Hi

Is it possible to have a script run in a loop (waiting for a change of state in the network interface), and if the loop continues for five minutes, to have it email the admin, and carry on in the loop?

Here is my loop:

#!/bin/bash

STATE=`echo "show State:/Network/Interface/bond0/Link" | scutil | grep -c "TRUE"`

while [ ! ${STATE} -eq 1 ]
do
echo "Link is inactive"
/bin/sleep 2
done

echo "Link is alive"

#!/bin/bash

STATE=`echo "show State:/Network/Interface/bond0/Link" | scutil | grep -c "TRUE"`

while [ ! ${STATE} -eq 1 ]
do
echo "Link is inactive"
/bin/sleep 2
STATE=`echo "show State:/Network/Interface/bond0/Link" | scutil | grep -c "TRUE"
done

echo "Link is alive"

Hey aju_kup

Thanks for pointing that error in the loop out to me I should know about things like that from previous posts.

Does anyone know how to answer my original query?

Mike

Hey all

As a follow-up, I figured it out. So, if anyone is searching for a similar solution, here is my proposed script in its entireity. It is used within OS X. Note that after 5 minutes, I actually exit the script and do not continue:

#!/bin/bash

# Script to call external monitor scripts
# This is run via Launchd

###
# Setup Email
###
C_NAME=`/usr/sbin/scutil --get ComputerName`
NOTIFY="admin@admin.com"
WARNING="There is a problem with the NIC's on the primary Moodle server."
###
# End setup Email
###

TIME="0"
# The loop is for 150 times
# 150x2(sleep time)=300s=5minutes
# If this script continues to fail
# after 5 minutes it will finish the loop
# and then email the admin with necessary info
		
while [ ${TIME} -ne 150 ]
do
        
	# Sleep for 2 seconds in between to preserve
	# processor hogging

   	/bin/sleep 2
      	
	# Check state of aggregated link and interfaces associated
	# with it. Apache will not run unitl NICs are up and running
	# so the remainder of this script will fail.
		
	if [ ! $(echo "show State:/Network/Interface/bond0/Link" | scutil \
	| grep -c "TRUE") -eq 1 ]
	then
		echo Loop ${TIME}
                echo "Interface is not active"
       	        TIME=$(( ${TIME} + 1 ))
   	else 
		echo "Interface is alive"

		# Sleep to allow Apache time to start.
		# If Apache fails after this amount of
		# time, then there is something wrong
		# and shutdown_services script will be
		# called to initiate failover
			
		/bin/sleep 30

		echo "Running monitor scripts"
			
		# Call check_apache.sh script
		APACHE="/Library/IPFailover/scripts/check_apache.sh"
			
		if [ -x ${APACHE} ]
		then
			/Library/IPFailover/scripts/check_apache.sh
		else
			echo "check_apache.sh does not exist - moving to next script"
			logger \
			"FAILOVER: check_apache.sh does not exist - moving to next script"
		fi

		# Call check_php.sh script
		PHP="/Library/IPFailover/scripts/check_php.sh"
			
		if [ -x ${PHP} ]
		then
			/Library/IPFailover/scripts/check_php.sh
		else
			echo "check_php.sh does not exist - moving to next script"
			logger \
			"FAILOVER: check_php.sh does not exist - moving to next script"
		fi

		# Call check_postgres.sh script
		POSTGRES="/Library/IPFailover/scripts/check_postgres.sh"
			
		if [ -x ${POSTGRES} ]
		then
			/Library/IPFailover/scripts/check_postgres.sh
		else
			echo "check_postgres.sh does not exist - moving to next script"
			logger \
			"FAILOVER: check_postgres.sh does not exist - moving to next script"
		fi

		# Exit the script and therefore the loop
       	        # Using "break" would only exit the loop
		# and not exit the script
       	        # We don't need to email Admins
       	        # when the script is successful
                        
		exit 0
	fi
done

echo "Finished looping"
echo "Mailing the admin"
logger "FAILOVER: Problem with Network Interfaces - mailing the admin and exiting monitor.sh script"

# Mail the System Administrator

echo From: $C_NAME - ${WARNING} > /tmp/monitor_warning.txt
echo ----- >> /tmp/monitor_warning.txt
echo "Note that there has been a bad response from the "monitor" shell script."\
>> /tmp/monitor_warning.txt
echo "This means that there is a problem with the Network Interfaces of $C_NAME providing the Moodle service." \
>> /tmp/monitor_warning.txt
echo ----- >> /tmp/monitor_warning.txt
echo "This error was written out at `date`." >> /tmp/monitor_warning.txt
echo ----- >> /tmp/monitor_warning.txt
echo "Please attend to this immediately." >> /tmp/monitor_warning.txt
cat /tmp/monitor_warning.txt | mail -s \
"From: ${C_NAME} - Primary NIC's are down" ${NOTIFY}

# Should I instigate failover here?

exit 0