backup-strategy（备份策略）

Backup Strategy Skill

Implement automated backup solutions for VPS servers to ensure quick recovery from security incidents or system failures.

What This Skill Does

This skill helps AI agents configure automated backup systems on VPS servers. Security isn't just prevention - it's recovery. If your server gets compromised, you need to rebuild quickly. Regular, off-server backups are essential for business continuity and disaster recovery.

Key capabilities:

• Create automated backup scripts
• Schedule regular backups with cron
• Implement retention policies (keep N days of backups)
• Compress and encrypt backup archives
• Store backups off-server (S3, remote server, etc.)
• Verify backup integrity
• Document restoration procedures

When to Use

Use this skill when you need to:

• Set up new server with backup strategy
• Implement disaster recovery plan
• Comply with data retention requirements
• Protect against ransomware and data loss
• Enable quick server rebuilds
• Meet business continuity requirements

Critical understanding: The backup must NOT be on the same server. If the server is compromised, local backups can be deleted or encrypted by attackers.

Prerequisites

• Root or sudo access to the server
• Sufficient disk space for temporary backups
• Off-server storage solution (S3, remote server, NAS, etc.)
• Understanding of what needs to be backed up
• Credentials for remote storage (if applicable)

What to Back Up

Critical Directories

/home                    # User home directories
/etc                     # System and application configuration
/var/www                 # Web server content
/var/lib/mysql           # MySQL databases (if using file-based)
/root                    # Root user home (if used)
/opt                     # Optional software installations
/usr/local               # Locally installed software

What NOT to Back Up

/tmp                     # Temporary files
/var/tmp                 # Temporary files
/proc                    # Virtual filesystem
/sys                     # Virtual filesystem
/dev                     # Device files
/run                     # Runtime data
/var/cache               # Cache files

Basic Backup Script

Simple Tar-Based Backup

Create /usr/local/bin/backup.sh:

#!/bin/bash
#
# Simple backup script using tar and gzip
#

# Configuration
BACKUP_DIR="/backup"
DATE=$(date +%Y-%m-%d)
BACKUP_NAME="backup-$DATE.tar.gz"
RETENTION_DAYS=7

# Create backup directory
mkdir -p "$BACKUP_DIR"

# Create compressed archive
echo "Creating backup: $BACKUP_NAME"
tar -czf "$BACKUP_DIR/$BACKUP_NAME" \
    --exclude='/backup' \
    --exclude='/proc' \
    --exclude='/sys' \
    --exclude='/dev' \
    --exclude='/run' \
    --exclude='/tmp' \
    --exclude='/var/tmp' \
    --exclude='/var/cache' \
    /home \
    /etc \
    /var/www \
    /root \
    2>/var/log/backup-error.log

# Check if backup was successful
if [ $? -eq 0 ]; then
    echo "Backup completed successfully"
    echo "Backup saved to: $BACKUP_DIR/$BACKUP_NAME"
else
    echo "Backup failed! Check /var/log/backup-error.log"
    exit 1
fi

# Delete old backups (keep last N days)
echo "Cleaning up old backups (keeping last $RETENTION_DAYS days)..."
find "$BACKUP_DIR" -name "backup-*.tar.gz" -mtime +$RETENTION_DAYS -delete

echo "Backup process complete"

Make it executable:

sudo chmod +x /usr/local/bin/backup.sh

Advanced Backup Strategies

Database Backups

MySQL/MariaDB:

#!/bin/bash
# MySQL backup script

DB_USER="root"
DB_PASS="your_password"
BACKUP_DIR="/backup/mysql"
DATE=$(date +%Y-%m-%d)

mkdir -p "$BACKUP_DIR"

# Backup all databases
mysqldump -u"$DB_USER" -p"$DB_PASS" --all-databases \
    --single-transaction \
    --quick \
    --lock-tables=false \
    > "$BACKUP_DIR/all-databases-$DATE.sql"

# Compress
gzip "$BACKUP_DIR/all-databases-$DATE.sql"

# Delete old backups
find "$BACKUP_DIR" -name "all-databases-*.sql.gz" -mtime +7 -delete

PostgreSQL:

#!/bin/bash
# PostgreSQL backup script

BACKUP_DIR="/backup/postgresql"
DATE=$(date +%Y-%m-%d)

mkdir -p "$BACKUP_DIR"

# Backup all databases
sudo -u postgres pg_dumpall > "$BACKUP_DIR/pg-backup-$DATE.sql"

# Compress
gzip "$BACKUP_DIR/pg-backup-$DATE.sql"

# Delete old backups
find "$BACKUP_DIR" -name "pg-backup-*.sql.gz" -mtime +7 -delete

Incremental Backups with rsync

#!/bin/bash
# Incremental backup using rsync

BACKUP_DIR="/backup/incremental"
CURRENT="$BACKUP_DIR/current"
DATE=$(date +%Y-%m-%d-%H%M%S)
SNAPSHOT="$BACKUP_DIR/$DATE"

# Create backup directory
mkdir -p "$BACKUP_DIR"

# Perform incremental backup
rsync -av --delete \
    --link-dest="$CURRENT" \
    --exclude='/backup' \
    --exclude='/proc' \
    --exclude='/sys' \
    /home \
    /etc \
    /var/www \
    "$SNAPSHOT"

# Update current symlink
rm -f "$CURRENT"
ln -s "$SNAPSHOT" "$CURRENT"

# Keep only last 10 snapshots
ls -1dt "$BACKUP_DIR"/2* | tail -n +11 | xargs rm -rf

Off-Server Storage

AWS S3 Backup

#!/bin/bash
# Backup to AWS S3

BACKUP_DIR="/backup"
S3_BUCKET="s3://my-backups/server-name"
DATE=$(date +%Y-%m-%d)
BACKUP_FILE="backup-$DATE.tar.gz"

# Create backup
tar -czf "$BACKUP_DIR/$BACKUP_FILE" /home /etc /var/www

# Upload to S3
aws s3 cp "$BACKUP_DIR/$BACKUP_FILE" "$S3_BUCKET/"

# Verify upload
if [ $? -eq 0 ]; then
    echo "Backup uploaded to S3 successfully"
    # Remove local copy after successful upload
    rm "$BACKUP_DIR/$BACKUP_FILE"
else
    echo "S3 upload failed!"
    exit 1
fi

# S3 lifecycle policy handles retention

SCP to Remote Server

#!/bin/bash
# Backup to remote server via SCP

BACKUP_DIR="/backup"
REMOTE_USER="backup"
REMOTE_HOST="backup-server.example.com"
REMOTE_DIR="/backups/webserver"
DATE=$(date +%Y-%m-%d)
BACKUP_FILE="backup-$DATE.tar.gz"

# Create backup
tar -czf "$BACKUP_DIR/$BACKUP_FILE" /home /etc /var/www

# Upload via SCP (requires SSH key authentication)
scp "$BACKUP_DIR/$BACKUP_FILE" "$REMOTE_USER@$REMOTE_HOST:$REMOTE_DIR/"

# Verify upload
if [ $? -eq 0 ]; then
    echo "Backup transferred successfully"
    rm "$BACKUP_DIR/$BACKUP_FILE"
else
    echo "Transfer failed!"
    exit 1
fi

Encrypted Backups

#!/bin/bash
# Create encrypted backup

BACKUP_DIR="/backup"
DATE=$(date +%Y-%m-%d)
BACKUP_FILE="backup-$DATE.tar.gz"
ENCRYPTED_FILE="backup-$DATE.tar.gz.gpg"
GPG_RECIPIENT="admin@example.com"

# Create compressed backup
tar -czf "$BACKUP_DIR/$BACKUP_FILE" /home /etc /var/www

# Encrypt with GPG
gpg --encrypt --recipient "$GPG_RECIPIENT" \
    --output "$BACKUP_DIR/$ENCRYPTED_FILE" \
    "$BACKUP_DIR/$BACKUP_FILE"

# Remove unencrypted version
rm "$BACKUP_DIR/$BACKUP_FILE"

# Upload encrypted backup (S3, SCP, etc.)
# ...

echo "Encrypted backup created: $ENCRYPTED_FILE"

Scheduling Backups with Cron

Edit Crontab

sudo crontab -e

Common Schedules

# Daily at 2 AM
0 2 * * * /usr/local/bin/backup.sh >> /var/log/backup.log 2>&1

# Weekly on Sunday at 3 AM
0 3 * * 0 /usr/local/bin/backup.sh

# Daily at 2 AM, keep 30 days
0 2 * * * /usr/local/bin/backup.sh && find /backup -name "backup-*.tar.gz" -mtime +30 -delete

# Every 6 hours
0 */6 * * * /usr/local/bin/backup.sh

# Monthly on the 1st at midnight
0 0 1 * * /usr/local/bin/backup.sh

Cron with Logging

# Daily backup with logging and email on failure
0 2 * * * /usr/local/bin/backup.sh > /var/log/backup-$(date +\%Y\%m\%d).log 2>&1 || mail -s "Backup Failed" admin@example.com < /var/log/backup-$(date +\%Y\%m\%d).log

Backup Verification

Check Backup Integrity

#!/bin/bash
# Verify backup archive integrity

BACKUP_FILE="/backup/backup-2024-01-31.tar.gz"

# Test gzip integrity
gzip -t "$BACKUP_FILE"

if [ $? -eq 0 ]; then
    echo "Backup archive is valid"
else
    echo "Backup archive is corrupted!"
    exit 1
fi

# Test tar contents
tar -tzf "$BACKUP_FILE" > /dev/null

if [ $? -eq 0 ]; then
    echo "Tar archive structure is valid"
else
    echo "Tar archive has errors!"
    exit 1
fi

List Backup Contents

# List files in backup
tar -tzf /backup/backup-2024-01-31.tar.gz | less

# Search for specific file
tar -tzf /backup/backup-2024-01-31.tar.gz | grep "config.php"

Restoration Procedures

Full System Restore

#!/bin/bash
# Restore from backup

BACKUP_FILE="/backup/backup-2024-01-31.tar.gz"

# WARNING: This will overwrite existing files!
echo "WARNING: This will restore files and may overwrite existing data!"
read -p "Continue? (yes/no): " CONFIRM

if [ "$CONFIRM" != "yes" ]; then
    echo "Aborted"
    exit 1
fi

# Extract to root
cd /
tar -xzf "$BACKUP_FILE"

echo "Restore complete. Review extracted files and restart services."

Restore Specific Directory

# Restore only /etc
tar -xzf /backup/backup-2024-01-31.tar.gz -C / etc/

# Restore specific file
tar -xzf /backup/backup-2024-01-31.tar.gz -C / etc/nginx/nginx.conf

Restore Database

# MySQL restore
gunzip < /backup/mysql/all-databases-2024-01-31.sql.gz | mysql -uroot -p

# PostgreSQL restore
gunzip < /backup/postgresql/pg-backup-2024-01-31.sql.gz | sudo -u postgres psql

Monitoring and Alerting

Email Notifications

#!/bin/bash
# Backup with email notification

BACKUP_SCRIPT="/usr/local/bin/backup.sh"
ADMIN_EMAIL="admin@example.com"

# Run backup
if $BACKUP_SCRIPT; then
    echo "Backup completed successfully on $(date)" | \
        mail -s "Backup Success - $(hostname)" "$ADMIN_EMAIL"
else
    echo "Backup failed on $(date)" | \
        mail -s "BACKUP FAILED - $(hostname)" "$ADMIN_EMAIL"
fi

Check Last Backup Age

#!/bin/bash
# Alert if backup is too old

BACKUP_DIR="/backup"
MAX_AGE_HOURS=26  # Alert if no backup in last 26 hours

LATEST_BACKUP=$(find "$BACKUP_DIR" -name "backup-*.tar.gz" -type f -printf '%T@ %p\n' | sort -n | tail -1 | cut -d' ' -f2-)

if [ -z "$LATEST_BACKUP" ]; then
    echo "No backups found!" | mail -s "BACKUP ALERT" admin@example.com
    exit 1
fi

AGE_SECONDS=$(($(date +%s) - $(stat -c %Y "$LATEST_BACKUP")))
AGE_HOURS=$((AGE_SECONDS / 3600))

if [ $AGE_HOURS -gt $MAX_AGE_HOURS ]; then
    echo "Last backup is $AGE_HOURS hours old!" | \
        mail -s "BACKUP TOO OLD" admin@example.com
fi

Security Best Practices

1. Off-server storage - Never rely solely on local backups
2. Encryption - Encrypt sensitive backups, especially if storing remotely
3. Access control - Restrict backup file permissions (600 or 640)
4. Test restores - Regularly test that backups can be restored
5. Monitor backup jobs - Alert on failures
6. Retention policy - Balance storage costs with recovery needs
7. Version backups - Keep multiple generations
8. Document procedures - Maintain restoration runbooks
9. Separate credentials - Don't store backup credentials on the server being backed up

Common Mistakes to Avoid

• ❌ Only backing up to the same server (single point of failure)
• ❌ Not testing restore procedures
• ❌ Backing up cached/temporary files (waste of space)
• ❌ Not encrypting backups containing sensitive data
• ❌ Setting retention too short (can't recover from old issues)
• ❌ Not monitoring backup success/failure
• ❌ Including backup directory in backup (infinite loop!)
• ❌ Not documenting what's backed up and how to restore

Additional Resources

See references/backup-locations.md for storage provider comparison.

See scripts/backup-full.sh for comprehensive backup script.

See scripts/backup-mysql.sh for database-specific backup.

Related Skills

• auto-updates - Keep backup tools updated
• ssh-hardening - Secure SSH for remote backups
• firewall-configuration - Protect backup storage access