This page describes the basic use of cron on HCoop systems.
It is very important that you read the Running Unattended Commands page to understand the general idea of how to run unattended commands.
Making a cron configuration file
Cron needs a special configuration file to tell it how to operate. It is suggested that this file be called ~/.crontab, though it can have any name.
For an explanation of the format of this file, see the crontab manual page (or run man crontab on the shell server).
Managing your cron jobs
Activating your cron configuration
Once you are satisfied with your setup, run the following command to activate your changes, assuming that your configuration file is called ~/.crontab:
Disabling your cron jobs
Just run crontab -r ~/.crontab
Checking your cron configuration
Run crontab -l to display the jobs cron is running for you.
One convention for making scripts to run commands in their own PAG, as specified by the Running Unattended Commands page, is to create the ~/scripts directory, and place your scripts there. Then, make a cron configuration file that looks something like this.
PATH=/afs/hcoop.net/common/bin:/usr/local/bin:/usr/bin:/bin 09 0 * * 0 run-in-pagsh --fg clean-mail ~/scripts/clean-mail 11 0 * * 0 run-in-pagsh --fg remove-tmp ~/scripts/remove-tmp 12 0 * * 0 run-in-pagsh --fg weekly-comment-check ~/scripts/weekly-comment-check */10 * * * * run-in-pagsh --fg make-git-projects ~/scripts/make-git-projects
The PATH line is needed so that you have access to run-in-pagsh. Make sure you don't try to reference any other environment variables like $HOME, because cron will not expand them properly.
The first cron command runs something 9 minutes after midnight (aka the "0th" hour), every day of the month, every month, provided that it is Sunday (aka the "0th" day of the week). In short: run something at 12:09 every Sunday.
The second and third commands are similar.
The fourth command runs something every 10 minutes.