DHIVA LABS · LEARN

Appearance

Check drive health ​

The Disks tab shows every physical drive with its live read and write speed and its health, and how full each filesystem is. Press 4 to open it.

Get full health data ​

Drive health data comes from SMART, the record every modern drive keeps of its own wear and errors. Reading it needs root:

bash
sudo ./sysvitals.py

Then press 4. Without sudo, you still get the model, temperature and throughput, but NVMe drives show SMART: permission denied (run with sudo for full SMART).

  • NVMe drives are read directly. Nothing else needs to be installed.
  • SATA drives (and other non-NVMe drives) are read through smartctl. Install smartmontools first; see Install. If smartctl is missing, or returns no health data (which is what usually happens without sudo), the drive shows SMART: unavailable (install smartmontools for SATA health).

SysVitals only reads from drives. It never changes drive settings or runs drive self-tests.

Read a drive's entry ​

Each drive starts with a cyan line:

text
▸ /dev/nvme0n1  <model>  fw <firmware>  SSD  <size>   R <read>/s  W <write>/s

That is the device name, model, firmware version (when known), SSD or HDD, the capacity, and the current read and write speed. The next line shows the serial number. Loop devices, RAM disks, zram, device-mapper and optical drives are not listed.

NVMe drives ​

With root, an NVMe drive shows these lines:

LineWhat it tells you
SMART: HEALTHYThe drive reports no critical warnings. SMART: CRITICAL in red is followed by the warnings, listed below.
Health:Life left, as a bar followed by % life remaining. Green above 30 %, yellow above 10 %, red at 10 % or less.
Wear used, spare, alarm atHow much of the rated wear is used, the spare capacity left, and the spare level at which the drive raises an alarm. Then a rough lifetime estimate.
Lifecycle:Power-on hours, power cycles, unsafe shutdowns, media errors and error-log entries.
Totals:Data written and read over the drive's life, minutes the controller was busy, and millions of write and read commands.
Temperature:The drive temperature, each sensor's reading, and minutes spent above the warning temperature, if any. Green below 55 °C, yellow below 70 °C, red at 70 °C and above.

The lifetime estimate divides the wear so far by the power-on hours, for example projected ~6.2y left at current write rate. New drives with less than 1 % wear show wear <1%, too early to project. Treat the figure as a rough guide: it assumes you keep writing at the same rate.

The critical warnings an NVMe drive can raise are:

  • spare below threshold
  • temperature out of range
  • NVM reliability degraded
  • media in read-only mode
  • volatile backup failed
  • persistent region read-only

DANGER

If a drive shows SMART: CRITICAL, media in read-only mode, a life bar in red, or a growing number of media errors, back up its data now. A failing drive can stop working without further warning.

SATA drives ​

With smartmontools installed and root, a SATA drive shows:

  • SMART: with the drive's own verdict, usually PASSED. Anything else is shown in red.
  • A Health life bar, if the drive reports a life-left or wear attribute.
  • Power-on hours, power cycles, reallocated sectors and temperature, when the drive reports them.

Reallocated sectors are areas of the disk the drive has replaced because they went bad. A count that keeps rising is a warning sign.

Filesystems ​

Below the drives, ▸ Filesystems lists each mounted filesystem on a /dev/ device: the device, where it is mounted, the filesystem type, a usage bar, the percentage used and the used and total size. The bar is green below 70 %, yellow below 90 % and red above.

In a snapshot ​

./sysvitals.py --dump prints the same information in short form under --- Disks ---: model, type, size and speed, then temp:, SMART: (OK or WARNING: with the warnings), life:, lifecycle: and totals: lines for NVMe drives. Run it with sudo for the full data.

Next steps ​