Scheduling Tasks with cron and at

1. One-Time Tasks with at

Imagine you need to run a resource-intensive script, but you want it to execute after business hours to avoid impacting users. Or perhaps you need to restart a server in 10 minutes but want to log off now. This is where the at command is perfect. It schedules a job for a single future execution.

The at command relies on a background service, or daemon, called atd. Just like the services we managed in the previous lesson, atd must be running for your scheduled jobs to execute. You can verify its status with sudo systemctl status atd.

Let's see how to use it in practice.

Key at Commands

To summarize the video and provide a quick reference:

The batch command, mentioned briefly in the video, is part of the at system. It schedules jobs to run only when the system load is low. While it's good to know it exists, it's far less common in modern practice than at or cron.

2. Recurring Tasks with cron

While at is for one-off tasks, the real power of automation comes from cron, which handles recurring jobs. cron is driven by a daemon, crond, that wakes up every minute to check for scheduled tasks to run.

The cron Architecture

Before diving into the commands, it's helpful to understand the overall structure of cron on a Linux system, which aligns with your preference for a theoretical overview. There isn't just one place to define a cron job.

We'll focus on the two most important methods for an administrator:

1. User crontab files: Each user can have their own list of scheduled jobs.
2. System-wide cron files: Located in /etc, these are for system-level tasks.

Method 1: User crontab (The Common Way)

As a user or administrator, you will most often manage jobs using your own "cron table" or crontab. You interact with it using the crontab command.

• crontab -e: Edit your crontab file.
• crontab -l: List the current jobs in your crontab.
• crontab -r: Remove your entire crontab (use with caution!).

The most challenging part of using cron is the syntax for specifying the schedule. A cron job definition is a single line with six fields: five for the time and one for the command.

Let's break down the syntax with a detailed walkthrough.

The five fields, in order, are:

1. Minute (0-59)
2. Hour (0-23)
3. Day of Month (1-31)
4. Month (1-12)
5. Day of Week (0-7, where both 0 and 7 represent Sunday)

An asterisk (*) in a field means "every". So, * * * * * means every minute of every hour of every day...

Now, let's learn about the special operators that make cron so flexible.

A fantastic tool for building and verifying your cron schedules is Crontab Guru. You can type in a schedule and it will translate it into plain English. Both videos recommended it, and it's a great resource to bookmark.

Method 2: System-wide cron

System-wide cron jobs are intended for services and system administration tasks, not individual users. You'll find them in a few places in the /etc/ directory. The key difference is that system cron files require a sixth field specifying the username to run the command as.

To recap the system-wide locations:

• /etc/crontab: The main system crontab file. You should generally avoid editing this directly.
• /etc/cron.d/: The preferred location to place custom system-wide cron files. Each file here must specify the user to run the job as.
• /etc/cron.daily/, /etc/cron.weekly/, etc.: Simple directories. Any executable script you place here will be run daily, weekly, etc. You don't use cron syntax in these files.

3. Access Control with .allow and .deny

On a multi-user system, you may want to control who can schedule at and cron jobs. This is managed through two simple files for each service: at.allow/at.deny and cron.allow/cron.deny.

Conclusion

You have now learned how to harness two of Linux's most powerful automation tools. This moves you from being a reactive administrator to a proactive one, saving time and ensuring consistency.

Key Takeaways:

• at is used to schedule a single, one-time task in the future. You manage it with at, atq, and atrm.
• cron is used for recurring, scheduled tasks. The crond daemon runs every minute to check for jobs.
• User crontabs are managed with crontab -e and use a five-field time format.
• System crontabs (like in /etc/cron.d/) add a sixth field for the username.
• Special directories like /etc/cron.daily offer a simpler way to schedule scripts without using cron syntax.
• Access to at and cron can be restricted for users via .allow and .deny files in /etc/.

Preview of the Next Lesson:

We've just learned how to schedule commands, but what if you need to run a sequence of commands with logic, variables, and conditions? In our next lesson, we will dive into shell scripting, where you'll learn to write your own basic bash scripts—the perfect tool to be scheduled with cron