nvme-cli 2.11

NVMe is a different protocol from SATA with its own command set and health reporting. nvme-cli speaks it directly.

The health log

sudo nvme list
sudo nvme smart-log /dev/nvme0

critical_warning     : 0
temperature          : 41 C
available_spare      : 100%
percentage_used      : 3%
media_errors         : 0

percentage_used is a vendor wear estimate where 100 means rated endurance reached. Drives often continue past it; treat it as a replacement planning signal.

available_spare dropping below its threshold is a genuine warning and one of the better failure predictors.

media_errors should be zero.

critical_warning non-zero means the drive is reporting a specific problem.

Thermal throttling

sudo nvme smart-log /dev/nvme0 | grep -i temp

NVMe drives reduce performance when hot, which is the usual explanation for a drive that starts a large transfer fast and slows partway through. Run a sustained fio job and watch the temperature.

Secure erase

sudo nvme format /dev/nvme0n1 --ses=1     # cryptographic erase

Discards the internal encryption key, making everything unrecoverable in seconds regardless of capacity. This is what to do before selling or returning a drive, and it is dramatically faster than overwriting.

Destructive with no meaningful confirmation. Check the device name twice.

Alongside smartctl

smartmontools reads the same health log in a familiar format and can alert through smartd, which is what you want for routine monitoring.

nvme-cli covers what smartctl does not: namespaces, format and secure erase, firmware slots, and vendor log pages.

A monthly check

for d in /dev/nvme?; do
  sudo nvme smart-log "$d" | grep -E 'critical_warning|percentage_used|available_spare|media_errors'
done

Run it from a systemd timer and you will know a drive is wearing out long before it fails. Our nvme-cli guide covers the rest.

Verify Your Download

Source releases are on GitHub.