Cron jobs are typically configured using five time fields (minutes, hours, day of month, month, day of week). For running every 2 days at 11 PM, we need to focus on the day-of-month field (*/2) combined with the specific hour (23).
# Example crontab entry (runs at 11:00 PM every 2 days)
0 23 */2 * * /path/to/your/script.sh
For systems requiring different interval logic, consider these alternatives:
# Method 1: Using day-of-month wildcard
0 23 1-31/2 * * /path/to/script
# Method 2: Date-based execution check
0 23 * * * [ $(($(date +\%d) \% 2)) -eq 0 ] && /path/to/script
When your cron job isn't working:
- Check system logs:
grep CRON /var/log/syslog - Verify cron service status:
systemctl status cron - Test command execution manually
- Check file permissions and paths
- Use absolute paths in scripts
- Redirect output to log files
- Include error handling in scripts
- Consider using systemd timers for complex schedules
# Example with logging
0 23 */2 * * /usr/bin/python3 /opt/scripts/nightly.py >> /var/log/nightly.log 2>&1
For servers in different time zones:
# Set system time zone
sudo timedatectl set-timezone America/New_York
# Or specify TZ in crontab
TZ=UTC
0 23 */2 * * /path/to/script
Cron jobs on CentOS/RHEL systems follow a specific time notation format:
* * * * * command_to_execute - - - - - | | | | | | | | | +----- Day of week (0-6) (Sunday=0) | | | +------- Month (1-12) | | +--------- Day of month (1-31) | +----------- Hour (0-23) +------------- Minute (0-59)
For running a job at 11 PM every two days, we have several approaches:
Method 1: Using Day-of-Month Increment
0 23 */2 * * /path/to/your/script.sh
This will run at 11 PM on even-numbered days (2nd, 4th, etc.). Note that months with 31 days will see consecutive executions on 30th and 31st.
Method 2: Date-Based Conditional Execution
0 23 * * * [ $(($(date +\%d)\%2)) -eq 0 ] && /path/to/your/script.sh
The arithmetic operation checks if current day is even before executing.
Method 3: Using Systemd Timers (Modern Alternative)
# /etc/systemd/system/your-service.timer [Unit] Description=Run every 48 hours at 11PM [Timer] OnCalendar=*-*-* 23:00:00 AccuracySec=1h Persistent=true Unit=your-service.service [Install] WantedBy=timers.target
If your scheduled job isn't running as expected:
- Verify cron service status:
systemctl status crond - Check mail for cron output:
mail -u your_user - Redirect output to log file:
0 23 */2 * * /path/script.sh >> /var/log/cron.log 2>&1 - Confirm script permissions:
chmod +x /path/script.sh - Test environment variables: Cron runs with minimal env - specify full paths
For complex scheduling requirements, consider a state file approach:
0 23 * * * /usr/bin/flock -n /tmp/yourjob.lock /bin/bash -c '[[ -f /var/run/yourjob.flag ]] && rm /var/run/yourjob.flag || (touch /var/run/yourjob.flag && /path/script.sh)'
This creates a toggle mechanism that alternates execution days while preventing overlapping runs.