An It-slave in the digital saltmine

I have hinted about automysqlbackup in an earlier blog. Of course an important script like this should be monitored by Nagios or op5 Monitor. In my opinion the correct way of doing this is using a passive check that sends in the result of the backup into Nagios or op5 Monitor. The backup script is supposed to run every 24 hour. If no result from the script has been reported to Nagios or op5 Monitor within the last 26 hours it will be set to WARNING status.

The Nagios or op5 Monitor configuration:

# service 'MySQL backup'

 define service{

 use                            default-service

 host_name                      dull

 service_description            MySQL backup

 check_command                  check_dummy!1 "No Data from MySQL backup"

 max_check_attempts             1

 active_checks_enabled          0

 check_freshness                1

 freshness_threshold            93600

 flap_detection_options         n

 contact_groups                 it-slav_mail,call_it-slav,it-slav_msn

 stalking_options               n

 }

Explanation:

The freshness_threshold 93600 makes Nagios or op5 Monitor to run the check_command if no result has been sent in the last 26 hours i.e. set status to WARNING and the Status Information to “No Data from MySQL Backup”.

I’ve modified the automysqlbackup.sh.2.5 so if there are any errors the message will be sent to a temporary file:

cat "$LOGERR"
cat "$LOGERR" > /tmp/automysqlbackuperror.log #this is added by peter@it-slav.net

A wrapper that starts automysqlbackup is created:

#!/bin/sh
 #This script is a wrapper for automysqlbackup
 #It runs automysqlbackup and sends the status of the automysqlbackup
 #to Nagios or op5 Monitor with nsca
 #By Peter Andersson
 #peter@it-slav.net
 #GPLv2

BASEDIR=/root/scripts/automysqlbackup/
 NAGIOSHOST=op5

cd $BASEDIR

./automysqlbackup.sh.2.5
 if [ $? != "0" ]
 then
 echo -n "dull   MySQL backup    2       CRITICAL:" >passive_check_data_failed
 cat /tmp/automysqlbackuperror.log >> passive_check_data_failed
 /opt/plugins/custom/send_nsca -H $NAGIOSHOST -c send_nscav2.cfg < passive_check_data_failed
 else
 /opt/plugins/custom/send_nsca -H $NAGIOSHOST -c send_nscav2.cfg < passive_check_data_passed > /dev/null
 fi

If the backup is successful the content of passive_check_data_passed will be sent to Nagios or op5 Monitor. passive_check_data_passed looks like this:

dull    MySQL backup    0       OK: MySQL backup success

If the backup script fails it will send the content in passive_check_data_failed via nsca to Nagios or op5 Monitor. An example of passive_check_data_failed:

dull    MySQL backup    2       CRITICAL:ERROR 1045 (28000): Access denied for user 'kalle'@'localhost' (using password: NO)

Create a cron job that starts the wrapper:

0     23       *       *       *       /root/scripts/automysqlbackup/automysqlwrapper.sh

Links:

• Earlier article about passive checks
• Earlier article about automyssqlbackup
• op5 Monitor
• Nagios

One Response to “Monitor MySQL backup automysqlbackup with op5 Monitor or Nagios”

1. Diigo’ the week (weekly) » The Bipeds’ Monitor Says:
   May 31st, 2009 at 2:10 am

   […] Monitor MySQL backup automysqlbackup with op5 Monitor or Nagios | An It-Slave in the digital saltmin… […]

Leave a Reply