There is a production mainframe job that copy a file between 2AM to 2:30AM on Unix. Sometimes, the job fails without giving out any error.
I would like to mimic the copy to find out the error. So, I am thinking to schedule a unix job to copy a small file ( or ftp a file or something similar) between 2AM to 2:30AM. I don't want to copy a large file though for 30 minutes. Is there a way to consistently copy a small file over and over again or copy a small file slowly? What is the command to catch error message when error occurs?