Smart_ata: notify on drive age (powered_on)

General
1

Hi,

Installed checkmk this weekend. Still finding my way around but the great documentation helps, thanks!

What I cannot figure out:

How can I create a rule that monitors the SMART parameter powered_on? Would like to receive a notification for drives running for more than x years.

The powered_on hours are displayed for the HDDs in the server in question.

Also managed to find the rule set “SMART ATA (incompatible with legacy plug-in)”. But when adding a rule, I only see the following parameters:

image
image609×230 16.3 KB

No powered_on… What am I missing?

Thanks in advance!

2

Solved this with a bash script.

#!/bin/bash


GREP_LINES="$(grep "Powered on:" /data/checkmk/cmk/var/nagios/nagios.log)"
#echo $GREP_LINES

counter=0
while IFS= read -r line; do
        #echo "Found: $line"

        model="$(echo "$line" | sed -E 's/.*SMART ([-A-Za-z0-9 _]+) Stats.*/\1/g')"

        if [[ ! " ${modelNames[*]} " =~ [[:space:]]${model}[[:space:]] ]]; then
                counter=$(($counter+1))
                modelNames[$counter]="$model"
        fi

done <<< "$GREP_LINES"

#echo "count: $counter"

WARNING_AGE=0
WARNING_CYCLES=0

while IFS= read -r line; do
        model="$(echo "$line" | sed -E 's/.*SMART ([-A-Za-z0-9 _]+) Stats.*/\1/g')"

        counter=0
        for modelDummy in "${modelNames[@]}"
        do
                counter=$(($counter+1))

                if [ "$model" == "${modelNames[$counter]}" ] ; then
#                       echo "$counter: $model"
#                       echo $line


                        HOURS="$(echo "$line" | sed -E 's/.*Powered on:[0-9 days]* ([0-9]*) hours.*/\1/g')"
                        if [ ${#HOURS} -gt 3 ] ; then
                                HOURS=0
                                YEARS="$(echo "$line" | sed -E 's/.*Powered on: ([0-9]*) year.*/\1/g')"
                        else
                                YEARS=0
                        fi

                        DAYS="$(echo "$line" | sed -E 's/.*Powered on:[0-9 years|1 year]* ([0-9]*) day.*/\1/g')"

                        CYCLES="$(echo "$line" | sed -E 's/.*Power cycles: ([0-9]*),.*/\1/g')"

                        ALL_YEARS[$counter]=$YEARS
                        ALL_DAYS[$counter]=$DAYS
                        ALL_CYCLES[$counter]=$CYCLES

#                       echo "years: $YEARS"
#                       echo "days: $DAYS"
#                       echo "hours: $HOURS"
#                       echo "cycles: $CYCLES"

                        if [ $YEARS -gt 4 ] ; then
                                if [ $DAYS -gt 300 ] ; then
                                        WARNING_AGE=$counter
                                fi
                        fi

                        if [ $CYCLES -gt 1000 ] ; then
                                WARNING_CYCLES=$counter
                        fi
                fi
        done
done <<< "$GREP_LINES"

if [ $WARNING_AGE -gt 0 ] ; then
        logger "CDA: HDD headsup"
        logger "CDA: Drive ${modelNames[$WARNING_AGE]} running for ${ALL_YEARS[$WARNING_AGE]} years and ${ALL_DAYS[$WARNING_AGE]} days. Power cycles: ${ALL_CYCLES[$WARNING_AGE]}."
fi

if [ $WARNING_CYCLES -gt 0 ] ; then
        logger "CDA: HDD/SSD headsup"
        logger "CDA: Drive ${modelNames[$WARNING_CYCLES]} running for ${ALL_YEARS[$WARNING_CYCLES]} year(s) and ${ALL_DAYS[$WARNING_CYCLES]} days. Power cycles: ${ALL_CYCLES[$WARNING_CYCLES]}."
fi

if { [ $WARNING_AGE -gt 0 ] || [ $WARNING_CYCLES -gt 0 ] ;} ; then
        logger "CDA: All HDD/SSD data just in case:"
        counter=0
        for modelDummy in "${modelNames[@]}"
        do
                counter=$(($counter+1))
                logger "CDA: ${modelNames[$counter]}: ${ALL_YEARS[$counter]} year(s) ${ALL_DAYS[$counter]} days, power cycles: ${ALL_CYCLES[$counter]}"
        done
fi

exit 0
3

The HOURS line should read as follows

HOURS="$(echo "$line" | sed -E 's/.*Powered on:[0-9 days]* ([0-9]*) hour.*/\1/g')"