smartctl disk monitoring nagios check

00-readme.md

check_smartctl_health

Run with no arguments (check all disks)

$ sudo ./check_smartctl_health
ok: all devices healthy (20 total)
$ echo $?
0

Run with a disk as an argument

$ sudo ./check_smartctl_health /dev/rdsk/c1t5000CCA03BB46E69d0s0
ok: all devices healthy (1 total)
$ echo $?
0

Run with a bad argument

$ sudo ./check_smartctl_health foo
unknown: failed to run - smartctl -a foo
$ echo $?
3

check_smartctl_health

#!/usr/bin/env bash
#
# Check for disk health using smartctl
#
# Call with a list of devices as arguments to run this check against, or omit
# the first argument and this script will run against all disks found.
#
# Author: Dave Eddy <dave@daveeddy.com>
# Date: September 13, 2023
# License: MIT

devices=("$@")

if (( ${#devices[@]} == 0)); then
	# no devices given, scan for them
	output=$(smartctl --scan)

	if (($? != 0)); then
		echo 'unknown: failed to enumerate devices'
		exit 3
	fi

	if [[ -z $output ]]; then
		echo 'unknown: no devices found to check'
		exit 3
	fi

	# store devices found in the "devices" array
	while read -r device _; do
		devices+=("$device")
	done <<< "$output"
fi

# get health for each device found - look for bad devices
bad_devices=()
for device in "${devices[@]}"; do
	# read device health
	output=$(smartctl -a "$device")

	if (($? != 0)) || [[ -z $output ]]; then
		echo "unknown: failed to run - smartctl -a $device"
		exit 3
	fi

	# keep track of devices that don't report "OK"
	if ! grep -q '^SMART Health Status:.*OK$' <<< "$output"; then
		bad_devices+=("$device")
	fi
done

num_devices=${#devices[@]}
num_bad=${#bad_devices[@]}

if (( num_bad == 0 )); then
	echo "ok: all devices healthy ($num_devices total)"
	exit 0
else
	echo "critical: $num_Bad devices unhealthy ($num_devices total)"
	exit 2
fi