Open-source test data management - store, share and reserve test data
539
Open-source test data management: one place where teams store, share and reserve test data, so parallel tests never fight over the same records. A REST API with client libraries for Java, Python and JavaScript, a web application for testers, full history of changes, and an MCP server for AI agents.
latest - the latest release3.20.1, 3.20... - a release, or the latest patch of a minor versionImages for linux/amd64 and linux/arm64. The image holds the jar of the GitHub release, run with Java 17
(Amazon Corretto) as user 1000, on port 8080.
Databucket needs PostgreSQL (developed and tested on 18). With Docker Compose:
services:
postgres:
image: postgres:18-alpine
environment:
POSTGRES_DB: databucket
POSTGRES_USER: databucket
POSTGRES_PASSWORD: change-me
volumes:
- pgdata:/var/lib/postgresql
healthcheck:
test: ["CMD-SHELL", "pg_isready -U databucket -d databucket"]
interval: 5s
retries: 10
databucket:
image: databucket/databucket-server:latest
depends_on:
postgres:
condition: service_healthy
ports:
- "8080:8080"
environment:
SPRING_DATASOURCE_URL: jdbc:postgresql://postgres:5432/databucket
SPRING_DATASOURCE_USERNAME: databucket
SPRING_DATASOURCE_PASSWORD: change-me
# signs the sign-in tokens: a random text of at least 32 characters, e.g. openssl rand -base64 48
JWT_SECRET: change-me-to-a-random-text-of-at-least-32-characters
volumes:
pgdata:
docker compose up, open http://localhost:8080 and sign in as super / super - change this password right away
(the user menu → Account). Then Management → Projects: create a project and import the DEMO dataset to look around.
JWT_SECRET is required: Databucket doesn't start without its own secret of at least 32 characters (since 3.20.2). Whoever knows it can sign in as any user, so keep it secret and don't reuse the examples.
Every setting is an environment variable (Spring Boot names), e.g.:
| Variable | |
|---|---|
SPRING_DATASOURCE_URL, SPRING_DATASOURCE_USERNAME, SPRING_DATASOURCE_PASSWORD | the PostgreSQL database (empty at the first start) |
JWT_SECRET | signs the sign-in tokens, at least 32 characters |
JWT_EXPIRE_HOURS | how long a sign-in is valid, 24 by default |
SERVER_SERVLET_CONTEXT_PATH | e.g. /databucket to serve under a path |
SPRING_MAIL_HOST, SPRING_MAIL_PORT, SPRING_MAIL_USERNAME, SPRING_MAIL_PASSWORD, SPRING_MAIL_FROM | the SMTP server for e-mails (sign-up, forgotten password) |
DATABUCKET_MCP_ENABLED | true turns on the MCP server for AI agents at /mcp |
SSO_ENABLED, SSO_PROVIDER, SSO_ISSUER_URI, SSO_CLIENT_ID, SSO_CLIENT_SECRET | single sign-on (OpenID Connect) |
All settings: Installation, Single sign-on. A complete local setup with single sign-on (Keycloak) and a mail catcher is the docker compose of the repository.
Pull the new image and start it on the same database; the Upgrading table lists the releases that need a manual step.
License: Apache 2.0.
Content type
Image
Digest
sha256:727c0f475…
Size
280.1 MB
Last updated
about 1 hour ago
docker pull databucket/databucket-server