PostgreSQL Cloud Backup

Automated PostgreSQL backups to any S3-compatible cloud storage.

Lightweight Docker container that runs pg_dump on a cron schedule, compresses with gzip, uploads to S3-compatible storage, and enforces a retention policy.

Supported Storage Providers

Any S3-compatible storage works out of the box:

Provider	Endpoint Example
Cloudflare R2	https://<account-id>.r2.cloudflarestorage.com
MinIO	https://minio.example.com:9000
AWS S3	https://s3.amazonaws.com
Backblaze B2	https://s3.us-west-000.backblazeb2.com
DigitalOcean Spaces	https://nyc3.digitaloceanspaces.com
Wasabi	https://s3.wasabisys.com

Quick Start

docker run -d --name pg-backup \
  -e POSTGRES_HOST=your-db-host \
  -e POSTGRES_USER=postgres \
  -e POSTGRES_PASSWORD=your-password \
  -e POSTGRES_DB=your-database \
  -e BACKUP_SCHEDULE=daily \
  -e BACKUP_TIME=03:00 \
  -e BACKUP_RETENTION=3 \
  -e S3_ENDPOINT=https://your-endpoint.com \
  -e S3_ACCESS_KEY=your-access-key \
  -e S3_SECRET_KEY=your-secret-key \
  -e S3_BUCKET=your-bucket \
  ghcr.io/dublyo/postgresql-cloudbackup:latest

Docker Compose

services:
  postgres:
    image: postgres:17
    environment:
      POSTGRES_PASSWORD: secret123
      POSTGRES_DB: myapp
    volumes:
      - pgdata:/var/lib/postgresql/data

  pg-backup:
    image: ghcr.io/dublyo/postgresql-cloudbackup:latest
    depends_on:
      - postgres
    environment:
      POSTGRES_HOST: postgres
      POSTGRES_USER: postgres
      POSTGRES_PASSWORD: secret123
      POSTGRES_DB: myapp
      BACKUP_SCHEDULE: daily
      BACKUP_TIME: "03:00"
      BACKUP_RETENTION: "3"
      BACKUP_ON_START: "true"
      S3_ENDPOINT: https://your-endpoint.com
      S3_ACCESS_KEY: your-access-key
      S3_SECRET_KEY: your-secret-key
      S3_BUCKET: my-backups
      S3_PATH_PREFIX: "myapp/"

volumes:
  pgdata:

Environment Variables

PostgreSQL Connection

Variable	Required	Default	Description
POSTGRES_HOST	Yes	—	PostgreSQL hostname or container name
POSTGRES_PORT	No	5432	PostgreSQL port
POSTGRES_USER	Yes	—	Database user
POSTGRES_PASSWORD	Yes	—	Database password
POSTGRES_DB	Yes	—	Database name to back up

Backup Schedule

Variable	Required	Default	Description
BACKUP_SCHEDULE	No	daily	daily, weekly, every6h, every12h, or a cron expression
BACKUP_TIME	No	03:00	Time to run (HH:MM, UTC)
BACKUP_RETENTION	No	3	Number of backups to keep (1-5+)
BACKUP_ON_START	No	false	Run backup immediately on container start
TZ	No	UTC	Timezone for schedule

S3 Storage

Variable	Required	Default	Description
S3_ENDPOINT	Yes	—	S3-compatible endpoint URL
S3_ACCESS_KEY	Yes	—	Access key ID
S3_SECRET_KEY	Yes	—	Secret access key
S3_BUCKET	Yes	—	Bucket name
S3_REGION	No	auto	Region (auto for R2, us-east-1 for MinIO)
S3_PATH_PREFIX	No	backups/	Folder prefix inside bucket

How It Works

1. Cron fires at the configured schedule
2. pg_dump exports the database in plain SQL format
3. gzip -9 compresses the dump (typically 80-90% reduction)
4. aws s3 cp uploads to your S3-compatible bucket
5. Retention lists backups and deletes the oldest beyond the retention count

Backup File Naming

backups/mydb-20260213-030000.sql.gz
backups/mydb-20260214-030000.sql.gz
backups/mydb-20260215-030000.sql.gz

Restoring a Backup

Download and restore manually:

# Download
aws s3 cp s3://your-bucket/backups/mydb-20260213-030000.sql.gz ./backup.sql.gz \
  --endpoint-url https://your-endpoint.com

# Restore
gunzip -c backup.sql.gz | psql -h your-host -U postgres -d your-database

Dublyo PaaS

This template is available on Dublyo as a one-click deploy. Deploy it alongside your PostgreSQL database and configure storage via the dashboard.

License

MIT