Skip to main content

Backups

How to backup and restore your Agam Space data.

What to Backup​

You need to backup two things:

  1. PostgreSQL database - User accounts, file metadata, encryption keys
  2. File storage - Encrypted file chunks

Both are required to restore your instance. Encrypted files are useless without the database.

Automated Backup Script​

Create Backup Script​

Create /opt/agam-space/backup.sh:

#!/bin/bash
BACKUP_DIR="/backups/agam-$(date +%Y%m%d)"
mkdir -p "$BACKUP_DIR"

# Backup database
docker exec agam-space-postgres-1 pg_dump -U postgres agam_space > "$BACKUP_DIR/database.sql"

# Backup files
cp -r /opt/agam-space/data/files "$BACKUP_DIR/"

# Compress
tar -czf "$BACKUP_DIR.tar.gz" -C /backups "$(basename $BACKUP_DIR)"
rm -rf "$BACKUP_DIR"

# Keep last 30 days
find /backups -name "agam-*.tar.gz" -mtime +30 -delete

echo "Backup completed: $BACKUP_DIR.tar.gz"

Make executable:

chmod +x /opt/agam-space/backup.sh

Test Backup​

Run manually to test:

/opt/agam-space/backup.sh

Check backup created:

ls -lh /backups/

Schedule Automatic Backups​

Add to crontab:

crontab -e

Add line:

0 2 * * * /opt/agam-space/backup.sh >> /var/log/agam-backup.log 2>&1

This runs daily at 2 AM and logs output.

Restore from Backup​

Full Restore​

Stop containers:

cd /opt/agam-space
docker-compose down

Extract backup:

cd /backups
tar -xzf agam-20241221.tar.gz

Restore database:

cd /opt/agam-space
docker-compose up -d postgres
sleep 5  # Wait for postgres to start
cat /backups/agam-20241221/database.sql | docker exec -i agam-space-postgres-1 psql -U postgres agam_space

Restore files:

rm -rf /opt/agam-space/data/files
cp -r /backups/agam-20241221/files /opt/agam-space/data/
chown -R 1000:1000 /opt/agam-space/data/files

Start all containers:

docker-compose up -d

Verify:

docker-compose logs -f

Database-Only Restore​

If you only need to restore database (files intact):

cat backup/database.sql | docker exec -i agam-space-postgres-1 psql -U postgres agam_space
docker-compose restart agam

Off-Site Backups​

Option 1: rsync to Remote Server​

Add to backup script after compression:

# Sync to remote server
rsync -avz --delete /backups/ user@backup-server:/backups/agam-space/

Option 2: Cloud Storage (S3, B2, etc.)​

Install rclone:

curl https://rclone.org/install.sh | sudo bash

Configure remote:

rclone config

Add to backup script:

# Upload to cloud
rclone copy /backups/ remote:agam-backups/

Option 3: Encrypted Cloud Backup​

Use restic for encrypted backups:

sudo apt install restic

Initialize repository:

restic -r /backups/restic-repo init

Add to backup script:

# Backup with restic
restic -r /backups/restic-repo backup /opt/agam-space/data

Backup Verification​

Test your backups regularly:

#!/bin/bash
# test-restore.sh

# Create test directory
TEST_DIR="/tmp/agam-restore-test"
mkdir -p "$TEST_DIR"

# Extract latest backup
LATEST_BACKUP=$(ls -t /backups/agam-*.tar.gz | head -1)
tar -xzf "$LATEST_BACKUP" -C "$TEST_DIR"

# Verify database backup
if [ -f "$TEST_DIR/*/database.sql" ]; then
    echo "✅ Database backup found"
else
    echo "❌ Database backup missing"
    exit 1
fi

# Verify files backup
if [ -d "$TEST_DIR/*/files" ]; then
    echo "✅ Files backup found"
else
    echo "❌ Files backup missing"
    exit 1
fi

# Cleanup
rm -rf "$TEST_DIR"

echo "✅ Backup verification passed"

Run monthly:

chmod +x /opt/agam-space/test-restore.sh
# Add to crontab: 0 3 1 * * /opt/agam-space/test-restore.sh

Backup Best Practices​

Retention policy:

  • Keep daily backups for 7 days
  • Keep weekly backups for 4 weeks
  • Keep monthly backups for 12 months

Storage:

  • Keep backups on different physical disk
  • Off-site backup to protect against site failure
  • Encrypt backups if storing in untrusted location

Testing:

  • Test restore process quarterly
  • Document restore procedures
  • Time how long restore takes

Monitoring:

  • Alert if backup fails
  • Alert if backup size changes dramatically
  • Track backup timing trends

Disaster Recovery Plan​

Document your recovery process:

  1. Preparation: Have backup location, credentials ready
  2. New server: Provision new server, install Docker
  3. Restore: Follow restore steps above
  4. DNS: Update DNS to point to new server
  5. Verify: Test login, file access, encryption
  6. Monitor: Watch for issues in first 24 hours

Recovery Time Objective (RTO): How long can you be down? Recovery Point Objective (RPO): How much data loss is acceptable?

For most self-hosters: Daily backups (24h RPO) with 4-hour RTO is reasonable.

Next Steps​

🚀 First Steps - Initial configuration

🔧 Configuration - Configure SSO, trusted devices, and more