Graph Modeling 10.2 Installation Guide

26/03/2026

This guide shows how to do a fresh installation of Graph Modeling (formerly PoolParty) for both development and production environments via Docker compose.

Tip

While this guide focuses on Docker compose, Helm charts are also available for deploying Graph Modeling 10.2 and higher on a Kubernetes cluster. You can find the official chart here.

Note

If you already have PoolParty 9.7 installed, consult the PoolParty 2025 R2 (9.7) to PoolParty 10 Migration Guide instead of this one. If you are installing a clean new version of Graph Modeling 10.2 instead of upgrading but using existing Keycloak, Elasticsearch, or Spark instances, the Migration Guide has helpful information for that.

The most up-to-date version of these instructions is available in the github compose-files project's README.md file, titled Docker compose file for Graph Modeling (formerly PoolParty). This document will help you navigate the README to find what you need.

Graph Modeling uses Docker to automate the installation of the configured set of components that make up the selected bundle. These components include Keycloak, Elasticsearch, Spark, Nginx, and GraphDB.

You can perform this installation from the Linux command line or, using the Docker Desktop, on macOS or Windows.

Important

Please note that due to rebranding of our software suite the names PoolParty and Graph Modeling are alternatively used in this guide for version 10.2 and higher. The image files continue having PoolParty reference in their designations.

A minimum of 32GB of memory (8 for Graph Modeling, 12 for Elasticsearch, and 12 for GraphDB); we recommend 44 GB (16 for PoolParty, 16 for Elasticsearch, and 12 for GraphDB)
A downloaded copy of the compose-files github project, which contains the README file mentioned above and other files that you will use as part of the installation
A recent version of Docker or Docker Desktop
The docker-compose-plugin, which is installed with the other Docker packages when you install Docker with the instructions on the Docker site
License files for Graph Modeling 10.2 and GraphDB

Caution

Please note that the internal database Keycloak is distributed with is only intended for non-productive environments. All configuration steps assume that you have already switched to an external PostgreSQL DB to be used by Keycloak to avoid any issues during deployment.

From a high level, installation of Graph Modeling 10.2 consists of three steps:

Copying the .env_template file included with the github project to a .env file that you will use to customize your installation
Setting appropriate environment variables in the .env file, and possibly other values in other configuration files
Running a Docker command that uses the downloaded Docker image and configuration settings to install and then start Graph Modeling

Configuration file settings vary depending on your deployment requirements. For development environments intended for internal use only, you can use the default settings with included Docker instances.

Production environments, however, typically require integration with existing external services such as PostgreSQL, Keycloak, and Spark rather than relying on the bundled Docker instances.

The configuration section of the README file describes how you must copy the .env_template file included with the github project to a .env file. This file is where you will set environment variables for Docker to use when it installs the image. Two especially important variables described by the README file are POOLPARTY_LICENSE and GRAPHDB_LICENSE , which must be set for your Graph Modeling instance to find the appropriate licenses.

The README file also lists several other administration passwords that are useful to set. Comments in the .env file describe dozens of other environment variables that you can set to customize your developer or production environment.

The following shows an example .env_template file with all of the potential settings and comments about their roles. (Be sure to use the version of this file included with the compose-files github project mentioned above under Prerequisites to ensure that you have the most recent version.)

The introduction to the Docker compose file for Graph Modeling README file describes additional configuration files with settings that you can use to, for example, add additional services or include SSL support. (For more details on running Graph Modeling with SSL support, see the Running with SLL section of the README file, which describes the .env variables to set to point to your SSL certificate and key and also the ssl.yaml file that is included with the github project to use with the docker compose command.)

The introduction to the README file also describes the Docker commands used to start up or shut down services. For a simple development environment Graph Modeling setup, the startup command may be as simple as this:

docker compose up -d

As this runs, it will display messages as it starts the various services. When complete, it will return to your command line prompt.

Note

When running the installation, the error message "Bind for 0.0.0.0:nn failed: port is already allocated" (where nn is a port number), means that the Docker installation process is trying to assign something to a port that is already used by another service. Use the appropriate operating system tool to find which service is using that port. If you can't unset that one, you can change the port being used by Graph Modeling in the docker-compose.yaml file.

You can add additional configuration files on the docker compose command line. For example, the following uses the production.yaml file included with the github compose-files project to add in the PostgreSQL service used by Keycloak:

docker compose -f docker-compose.yaml -f production.yaml up -d

Note

To learn more about running a development configuration of Graph Modeling without Keycloak, see the Development section of the README file.

You can configure the use of an external Spark service by by adding the spark.yaml file from the github project to the command line with appropriate settings:

docker compose up -f docker-compose.yaml -f spark.yaml up -d

To deploy the services that complement Graph Modeling 10.2, you can use the addons.yaml file with the -f parameter in that command line. This file includes the Application Development Framework (ADF), the Semantic Workbench services, GraphViews, UnifiedViews, and the Mirror App.

Warning

Your Graph Modeling license must have the Semantic Workbench capability in order to use this service. Before starting the services, copy the nginx configuration files from files/nginx/addons to files/nginx/includes/extra_includes. These will expose the service through nginx on the /ADF, /SemanticWorkbench, /GraphViews, /UnifiedViews , and /PoolParty/mirror context paths.

Once the docker compose up command finishes, you can confirm that you have installed Graph Modeling by sending your browser to http://poolparty.127.0.0.1.nip.io/PoolParty, which will display this:

Note

nip.io is used here because it is an easy DNS service to use for demos like this. For anything else other then local development or testing, you would need a real DNS name that resolves to the machine where PoolParty is deployed.

On the screen shown above, log in with the user name superadmin and the password poolparty. Once you log in, it will have you create a new password.

By default Graph Modeling uses fixed Elasticsearch index names (such as conceptdata, searchdata). This functionality can be enabled after having successfully migrated to Graph Modeling 10.2.

To run multiple Graph Modeling instances against a single Elasticsearch cluster, each instance must be assigned unique index names using a prefix, a suffix, or both.

Tip

For more details please refer to the README files in the docker image.

In this section: