Shell Programming and Scripting

I have a rather simple routine that I would like to fine tune. I am having scripters block if that exists. Trying to tally up that amount of subnets after I performed an awk and uniq from zipped apache logs. For instance, on a solaris 10 system:

This does work
gunzip -c access_log1.200834.gz | awk '{print $1}' | uniq -c | sort -nr > log_tally

log_tally would consist of something like:

2 164.156.43.3
1 10.1.1.3
2 10.1.1.1
756 192.168.1.4
328 192.168.1.4
10 164.156.43.3
2 164.156.43.3
1 10.1.1.3
2 10.1.1.1
1 10.1.1.3
3 10.1.1.1

I would like to consolidate and add the duplicate sunbets so that I can have a top 10 or 20 subnet listing. Not mention sort it in reverse order.

I found this on this forum but I need more sophistication:
awk -F" " '{x +=$1} END {print "SUM: "x}' log_tally

As you can imagine, this will take me a while as I would have to do this for every file with its respective subnet. YUCK.

How would I combine these two routines and give me sums for each deduplicated subnet in reverse order?

Any help would be greatly appreciated.

Have you solved this problem of yours? If not, how do you know what "subnet" an IP address is using??

otheus,

I did fix the issue by using an extra sort command...for instance:

gunzip -c access_log1.200834.gz | awk '{print $1}' | sort | uniq -c | sort -nr | awk '{print $2}' | cut -d "." -f 1-3 | sort | uniq -c |sort -nr > log_tally

Kinda long but it listed and consolidated the top subnets out to the 3rd octet in order.

Thanks for your concern.