Skip to content

README.md

Latest commit

 

History

History
146 lines (102 loc) · 4.71 KB

README.md

File metadata and controls

146 lines (102 loc) · 4.71 KB
Raw

Deploy HugeGraph Server with Docker

Note:

  1. The HugeGraph Docker image is a convenience release, not an official ASF distribution artifact. See the ASF Release Distribution Policy for details.

  2. Use release tags (for example, 1.7.0) for stable deployments. Use latest only for development or testing.

1. Deploy

Use Docker to quickly start a standalone HugeGraph Server with RocksDB.

  1. Using docker run

    Use docker run -itd --name=graph -p 8080:8080 hugegraph/hugegraph:1.7.0 to start hugegraph server.

  2. Using docker compose

    To deploy only the server, use docker compose up -d. The compose file is as follows:

    version: '3'
services:
  graph:
    image: hugegraph/hugegraph:1.7.0
    ports:
      - 8080:8080

2. Create Sample Graph on Server Startup

To preload sample data on startup, set PRELOAD=true.

To customize the preload, mount your own Groovy script.

  1. Using docker run

    Use docker run -itd --name=graph -p 8080:8080 -e PRELOAD=true -v /path/to/script:/hugegraph-server/scripts/example.groovy hugegraph/hugegraph:1.7.0 to start hugegraph server.

  2. Using docker compose

    Use docker compose up -d to start quickly. The compose file is below. example.groovy is a predefined script. Replace it with your own script to preload different data:

    version: '3'
services:
  graph:
    image: hugegraph/hugegraph:1.7.0
    environment:
      - PRELOAD=true
    volumes:
      - /path/to/script:/hugegraph-server/scripts/example.groovy
    ports:
      - 8080:8080

  3. Using start-hugegraph.sh

    If you deploy HugeGraph Server without Docker, you can also pass -p true to bin/start-hugegraph.sh.

3. Enable Authentication

  1. Using docker run

    Use docker run -itd --name=graph -p 8080:8080 -e AUTH=true -e PASSWORD=xxx hugegraph/hugegraph:1.7.0 to enable authentication.

  2. Using docker compose

    Set the environment variables in the compose file:

    version: '3'
services:
  server:
    image: hugegraph/hugegraph:1.7.0
    container_name: graph
    ports:
      - 8080:8080
    environment:
      - AUTH=true
      - PASSWORD=xxx

4. Run OpenTelemetry

CAUTION:

The docker-compose-trace.yaml uses Grafana and Grafana Tempo, both of which are licensed under AGPL-3.0. Use this template for testing only.

  1. Start the OpenTelemetry collector

    # Run from the repository root
docker compose -f hugegraph-server/hugegraph-dist/docker/example/docker-compose-trace.yaml -p hugegraph-trace up -d

  2. Enable the OpenTelemetry agent

    ./start-hugegraph.sh -y true

  3. Stop the OpenTelemetry collector

    # Run from the repository root
docker compose -f hugegraph-server/hugegraph-dist/docker/example/docker-compose-trace.yaml -p hugegraph-trace stop

  4. References

5. Distributed Cluster (PD + Store + Server)

For a full distributed HugeGraph cluster with PD, Store, and Server, use the 3-node compose file in the docker/ directory at the repository root.

Prerequisites: Allocate at least 12 GB memory to Docker Desktop (Settings → Resources → Memory). The cluster runs 9 JVM processes.

cd docker
HUGEGRAPH_VERSION=1.7.0 docker compose -f docker-compose-3pd-3store-3server.yml up -d

See docker/README.md for the full setup guide, environment variable reference, and troubleshooting.

6. Process Supervision & Health Checks

All four HugeGraph Docker images (PD, Store, Server, Server-hstore) include native HEALTHCHECK instructions. docker ps shows real health status:

Image Health endpoint
hugegraph/hugegraph GET /versions on port 8080
hugegraph/hugegraph-hstore GET /versions on port 8080
hugegraph/hugegraph-pd GET /v1/health on port 8620
hugegraph/hugegraph-store GET /v1/health on port 8520

The entrypoints supervise the Java process directly — when Java exits, the container exits. If started with a restart policy (the provided compose files use restart: unless-stopped), Docker will bring it back automatically. The old cron-based monitor (-m true) is for VM/bare-metal deployments only and is not used in Docker images.