If you aren't able to modify @INC or get modules installed, and have a lot of data to sort, perl may not be the best answer for you.

Do you have Unix on that AIX box? If so, I recommend sticking with the Unix sort program. A command like:

sort file.txt > file.sorted.txt

should do the trick.

Unix sort is handy with large files, breaking them down to many smaller files and reassembling them as it moves. It uses a good balance of cpu, ram, and drive space, in my experience. I used to use it to help sift through data on some syslog servers at my old job, and actually replaced a production perl script with a Unix grep/sort one-liner due to the performance boost (over my sloppy perl script, that is).