer

SMART & NVMe Health Report Explainer

Paste smartctl or nvme-cli output to extract the health fields that deserve attention. The report never leaves your browser.

Reports can contain a serial number. Redact it before posting the report publicly.

What the analyzer checks

It recognizes common smartctl and nvme-cli labels for overall health, critical warnings, temperature, percentage used, spare capacity, power-on hours, unsafe shutdowns, media errors, error-log entries, reallocated sectors, pending sectors and uncorrectable sectors.

This is a triage tool, not a pass/fail warranty test. Vendor thresholds, controller behavior and workload context still matter. An unrecognized field stays unknown instead of being guessed.

How to capture a complete report

On Linux, use sudo smartctl -x /dev/nvme0 for an NVMe controller or sudo smartctl -x /dev/sdX for a SATA drive. NVMe users can also run sudo nvme smart-log /dev/nvme0. Confirm the device path first with lsblk or nvme list.

What deserves immediate attention

  • A failed overall-health result or non-zero NVMe critical warning.
  • Media/data-integrity errors, pending sectors or uncorrectable sectors above zero.
  • Unexpectedly rapid wear growth, declining spare capacity or persistent high temperature.

One snapshot cannot show a trend. Save dated reports and compare them after heavy writes or when a drive begins producing errors. See our EQi12 SSD and USB evidence for an example of separating device identity, benchmark results and health data.

A real report, annotated

The capture below is a read-only smartctl -x from the 512 GB WD PC SN540 (NVMe 1.4) inside our Beelink EQi12 test unit, taken on 2026-07-11 with smartmontools 7.5. The raw file is published in the eqi12-measurement-data repository under storage/, with the serial number removed.

FieldReportedHow to read it
SMART overall-healthPASSEDA vendor summary, not a guarantee. A drive can pass and still carry a non-zero critical warning bit.
Critical Warning0x00No bits set. This is a live state, not a latched fault.
Temperature38 °CSensors 1 and 2 both reported 38 °C, so this is unlikely to be one bad sensor.
Available Spare100%, threshold 10%Full replacement-block reserve. Trend the gap to the threshold, not the absolute percentage.
Percentage Used0%The counter moves in whole-percent steps, so 733 GB of writes was not enough to move it. Use Data Units Written for small changes.
Data Units Read / Written1,442,327 [738 GB] / 1,433,247 [733 GB]733 GB written across 89 power-on hours is the number to trend.
Controller Busy Time46Minutes the controller was busy. A sudden step up between two reports usually means the workload changed, not the drive.
Power On Hours / Power Cycles89 / 10The cycle count gives context for the unsafe-shutdown counter.
Unsafe Shutdowns2Two across 10 cycles. Correlate with power-loss events before blaming the drive.
Media and Data Integrity Errors0The field that actually drives a replacement decision.
Error Information Log Entries0The log also reported "No Errors Logged" and no self-test history.
Warning / Critical Composite Temperature Time0 / 0Cumulative minutes at or above each threshold. Both zero means this drive has never recorded a thermal excursion.

Turning one report into a trend

Compare three things between two dated reports and ignore the rest: Data Units Written per power-on hour, Available Spare, and any field that should stay at zero and does not. Everything else is context. Our Critical Warning 0x02 guide explains why the two composite temperature counters are the only record of an overheat you did not watch happen, and the SSD and USB port evidence shows how to separate a degraded interface path from a genuine health problem.

Frequently asked questions

What SMART values mean my drive is failing?

Immediate red flags: overall health "FAILED", non-zero NVMe critical warning, media/data-integrity errors above zero, pending sectors, or uncorrectable sectors. Any of these warrant backup and replacement planning.

Is a high temperature dangerous for an NVMe drive?

Most NVMe drives operate safely up to 70–75°C. Sustained temperatures above 70°C can accelerate wear and trigger thermal throttling. Check the drive vendor’s specification and improve airflow if temperatures are consistently high.

How quickly does NVMe wear accumulate?

Wear depends on write volume. A typical consumer NVMe rated for 600 TBW writing 20 GB per day would take about 82 years to reach its endurance limit. Server workloads with heavy writes can reach the limit much faster. Monitor "Percentage Used" monthly.

What does "Available Spare" mean?

Available Spare (or Reserved Block Count) is the percentage of replacement blocks remaining. When it drops below the vendor threshold (often 10%), the drive may enter read-only mode. A declining spare percentage indicates active wear and approaching end of life.

Should I run SMART self-tests?

Yes. Short self-tests run in minutes and check critical components. Extended self-tests can take hours but scan the full media surface. Schedule short tests weekly and extended tests monthly. Results appear in the SMART error log.

How often should I check drive health?

For a home server, check SMART values monthly and after any unusual event (power loss, crash, performance drop). Save dated reports to track trends. A single healthy snapshot is less useful than a series showing stable values over time.