dbk

Automated backup for databases running in Docker containers, with support for cron scheduling, S3/Google Drive upload, and Discord notifications.

---

How It Works

This container periodically backs up PostgreSQL databases (or others, if adapted), saving dumps locally and/or sending them to remote destinations. The schedule is fully configurable via environment variable.

How to Use

1. Prerequisites

• Docker installed
• The database must be running in a container and on the same Docker network as the backup container

---

2. rclone configuration (Google Drive / S3)

To use Google Drive or S3 as backup destinations, you must provide a valid rclone.conf file with the necessary credentials. Learn how to generate your configuration file for Google Drive or/and S3 in the official rclone documentation. For an example of how your file should look, see the rclone/rclone.example.conf file in this repository.

• For Google Drive: rclone Google Drive documentation
• For S3 (Amazon or compatible): rclone S3 documentation

After generating your rclone.conf: Place the rclone.conf file inside a directory named rclone (for example, ./rclone/rclone.conf). This directory should be in the same location where you run your docker run or docker compose commands. This way, you can mount the entire rclone directory as a volume into the container, as shown in the usage examples below. The configuration file will be used to enable backup to cloud providers.

---

3. Using the published image

Pull the image from Docker Hub:

docker pull f4b1002/dbk:latest

Create your .env file (example):

DB_HOST=my_postgres
DB_USER=user
DB_PASSWORD=secret
DB_NAME=mydb
DB_PORT=5432

# Number of days to keep the dumps (optional, default: 5)
DUMP_RETENTION_DAYS=

S3_PATH=bucket_name/dumps # start with bucket name
GDRIVE_PATH=dumps

CRON_SCHEDULE=0 3 * * *
WEBHOOK_URL=https://discord.com/api/webhooks/...

Custom pg_dump options

You can use the PG_DUMP_EXTRA_OPTS environment variable to pass custom options to pg_dump. For example, to dump only a specific schema, set:

PG_DUMP_EXTRA_OPTS=-n "myschema"

Example directory structure

Before running the container, your project directory should look like this:

project-root/
├── .env
├── pg_dumps/
└── rclone/
    └── rclone.conf

Run the backup container

Important: Replace db-network with your database network name.

docker run -d \
  --name dbk \
  --env-file .env \
  -v $(pwd)/pg_dumps:/pg_dumps \
  -v $(pwd)/rclone:/root/.config/rclone \
  --network db-network \
  f4b1002/dbk:latest

---

3. Building the image locally

Clone the repository:

git clone https://github.com/Fabiopf02/dbk.git
cd dbk

Build the image:

docker compose build

---

Environment Variables

Variable	Description	Required	Example
DB_HOST	Database host	Yes	my_postgres
DB_USER	Database user	Yes	postgres
DB_PASSWORD	Database password	Yes	mypassword
DB_NAME	Database name	Yes	mydb
DB_PORT	Database port	No	5432
CRON_SCHEDULE	Backup schedule (cron)	Yes	0 3 * * *
S3_PATH	S3 path	No	bucket_name/dumps
GDRIVE_PATH	Google Drive path	No	dumps
WEBHOOK_URL	Discord webhook URL (notifications, optional)	No	https://discord.com/api/webhooks/...
DUMP_RETENTION_DAYS	Number of days to keep dumps (optional, default: 5)	No	7 (default: 5)
PG_DUMP_EXTRA_OPTS	Extra options to pass to pg_dump (e.g., -n "myschema")	No	-n "myschema"

---

Tips & Notes

• The backup container must be on the same Docker network as the database.
• Use the database container name as the value for DB_HOST.
• Backups are saved in /pg_dumps inside the container (mount a local volume for persistence).
• You can change the schedule anytime by updating CRON_SCHEDULE and restarting the container.