Model Graph Tools Setup

This repository contains documentation, configuration and scripts on how to get started with the model graph tools. The model graph tools are a set of tools and services to analyze and work with the WildFly management model:

The analyzer creates a Neo4j database containing the management model of a given WildFly version. For ech major WildFly version starting with version 10.0.0.Final there are predefined Neo4j database images available at https://quay.io/repository/modelgraphtools/neo4j.

The setup described here use these images. If you want to use another WildFly version or want to analyze a version of JBoss EAP, use the analyzer command line tool to create your own Neo4j database.

The model services use these Neo4j databases and expose a REST API. The API service keeps a registry of all running model services and provides a unified access to the browser - a SPA with an UI to browse, query and compare the WildFly management model.

Getting all services up and running requires some plumbing. There are scripts in the different directories to make this as easy as possible.

Development

To run the model graph tools with two model services in development mode use a combination of the following scripts:

git clone https://github.com/model-graph-tools/api.git
git clone https://github.com/model-graph-tools/model.git
git clone https://github.com/model-graph-tools/browser.git

api/start-redis.sh
api/dev.sh
model/start-modeldb.sh 25
model/dev.sh 25
model/start-modeldb.sh 24
model/dev.sh 24
browser/dev.sh

The last command will open a browser at http://localhost:3000.

Docker Compose

To run the model graph tools using docker compose, use the jbang script compose/compose.java. The script accepts a port for the browser (defaults to 80), and expects a list of WildFly versions. The output is a docker compose file, which starts all services using the proper configuration values:

git clone https://github.com/model-graph-tools/setup.git

cd setup/compose
jbang compose.java 25 24 23
docker compose up -d
docker logs -fn 100 mgt-mgt-api-1

Use docker logs -fn 100 mgt-mgt-api-1 to follow the log file of the API service and wait until you see log messages about the registration of the model services. Then open http://localhost.

To shut everything down, use docker compose down.

OpenShift

To run the model graph tools on OpenShift, log in to your cluster and create a project. Then use the jbang script openshift/openshift.java. The script expects a list of WildFly versions. The output is a bash script, which creates and starts all necessary OpenShift resources:

git clone https://github.com/model-graph-tools/setup.git

cd setup/openshift
jbang openshift.java 25 24 23
./oc-setup.sh
oc logs -f deployment/mgt-api

Use oc logs -f deployment/mgt-api to follow the log file of the API service and wait until you see log messages about the registration of the model services. Then open the URL returned from oc get -o template route/mgt-browser --template={{.spec.host}}.

To shut everything down and remove all created OpenShift resources, use oc-delete.sh.