Backup Rotation Script in Bash

Overview

Backup rotation keeps a fixed number of backups and deletes old ones automatically. This script uses tar for compression and find for cleanup. It supports daily, weekly, and monthly retention policies so you always have recent and historical backups without filling the disk.

When to Use

• You need automated backups for a web app or database
• You want to keep daily backups for 7 days, weekly for 4 weeks, monthly for 6 months
• You are setting up a cron job for periodic backups
• You want compressed archives with automatic cleanup

Solution

Basic backup with timestamp

#!/bin/bash

SOURCE_DIR="/var/www/myapp"
BACKUP_DIR="/backups"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)

tar -czf "${BACKUP_DIR}/backup_${TIMESTAMP}.tar.gz" -C "${SOURCE_DIR}" .
echo "Backup created: backup_${TIMESTAMP}.tar.gz"

Backup with retention policy

#!/bin/bash

SOURCE_DIR="/var/www/myapp"
BACKUP_DIR="/backups"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="${BACKUP_DIR}/backup_${TIMESTAMP}.tar.gz"

# Create backup
tar -czf "${BACKUP_FILE}" -C "${SOURCE_DIR}" .

# Retention: keep last 7 daily backups
find "${BACKUP_DIR}" -name "backup_*.tar.gz" -type f -mtime +7 -delete

echo "Backup created: ${BACKUP_FILE}"
echo "Old backups cleaned (kept last 7 days)"

Tiered retention (daily, weekly, monthly)

#!/bin/bash

SOURCE_DIR="/var/www/myapp"
BACKUP_DIR="/backups"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
DAY_OF_WEEK=$(date +%u)  # 1=Monday, 7=Sunday
DAY_OF_MONTH=$(date +%d)

# Create daily backup
tar -czf "${BACKUP_DIR}/daily_${TIMESTAMP}.tar.gz" -C "${SOURCE_DIR}" .

# Weekly full backup on Sunday
if [ "$DAY_OF_WEEK" = "7" ]; then
    cp "${BACKUP_DIR}/daily_${TIMESTAMP}.tar.gz" "${BACKUP_DIR}/weekly_${TIMESTAMP}.tar.gz"
fi

# Monthly full backup on the 1st
if [ "$DAY_OF_MONTH" = "01" ]; then
    cp "${BACKUP_DIR}/daily_${TIMESTAMP}.tar.gz" "${BACKUP_DIR}/monthly_${TIMESTAMP}.tar.gz"
fi

# Cleanup: 7 daily, 4 weekly, 6 monthly
find "${BACKUP_DIR}" -name "daily_*.tar.gz" -type f -mtime +7 -delete
find "${BACKUP_DIR}" -name "weekly_*.tar.gz" -type f -mtime +28 -delete
find "${BACKUP_DIR}" -name "monthly_*.tar.gz" -type f -mtime +180 -delete

echo "Backup completed with tiered retention"

Database backup with rotation

#!/bin/bash

DB_NAME="myapp"
DB_USER="postgres"
BACKUP_DIR="/backups/db"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)

mkdir -p "${BACKUP_DIR}"

# Dump database
pg_dump -U "${DB_USER}" "${DB_NAME}" | gzip > "${BACKUP_DIR}/db_${TIMESTAMP}.sql.gz"

# Keep last 14 days
find "${BACKUP_DIR}" -name "db_*.sql.gz" -type f -mtime +14 -delete

echo "Database backup created: db_${TIMESTAMP}.sql.gz"

Cron setup

# Run daily at 2 AM
0 2 * * * /opt/scripts/backup.sh >> /var/log/backup.log 2>&1

# Run every 6 hours
0 */6 * * * /opt/scripts/backup.sh >> /var/log/backup.log 2>&1

Backup with integrity check

#!/bin/bash

SOURCE_DIR="/var/www/myapp"
BACKUP_DIR="/backups"
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
BACKUP_FILE="${BACKUP_DIR}/backup_${TIMESTAMP}.tar.gz"

# Create backup
tar -czf "${BACKUP_FILE}" -C "${SOURCE_DIR}" .

# Verify integrity
if gzip -t "${BACKUP_FILE}" 2>/dev/null; then
    echo "OK: ${BACKUP_FILE} verified"
    find "${BACKUP_DIR}" -name "backup_*.tar.gz" -type f -mtime +7 -delete
else
    echo "ERROR: ${BACKUP_FILE} is corrupt"
    rm -f "${BACKUP_FILE}"
    exit 1
fi

Explanation

The find command with -mtime +N matches files older than N days. -delete removes them. This is the simplest retention mechanism: every time the script runs, it creates a new backup and cleans files older than the retention window.

Tiered retention uses file naming prefixes (daily_, weekly_, monthly_) to apply different retention periods. The weekly backup is a copy of Sunday’s daily backup. The monthly backup is a copy of the 1st’s daily backup. This avoids creating multiple large archives on the same day.

gzip -t tests archive integrity without extracting. Run it after every backup to catch disk errors or incomplete writes.

Variants

Approach	Tool	Compression	Use When
tar + find	tar, gzip	gzip	Filesystem backups
rsync + find	rsync	None (sync)	Incremental backups
pg_dump + gzip	pg_dump	gzip	PostgreSQL backups
mysqldump + gzip	mysqldump	gzip	MySQL backups
restic	restic	Deduplication	Large-scale, encrypted backups

Guidelines

• Always test backup restoration. A backup you cannot restore is useless.
• Use gzip -t to verify archive integrity after creation.
• Store backups on a separate disk or remote server. A backup on the same disk fails when the disk fails.
• Set retention policies based on recovery point objectives (RPO). 7 daily + 4 weekly + 6 monthly covers most needs.
• Log backup operations. Append output to a log file via cron redirection.

Common Mistakes

• Not testing restores. Run a restore drill monthly to verify backups work.
• Storing backups on the same disk as the source. Disk failure loses everything.
• Using rm instead of find -delete. find -delete is safer because it only matches the pattern.
• Not setting mkdir -p for the backup directory. The script fails if the directory does not exist.
• Forgetting to make the script executable: chmod +x backup.sh.

Performance Tips

1. Use parallel compression. pigz is a multi-threaded gzip replacement:

# 4x faster on multi-core systems
tar -I pigz -cf "${BACKUP_FILE}" -C "${SOURCE_DIR}" .

2. Exclude unnecessary files. Skip caches, logs, and temp files:

tar -czf "${BACKUP_FILE}" \
    --exclude="*.log" \
    --exclude="*.tmp" \
    --exclude="cache/*" \
    --exclude="node_modules/*" \
    -C "${SOURCE_DIR}" .

3. Use ionice for low-priority I/O. Don’t let backups starve production:

ionice -c 3 tar -czf "${BACKUP_FILE}" -C "${SOURCE_DIR}" .