Shell Script To Notify When Server Uptime Has Reached Threshold

Question

I'm looking to cleanup the following shell script that reads in the uptime of a Linux server and will mail a certain user when the threshold has been met so that we know when to reboot through a daily cron job. The below will read in the uptime and output the result to a file, which will then be emailed. I'd also like to add in the hostname of the server either the report or the email heading. Fairly new to shell scripting so any/all tips are appreciated. Thanks!

#!/bin/bash
timeup () { uptime | awk  '{print $3}'; }
UPTIME=100
if [ $(timeup) -ge $UPTIME ]; then

    #output uptime to report
    uptime | awk '{print $3,$4}' | sed 's/,//' >> /opt/scripts/report.out 

    #mail report 
    cat /opt/scripts/report.out | mail -s "Server Needs To Be Rebooted" "your@emailaddress.com"

    #remove old file
    rm -f /opt/scripts/report.out   
fi

Answer 1

timeup () { uptime | awk '{print $3}'; } timeup () { uptime | awk '{print $3}'; } result will be something like that: " 45:45, ". There are 2 problems with this:

• return value will be string instead of integer therefore if line will send an error message
• "," sign at the and of the string

This is better (but not so nice:) ): timeup () { uptime | awk '{print $3}' | awk -F\: '{print $1}';} timeup () { uptime | awk '{print $3}' | awk -F\: '{print $1}';}

Second awk command divide hour and minute, and print only hours.

I prefer sendEmail script instead of mail command: http://manpages.ubuntu.com/manpages/xenial/man1/sendEmail.1.html

I'd put hostname in the subject: "Server Needs To Be Rebooted: $(hostname -f)"

hostname -f command print out FQDN

You can merge uptime... and mail command lines and remove rm command:

uptime | awk '{print $3,$4}' | sed 's/,//' | mail -s "Server Needs To Be Rebooted" "your@emailaddress.com"