Cron has been scheduling Unix tasks since 1975. It’s simple, reliable, and will silently fail in ways that waste hours of debugging. Here’s how to use it properly.

Cron Syntax

│││││└─││││└───│││└─────││└───────│└─────────c─────omDMDHMmaoaoiaynyunntrudohotff(e(0w1m-(e-o20e1n3-k2t)5)h9()0(-17-,310)and7areSunday)

Common Schedules

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

# Every minute
* * * * * /script.sh

# Every 5 minutes
*/5 * * * * /script.sh

# Every hour at minute 0
0 * * * * /script.sh

# Daily at 3 AM
0 3 * * * /script.sh

# Weekly on Sunday at midnight
0 0 * * 0 /script.sh

# Monthly on the 1st at 6 AM
0 6 1 * * /script.sh

# Every weekday at 9 AM
0 9 * * 1-5 /script.sh

# Every 15 minutes during business hours
*/15 9-17 * * 1-5 /script.sh

The Silent Failure Problem

Cron’s default behavior: run command, discard output, send errors to email (which you probably haven’t configured).

Always redirect output:

1
2
3
4
5
6
7
8

# Log everything
0 3 * * * /backup.sh >> /var/log/backup.log 2>&1

# Log with timestamp
0 3 * * * /backup.sh 2>&1 | ts '[%Y-%m-%d %H:%M:%S]' >> /var/log/backup.log

# Discard output only if you really don't care
0 * * * * /cleanup.sh > /dev/null 2>&1

Environment Gotcha

Cron runs with a minimal environment. Your script works in terminal but fails in cron because:

  • PATH is minimal (/usr/bin:/bin)
  • No user profile loaded
  • No environment variables from .bashrc

Fix 1: Use full paths

1
2
3
4
5

# Bad
0 * * * * python /scripts/task.py

# Good
0 * * * * /usr/bin/python3 /scripts/task.py

Fix 2: Set PATH in crontab

1
2
3
4

PATH=/usr/local/bin:/usr/bin:/bin
SHELL=/bin/bash

0 * * * * /scripts/task.sh

Fix 3: Source environment in script

1
2
3

#!/bin/bash
source /home/user/.bashrc
# rest of script

Lock Files (Prevent Overlap)

If a job runs longer than its interval, you get overlapping runs:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15

#!/bin/bash
LOCKFILE=/tmp/myjob.lock

# Exit if already running
if [ -f "$LOCKFILE" ]; then
    echo "Already running"
    exit 1
fi

# Create lock
trap "rm -f $LOCKFILE" EXIT
touch "$LOCKFILE"

# Your actual work here
/do/the/thing

Or use flock:

1

0 * * * * flock -n /tmp/myjob.lock /scripts/myjob.sh

The -n flag makes it fail immediately if lock is held.

Timeout Protection

Prevent runaway jobs:

1

0 * * * * timeout 3600 /scripts/long-job.sh

Kills the job after 1 hour.

Error Notification

Get notified when things fail:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

#!/bin/bash
set -e  # Exit on error

cleanup() {
    if [ $? -ne 0 ]; then
        echo "Job failed at $(date)" | mail -s "Cron failure: backup" admin@example.com
    fi
}
trap cleanup EXIT

# Your job here
/run/backup

Or use a monitoring service:

1

0 3 * * * /backup.sh && curl -s https://hc-ping.com/your-uuid

Logging Best Practices

 1
 2
 3
 4
 5
 6
 7
 8
 9
10

#!/bin/bash
LOG=/var/log/myjob.log
exec 1>>$LOG 2>&1  # Redirect all output to log

echo "=== Started at $(date) ==="

# Your commands here
/do/stuff

echo "=== Finished at $(date) ==="

Rotate logs:

1
2
3
4
5
6
7
8

# /etc/logrotate.d/myjob
/var/log/myjob.log {
    weekly
    rotate 4
    compress
    missingok
    notifempty
}

Testing Cron Jobs

Test the command manually first:

1
2

# Run exactly as cron would
env -i /bin/bash -c '/path/to/script.sh'

Check cron is running:

1
2
3

systemctl status cron
# or
systemctl status crond  # RHEL/CentOS

Watch the logs:

1
2
3

tail -f /var/log/syslog | grep CRON
# or
journalctl -u cron -f

List your crontab:

1

crontab -l

User vs System Crontabs

User crontab (crontab -e):

  • Stored in /var/spool/cron/crontabs/
  • Runs as that user
  • No user field in syntax

System crontab (/etc/crontab):

  • Has extra user field
  • Runs as specified user
1
2
3
4
5

# /etc/crontab has user field
0 3 * * * root /scripts/backup.sh

# User crontab doesn't
0 3 * * * /scripts/backup.sh

Drop-in directory (/etc/cron.d/):

  • Like system crontab (has user field)
  • One file per job
  • Easier to manage with config management

Cron Alternatives

Systemd timers (modern Linux):

 1
 2
 3
 4
 5
 6
 7
 8
 9
10

# /etc/systemd/system/backup.timer
[Unit]
Description=Daily backup

[Timer]
OnCalendar=daily
Persistent=true

[Install]
WantedBy=timers.target

1
2
3
4
5
6
7

# /etc/systemd/system/backup.service
[Unit]
Description=Backup job

[Service]
Type=oneshot
ExecStart=/scripts/backup.sh

1
2

systemctl enable --now backup.timer
systemctl list-timers

Benefits: better logging, dependency management, resource controls.

Common Mistakes

Wrong time zone:

1
2
3

# Set in crontab
TZ=America/New_York
0 9 * * * /morning-job.sh

Day-of-month AND day-of-week:

1
2
3
4

# This runs if EITHER matches (not both!)
0 0 15 * 5  # 15th of month OR Fridays

# For "15th if it's a Friday", use a wrapper script

Forgetting the newline: Crontab needs a trailing newline after the last entry.

Special characters in commands:

1
2

# Percent signs need escaping
0 * * * * echo "$(date +\%Y-\%m-\%d)" >> /log

Production Checklist

  • Full paths for all commands
  • Output redirected to log file
  • Lock file prevents overlap
  • Timeout prevents runaway
  • Error notification configured
  • Log rotation set up
  • Tested with minimal environment
  • Documented what the job does

Cron is a 50-year-old tool that still runs critical infrastructure. Respect it by setting it up properly.