Greetings!
I have a simple script running on my firewall that simply sends a
single ping packet to a web site I trust every minute. If the ping
fails, I assume my DSL line went down... again.
All works fine, until the router sends back a "no route to host" type
message. When this happens, the ping command simply hangs. It never
returns and I quickly have a few hundred ping commands sleeping in my
process table.
I get around this problem by running a script every hour that looks for
> 5 ping commands and runs a killall if it is true. However, is this a
bug in ping? Since I am seeing the "no route" message, I have to assume
ping knows what it is getting, it simply never exits.
Here's my script. (Be kind. I'm not a programmer.)
#!/bin/sh
ret_var=0
LOG_FILE=/var/log/ping_log
TARGET_IP="www.kcpt.org"
# echo log to $LOG_FILE
if ping -c 1 $TARGET_IP >> /dev/null { If you put >> $LOG_FILE here
you can see the no route stuff }
{ The lines below are never executed if the ping packet returns no
route to host message. }
then
echo `date` "Ping Succeeded" $TARGET_IP >> $LOG_FILE
else
echo `date` "Ping Failed" $TARGET_IP >> $LOG_FILE
fi
{ Notification stuff goes here, but you get the idea. }
Thanx!
-Michael
