My advice is to avoid using threads, especially as you are new to Perl.

It is unlikely that processing the logs in threads will be much faster than processing the logs from a single threaded program launched as multiple instances. Most of the time of your programm will likely be spent on disk IO anyway.

I recommend writing the program that processes a single log file first. Then I would launch that single program through runN, which will process all the files in parallel with it.

As for logging, I recommend writing the output to a separate log file for every input file, and merging these log files after all files have been processed.

kindly give some sample code for this threading in perl.

I could have sworn there was an example perlthrtut ... maybe these will help

• Perl Threads Boss/Worker Example
• thread/fork boss/worker framework recommendation
• threads: work crew memory leak
• WorkCrew threading
• Re: TCP Server hangs with multiple client connections, Thread::Queue

now split it to 5 parts and then process this 5 parts at simultaneously.

Why 5 parts? (Rather than 4 or 8 or ... )

Since you’re using Unix, do it “the Unix way.”   Write a script that does its part, then invoke as many parallel-executing copies of that script from another script or the command-line.   The program does not take on all the fancy-pants work of doing everything:   the shell does; the user does.   Each instance of the worker-bee (Perl) script simply finds another unit of work to do and does it, running until it finds no more work to do.   It has no awareness of any program other than itself.