Hi, I am a newer in Bourne shell. I have a log file which contain many lines(information about file used), now I want to count the 10 most requested files in the log, and display the 10 in descending order. I don't know how to count the duplicate files and sort them in descending order, any command for that purpose? can anybody tell me how to accomplish that?

If you are talking about monitoring web traffic, there are plenty of log analyzing tools already. sort -u will give you unique results although you would want to only look at the portion of each log entry to would give you a unique result, ie exclude the date/time stamp and who asked for the file. awk will help you look at portions of a line. grep -c will give you a count of occurances. sort -n -r will sort numerically in decending order.