Guides

Cron Jobs Explained: Schedule Backups and Log Cleanup

By Ravindra Bagale, DevOps trainer · LinkedIn · Published: 28 October 2026 · Last updated: 28 October 2026

Cron schedules backups and log cleanup on Linux. Learn fields, redirects and PATH fixes used on EC2 and on-prem hosts.

Language: English · मराठी · हिंदी

Why cron still matters

Cloud schedulers exist, but every Amazon EC2 and on-prem box still needs host-local cleanup. Missed backups cost INR and dollars alike.

Why: Unattended work must be reliable. What: crontab fields, backup and cleanup jobs, failure fixes. How: Build a script, schedule it, prove the log.

Concept 1: Crontab fields

Why: Wrong hour means wrong night. What: min hour dom mon dow + command. How: Memorize with one daily example.

Example 1: Crontab fields

Type this:

printf '%s\n' 'min hour dom mon dow command' '0 2 * * * /home/ec2-user/bin/backup.sh'

You will see:

min hour dom mon dow command
0 2 * * * /home/ec2-user/bin/backup.sh

What it means: 0 2 * * * means 02:00 every day.

Example 2: List crontab

Type this:

crontab -l 2>/dev/null || echo no-crontab-yet

You will see:

no-crontab-yet

What it means: Empty until you install one.

Example 3: Five field quiz

Type this:

python3 -c "print('*/5 * * * * -> every 5 minutes')"

You will see:

*/5 * * * * -> every 5 minutes

What it means: Star means every unit.

Concept 2: Backup and cleanup scripts

Why: Cron only launches; the script does the work. What: tar backup; find for old logs. How: Dry-run deletes with -print.

Example 4: Backup script stub

Type this:

printf '%s\n' '#!/usr/bin/env bash' 'set -euo pipefail' 'tar -czf /tmp/app-$(date +%F).tgz /var/www/app' > /tmp/backup.sh
chmod 700 /tmp/backup.sh
head -3 /tmp/backup.sh

You will see:

#!/usr/bin/env bash
set -euo pipefail
tar -czf /tmp/app-$(date +%F).tgz /var/www/app

What it means: Date-stamped tarball pattern.

Example 5: Log cleanup find

Type this:

printf '%s\n' 'find /var/log/myapp -name "*.log" -mtime +7 -delete'

You will see:

find /var/log/myapp -name "*.log" -mtime +7 -delete

What it means: Delete logs older than 7 days — test with -print first.

Example 6: Dry-run delete

Type this:

printf '%s\n' 'find /tmp/demo-logs -name "*.log" -mtime +7 -print'

You will see:

find /tmp/demo-logs -name "*.log" -mtime +7 -print

What it means: Always print before delete on production.

Concept 3: Redirects and MAILTO

Why: Silent cron jobs hide failure. What: >> log 2>&1; optional MAILTO=. How: Always capture stderr.

Example 7: Redirect pattern

Type this:

printf '%s\n' '0 2 * * * /home/ec2-user/bin/backup.sh >> /var/log/backup.log 2>&1'

You will see:

0 2 * * * /home/ec2-user/bin/backup.sh >> /var/log/backup.log 2>&1

What it means: 2>&1 merges stderr into the log.

Example 8: MAILTO off

Type this:

printf '%s\n' 'MAILTO=' '0 3 * * 0 /home/ec2-user/bin/weekly.sh'

You will see:

MAILTO=
0 3 * * 0 /home/ec2-user/bin/weekly.sh

What it means: Empty MAILTO stops email floods.

Example 9: Absolute interpreter

Type this:

printf '%s\n' '0 2 * * * /usr/bin/bash /home/ec2-user/bin/backup.sh'

You will see:

0 2 * * * /usr/bin/bash /home/ec2-user/bin/backup.sh

What it means: Avoid relying on cron PATH.

Concept 4: Common failures

Why: PATH and permissions cause most tickets. What: absolute paths; matching user; verify syslog. How: Test by hand before cron.

Example 10: Common failure PATH

Type this:

printf '%s\n' 'ERROR: aws: command not found' 'FIX: /usr/local/bin/aws or source profile'

You will see:

ERROR: aws: command not found
FIX: /usr/local/bin/aws or source profile

What it means: Cron does not load your interactive shell profile by default.

Example 11: Permission failure

Type this:

printf '%s\n' 'Permission denied' 'FIX: script owned by same user; mode 700'

You will see:

Permission denied
FIX: script owned by same user; mode 700

What it means: Root cron vs ec2-user paths differ.

Example 12: Verify last run

Type this:

printf '%s\n' 'grep backup /var/log/syslog | tail -3' '# or journalctl on some hosts'

You will see:

grep backup /var/log/syslog | tail -3
# or journalctl on some hosts

What it means: Confirm the job actually fired.

Install pattern

1. Write /home/ec2-user/bin/backup.sh
2. chmod 700
3. crontab -e and add the line with redirects
4. Wait for the minute or trigger with a short schedule in lab
5. Read the log

Suppose: Retail app in Pune

Nightly tar of /var/www to /backup. Rotate after 7 days.

Suppose: US SaaS worker

Clean /var/log/myapp weekly. PATH bug fixed with /usr/bin/find.

Check yourself (quiz)

Fields → script → crontab → logs → verify. Cron is still core DevOps. Did you understand? Let us try it once more.