Skip to main content

Backup & Recovery Guide

Complete guide for backing up and recovering the SBM CRM Platform.

Backup Strategy​

Backup Types​

  1. Database Backups - Daily full backups, hourly incremental
  2. File Backups - Daily backups of uploaded files
  3. Configuration Backups - Weekly backups of configuration files
  4. Log Backups - Monthly archives of application logs

Backup Schedule​

  • Database: Daily at 2:00 AM
  • Files: Daily at 3:00 AM
  • Configuration: Weekly on Sunday at 1:00 AM
  • Retention: 30 days for daily, 90 days for weekly

Database Backups​

Automated Backup Script​

Create /usr/local/bin/backup-sbmcrm-db.sh:

#!/bin/bash

# Configuration
BACKUP_DIR="/backups/sbmcrm/database"
DATE=$(date +%Y%m%d_%H%M%S)
DB_NAME="sbmcrm_production"
DB_USER="sbmcrm"
DB_HOST="localhost"
RETENTION_DAYS=30

# Create backup directory
mkdir -p $BACKUP_DIR

# Perform backup
pg_dump -h $DB_HOST -U $DB_USER -F c -b -v -f "$BACKUP_DIR/sbmcrm_$DATE.dump" $DB_NAME

# Compress backup
gzip "$BACKUP_DIR/sbmcrm_$DATE.dump"

# Remove old backups
find $BACKUP_DIR -name "sbmcrm_*.dump.gz" -mtime +$RETENTION_DAYS -delete

# Upload to S3 (optional)
aws s3 cp "$BACKUP_DIR/sbmcrm_$DATE.dump.gz" s3://your-backup-bucket/database/

echo "Backup completed: sbmcrm_$DATE.dump.gz"

Set Up Cron Job​

# Edit crontab
crontab -e

# Add daily backup at 2 AM
0 2 * * * /usr/local/bin/backup-sbmcrm-db.sh >> /var/log/sbmcrm/backup.log 2>&1

Manual Backup​

# Full backup
pg_dump -h localhost -U sbmcrm -F c -b -v -f backup.dump sbmcrm_production

# Compressed backup
pg_dump -h localhost -U sbmcrm sbmcrm_production | gzip > backup.sql.gz

# Backup specific table
pg_dump -h localhost -U sbmcrm -t customers sbmcrm_production > customers_backup.sql

File Backups​

Backup Uploaded Files​

#!/bin/bash

BACKUP_DIR="/backups/sbmcrm/files"
DATE=$(date +%Y%m%d_%H%M%S)
UPLOAD_DIR="/var/www/sbmcrm/uploads"
RETENTION_DAYS=30

mkdir -p $BACKUP_DIR

# Create tar archive
tar -czf "$BACKUP_DIR/uploads_$DATE.tar.gz" -C $UPLOAD_DIR .

# Remove old backups
find $BACKUP_DIR -name "uploads_*.tar.gz" -mtime +$RETENTION_DAYS -delete

# Upload to S3
aws s3 cp "$BACKUP_DIR/uploads_$DATE.tar.gz" s3://your-backup-bucket/files/

Configuration Backups​

Backup Configuration Files​

#!/bin/bash

BACKUP_DIR="/backups/sbmcrm/config"
DATE=$(date +%Y%m%d)
CONFIG_DIR="/opt/sbmcrm/config"

mkdir -p $BACKUP_DIR

# Backup configuration
tar -czf "$BACKUP_DIR/config_$DATE.tar.gz" -C $CONFIG_DIR .

# Keep last 90 days
find $BACKUP_DIR -name "config_*.tar.gz" -mtime +90 -delete

Recovery Procedures​

Database Recovery​

Full Database Restore​

# Stop application
sudo systemctl stop sbmcrm-api

# Drop existing database (WARNING: This deletes all data)
dropdb -h localhost -U sbmcrm sbmcrm_production

# Create new database
createdb -h localhost -U sbmcrm sbmcrm_production

# Restore from backup
pg_restore -h localhost -U sbmcrm -d sbmcrm_production backup.dump

# Or from compressed SQL
gunzip < backup.sql.gz | psql -h localhost -U sbmcrm sbmcrm_production

# Start application
sudo systemctl start sbmcrm-api

Point-in-Time Recovery​

For point-in-time recovery, use WAL archiving:

# Enable WAL archiving in postgresql.conf
wal_level = replica
archive_mode = on
archive_command = 'cp %p /backups/wal/%f'

# Restore to specific time
pg_basebackup -D /var/lib/postgresql/restore -Ft -z -P
# Then apply WAL files up to target time

File Recovery​

# Extract files from backup
tar -xzf uploads_20240120.tar.gz -C /var/www/sbmcrm/uploads

# Restore specific file
tar -xzf uploads_20240120.tar.gz path/to/file.jpg -C /var/www/sbmcrm/uploads

Backup Verification​

Verify Backup Integrity​

# Test database backup
pg_restore --list backup.dump

# Verify file backup
tar -tzf uploads_20240120.tar.gz | head -10

Automated Verification​

#!/bin/bash

BACKUP_FILE="$1"

# Verify database backup
if pg_restore --list "$BACKUP_FILE" > /dev/null 2>&1; then
    echo "Database backup is valid"
else
    echo "Database backup is corrupted!"
    exit 1
fi

Disaster Recovery Plan​

Recovery Time Objectives (RTO)​

  • Critical Systems: 4 hours
  • Non-Critical Systems: 24 hours

Recovery Point Objectives (RPO)​

  • Database: 1 hour (hourly backups)
  • Files: 24 hours (daily backups)

Recovery Steps​

  1. Assess Damage

    • Identify what needs to be recovered
    • Determine recovery point

  2. Prepare Environment

    • Set up new servers if needed
    • Restore base system

  3. Restore Backups

    • Restore database
    • Restore files
    • Restore configuration

  4. Verify System

    • Test application functionality
    • Verify data integrity
    • Check service health

  5. Resume Operations

    • Start services
    • Monitor for issues
    • Notify stakeholders

Backup Storage​

Local Storage​

  • Fast access for quick recovery
  • Limited retention
  • Vulnerable to local disasters

Cloud Storage (S3, Azure Blob)​

  • Off-site protection
  • Scalable storage
  • Cost-effective for long retention

Hybrid Approach​

  • Recent backups: Local storage
  • Older backups: Cloud storage
  • Best of both worlds

Monitoring​

Backup Monitoring​

Set up alerts for:

  • Backup failures
  • Backup size anomalies
  • Backup duration issues
  • Storage space warnings

Health Checks​

# Check last backup time
ls -lt /backups/sbmcrm/database/ | head -2

# Check backup size
du -sh /backups/sbmcrm/database/

# Verify backup age
find /backups/sbmcrm/database/ -mtime +1 -name "*.dump.gz"

Best Practices​

  1. Test Restores Regularly - Monthly restore tests
  2. Multiple Backup Locations - Local + cloud
  3. Encrypt Backups - Protect sensitive data
  4. Document Procedures - Clear recovery steps
  5. Monitor Backups - Automated monitoring
  6. Version Control - Track backup versions
  7. Retention Policy - Clear retention rules