DEFINE VARIABLES
PING_TARGET="google.ca"
LOG_DIR=/var/log/conn
LOG_FILE=$LOG_DIR"/conn.log"
Here, I am just setting a few variables up near the top of the script. It makes changing them a little easier.
PING_COMMAND="ping -W 1 -c 1"
This is the basic command that I will be using.
-W 1 - set a timeout on the ping for 1000 miliseconds (1 second)
-c 1 - send only 1 ping
LOG DIRECTORY AND FILE
if [ "`id -u`" != "0" ]
then
echo "You need root privileges!"
exit
fi
if [ ! -e $LOG_DIR ]
then
mkdir $LOG_DIR
fi
if [ ! -e $LOG_FILE ]
then
touch $LOG_FILE
chmod 757 $LOG_FILE
echo "Date/Time, Uptime (in secs), CPU Load, Next Hop IP, Ping Time (in ms), Target IP, Ping Time (in ms)," >> $LOG_FILE
fi
First, check to make sure the script is being run with root privileges. Then, check for the log directory then the log file; if either is missing, create.
LOOP
COUNTER=1
while [ $COUNTER -le 6 ]
do
...
COUNTER=$(( COUNTER+1 ))
sleep 10
done
I want my script to send a series of pings every 10 seconds. Since cron jobs can only run every minute at max, I need to set up a loop to run the commands for a total of 6 times, every 10 seconds. That is what this loop consists of.
COMMANDS (all within the loop)
FINAL_LOG=$(date),
Log the date/time stamp.
UPTIME=$(cat /proc/uptime | awk '{ print $1 }')
FINAL_LOG+=$UPTIME,
Log the ping tester's uptime.
CPU=$(cat /proc/loadavg | awk '{print $1}')
FINAL_LOG+=$CPU,
Log the device's current CPU usage. One of the possibilities here is that the issues *MAY* be (or partially) due to CPU load on the device. This logs the avg CPU usage for the past 60 seconds.
DEFAULT_ROUTE=$(/sbin/route | grep default | awk '{ print $2 }')
Log the default route IP.
wait
Wait for all the above commands to complete
if [ -z "$DEFAULT_ROUTE" ]
Check if the default route exists.
then
FINAL_LOG+="No Default Route, ,";
If not, log that info.
else
PING_DR="$($PING_COMMAND $DEFAULT_ROUTE 2>&1)"
Ping the default gateway.
DR_IP=$(awk -F " " '{print $3;}' <<< $PING_DR)
FINAL_LOG+=$DR_IP,
Extract the IP address from the ping response and log.
DRPINGTIME=$(awk -F " " '{print $14;}' <<< $PING_DR)
DRPINGTIME=${DRPINGTIME:5}
FINAL_LOG+=$DRPINGTIME,
Extract the ping time to the default gateway.
fi
PING_OUTPUT="$($PING_COMMAND $PING_TARGET 2>&1)"
Ping the PING_TARGET
if [[ $PING_OUTPUT == *"unknown host"* ]]
then
FINAL_LOG+="DNS ERROR, ,";
If the ping output shows "unknown host", this would be due to a DNS failure. Log this info if this is the case
elif [[ $PING_OUTPUT == *"0 received"* ]]
then
FINAL_LOG+="Ping Timeout, ,";
If the ping timed out, log that info
else
PING_IP=$(awk -F " " '{print $3;}' <<< $PING_OUTPUT)
FINAL_LOG+=$PING_IP,
Log the actual IP Address we are pinging (since we are pinging a DNS name, the IP is likely to change).
PINGTIME=$(awk -F " " '{print $15;}' <<< $PING_OUTPUT)
PINGTIME=${PINGTIME:5}
FINAL_LOG+=$PINGTIME,
Again, extract and log the ping response time.
fi
echo $FINAL_LOG >> $LOG_FILE
Log the data to the log file.
I am sure that someone, somewhere will look at this script and cringe, or make faces... Likely because there already exists something that will do something somewhat similar.But I needed something very lightweight, requiring little to no install and setup, useable on just about any/all Linux machines, and using (as much as possible) built-in commands like ping. This is also serving as an exercise in Linux and shell scripting. So please, feel free to download it, look it over, poke and prod at it, and give me your thoughts below.
 Ping Connectivity
1
|
|
 Monday, 07 September 2015 15:57  1.17 KB  411 |
Download |
