Hi all,

I have a log file of the below format.

[asdfg@st9872] 20081016:000042 asdflasjdf asljfljs asdflasjf safjl
[kronos@jd0121] 20081016:000229 /lask/ajlsdf/askdfjsa
[cak416@po0121] 20081016:000229 /lashflas /askdfaslj hsfhsahf
[close@as0121] 20081016:000304 lasflasj ashfashd
[cak416@sadfjk0121] 20081016:000304 lajfasdf ashfashdfhs

I need to generate a report based on the log file based on the timestamp. I need to search for particular errror messages that were encountered on an hourly basis from the log file. The second field here is the timestamp which is of the form yyyymmdd:hhmmss eg:20081016:000042. I need to create hour by hour statistics from this log file.

like errors encountered between 20081016:000042 and 20081016:230000

I tried using awk inside a shell script, but it was crude way doing things and it depended on the user input for the time, but what I need the script to do is to capture the information automatically from the log file based an hour by our basis.

below is the script that I tried. I am new to awk and I am clueless of doing this.

the variables nmbr1 and nmbr2 are being asked from the user through a shell script.

filename="/opt/tmp/ca/fds/logs/proof.out_20"
filename2="${filename}${dt}"
awk -v nmbr1=${nmbr1} -v nmbr2="$nmbr2" -v outfile="$outfile" '
{
FS=":";
if ((($2-$6) > nmbr1) && (($2-$6) < nmbr2)){
print $2,$3 >>outfile;
}
}
END {}' $filename2

#/home/i4zuss/proof.out_20081016
result=`grep "returned 7" $outfile | wc -l`
echo "Failure Return code 7 =" $result >>/tmp/results.out;
result=`grep "returned 5" $outfile | wc -l`
echo "Failure Return code 5 =" $result >>/tmp/results.out;
result=`grep "Max=30" $outfile | grep "acm" | wc -l`

Let me know if you guys have any idea abt how to solve this.