Skip to main content

Viglet Dumont: Installation Guide

Viglet Dumont is an open-source data extraction platform that connects content sources to search engines. The connector application (dumont-connector.jar) provides the pipeline infrastructure (message queue, batch processing, strategies, and indexing plugins) while connector plugins (AEM, Web Crawler) provide the actual content extraction capabilities.

The connector JAR alone does not crawl content

dumont-connector.jar is the pipeline engine. To actually extract content from sources, you must add a connector plugin JAR (AEM or Web Crawler) to the classpath via -Dloader.path. Without a plugin, the connector has no data source to crawl.

How It Works​

dumont-connector.jar          ← Pipeline engine (queue, strategies, indexing)
  └── libs/                    ← Plugin directory (added via -Dloader.path)
       ├── aem-plugin.jar      ← AEM connector plugin
       ├── web-crawler-plugin.jar  ← Web Crawler connector plugin
       └── my-custom-plugin.jar    ← Your custom extensions (optional)

The connector JAR is built with Spring Boot's ZIP layout (PropertiesLauncher), which allows loading additional JARs from external directories at runtime. When you pass -Dloader.path=libs, Spring Boot scans the libs/ directory and adds all JARs to the application classpath, making the connector plugins available to the pipeline.

Installing Java​

Dumont requires Java.

note

Dumont only supports Java 21 or later.

Set the JAVA_HOME environment variable and add $JAVA_HOME/bin to your PATH.

Hardware Requirements​

MinimumRecommended
Single 1.5 GHz processor, 1 GB available memoryDual 2 GHz processors, 2 GB available memory

In addition:

  • 500 MB of disk space for software
  • Additional disk space proportional to the content being crawled (Web Crawler caches visited URLs)

Installing Dumont​

Option 1: Docker (fastest)​

Pull the pre-built image published to the GitHub Container Registry (ghcr.io) by the Publish Docker Image GitHub Action:

docker pull ghcr.io/openviglet/dumont-ce:latest
docker run -p 30130:30130 ghcr.io/openviglet/dumont-ce:latest
Image tags

Besides latest, every run also publishes the project version and a commit tag (sha-<short>). Pin a specific tag in production instead of tracking latest.

Option 2: Download JARs​

Download dumont-connector.jar and the connector plugin JARs from the releases page:

mkdir -p /appl/viglet/dumont/server/libs

# Main connector engine
cp dumont-connector.jar /appl/viglet/dumont/server/

# Connector plugins (choose one or both)
cp aem-plugin.jar /appl/viglet/dumont/server/libs/
cp web-crawler-plugin.jar /appl/viglet/dumont/server/libs/

Option 3: Build from source​

git clone https://github.com/openviglet/dumont.git
cd dumont
mvn clean install

# Copy the connector engine
cp connector/connector-app/target/dumont-connector.jar /appl/viglet/dumont/server/

# Copy connector plugins
mkdir -p /appl/viglet/dumont/server/libs
cp aem/aem-plugin/target/aem-plugin.jar /appl/viglet/dumont/server/libs/
cp web-crawler/wc-plugin/target/web-crawler-plugin.jar /appl/viglet/dumont/server/libs/

Starting with a Connector Plugin​

The key to running Dumont is the -Dloader.path JVM property. It tells Spring Boot's PropertiesLauncher where to find the connector plugin JARs.

With AEM Plugin​

java -Dloader.path=libs \
     -Dturing.url=http://localhost:2700 \
     -Dturing.apiKey=<YOUR_TURING_API_KEY> \
     -jar dumont-connector.jar

With Web Crawler Plugin​

java -Dloader.path=libs \
     -Dturing.url=http://localhost:2700 \
     -Dturing.apiKey=<YOUR_TURING_API_KEY> \
     -jar dumont-connector.jar
One connector plugin per JVM

Currently, only one connector plugin can be active per JVM instance. To run both AEM and Web Crawler, start two separate instances of dumont-connector.jar, each with its own libs/ directory containing the appropriate plugin, and each on a different port.

With Custom AEM Extensions​

If you have a custom AEM extension (like the aem-plugin-sample), add its JAR to the same libs/ directory:

ls libs/
aem-plugin.jar
aem-plugin-sample.jar    # Your custom extensions

java -Dloader.path=libs \
     -Dturing.url=http://localhost:2700 \
     -Dturing.apiKey=<YOUR_TURING_API_KEY> \
     -jar dumont-connector.jar

The configuration JSON references extension classes by fully-qualified name (e.g., com.viglet.dumont.connector.aem.sample.ext.DumAemExtSampleModelJson). These classes are found on the classpath because the extension JAR is in the libs/ directory.

loader.path accepts directories or individual JARs

You can point to a directory or a specific JAR: -Dloader.path=libs or -Dloader.path=/path/to/aem-plugin.jar

Database Configuration​

By default, Dumont uses H2 as its embedded database for tracking indexing state. For production, replace with PostgreSQL.

H2 (default: development only)​

No configuration needed. The database is created automatically at ./store/db/dumontDB.

Create a database and user:

CREATE USER dumont PASSWORD '<password>' WITH CREATEUSER;
CREATE DATABASE dumont_db;
ALTER DATABASE dumont_db OWNER TO dumont;

Configure via JVM properties:

java -Dloader.path=libs \
     -Dspring.datasource.url=jdbc:postgresql://localhost:5432/dumont_db \
     -Dspring.datasource.username=dumont \
     -Dspring.datasource.password=<password> \
     -Dspring.datasource.driver-class-name=org.postgresql.Driver \
     -Dturing.url=http://localhost:2700 \
     -Dturing.apiKey=<YOUR_TURING_API_KEY> \
     -jar dumont-connector.jar

Connecting to Viglet Turing​

Dumont sends content to Turing by default. Configure the connection:

turing:
  url: http://localhost:2700
  apiKey: <YOUR_TURING_API_KEY>

Create an API Token in Turing → Administration → API Tokens and use it as the apiKey value.

API Token Required

Dumont cannot send content to Turing without a valid API Token. Create one in the Turing admin console before starting Dumont.

Creating a Linux Service​

Each connector plugin runs as a separate service. Create a dedicated directory per connector with its own configuration.

Directory Layout (AEM example)​

/appl/viglet/dumont/aem/
├── dumont-connector.jar
├── dumont-connector.properties    # Spring Boot external config
└── libs/
    └── aem-plugin.jar             # Connector plugin

External Properties File​

Create /appl/viglet/dumont/aem/dumont-connector.properties:

turing.url=http://localhost:2700
turing.apiKey=<YOUR_TURING_API_KEY>
server.port=30130

Systemd Service File​

As root, create /etc/systemd/system/dumont-aem.service:

[Unit]
Description=Viglet Dumont (AEM)
After=syslog.target

[Service]
User=turing
Group=turing
WorkingDirectory=/appl/viglet/dumont/aem
ExecStart=/appl/java/jdk21/bin/java -Xmx512m -Xms512m \
  -Dloader.path=/appl/viglet/dumont/aem/libs/ \
  -jar /appl/viglet/dumont/aem/dumont-connector.jar \
  --spring.config.additional-location=file:/appl/viglet/dumont/aem/dumont-connector.properties
SuccessExitStatus=143

[Install]
WantedBy=multi-user.target

Enable and start the service:

systemctl daemon-reload
systemctl enable dumont-aem.service
systemctl start dumont-aem.service

Running Multiple Connectors​

Since only one connector plugin is supported per JVM, run separate services for each connector, each with its own directory, config, and port:

/appl/viglet/dumont/
├── aem/                           # AEM connector (port 30130)
│   ├── dumont-connector.jar
│   ├── dumont-connector.properties
│   └── libs/
│       └── aem-plugin.jar
└── webcrawler/                    # Web Crawler connector (port 30131)
    ├── dumont-connector.jar
    ├── dumont-connector.properties
    └── libs/
        └── web-crawler-plugin.jar

Create a separate systemd service for each (e.g., dumont-aem.service, dumont-webcrawler.service), each pointing to its own WorkingDirectory and properties file with a different server.port.

Standalone Connectors (Database & FileSystem)​

The Database and FileSystem connectors are standalone CLI tools: they do not run as plugins inside the connector application. Instead, they connect to a running Dumont instance via its REST API.

Database Connector​

java -cp dumont-db-indexer.jar com.viglet.dumont.connector.db.DumDbImportTool \
  --server http://localhost:30130 \
  --api-key <API_KEY> \
  --driver org.mariadb.jdbc.Driver \
  --connect "jdbc:mariadb://localhost:3306/products" \
  --query "SELECT id, name, description FROM products" \
  --site ProductCatalog \
  --locale en_US

FileSystem Connector​

java -cp dumont-filesystem-indexer.jar com.viglet.dumont.filesystem.DumFSImportTool \
  --source-dir /mnt/shared/documents \
  --server http://localhost:30130 \
  --api-key <API_KEY> \
  --site InternalDocs \
  --locale en_US

These tools run independently and can be scheduled via cron jobs or CI/CD pipelines.

Docker Compose​

A full stack with Dumont, Turing, Solr, and MariaDB:

services:
  dumont:
    image: ghcr.io/openviglet/dumont-ce:latest
    ports:
      - "30130:30130"
    environment:
      TURING_URL: http://turing:2700
      TURING_APIKEY: <YOUR_API_KEY>

  turing:
    image: ghcr.io/openviglet/turing-ce:latest
    ports:
      - "2700:2700"

  solr:
    image: solr:9
    ports:
      - "8983:8983"

  mariadb:
    image: mariadb:11
    environment:
      MARIADB_ROOT_PASSWORD: root
      MARIADB_DATABASE: turing
      MARIADB_USER: turing
      MARIADB_PASSWORD: turing

Verifying the Installation​

After starting Dumont, verify it is running:

curl http://localhost:30130/api/v2/connector/status

A successful response confirms the service is up and ready to accept connector configurations.

Directory Structure Summary​

A complete production deployment with two connectors:

/appl/viglet/dumont/
├── aem/                                # AEM connector instance
│   ├── dumont-connector.jar            # Pipeline engine
│   ├── dumont-connector.properties     # Turing URL, API key, port
│   ├── libs/
│   │   └── aem-plugin.jar             # AEM connector plugin
│   └── store/                          # Auto-created at runtime
│       ├── db/                         # H2 indexing database
│       ├── queue/                      # Artemis persistent queue
│       └── logs/                       # Application logs
└── webcrawler/                         # Web Crawler connector instance
    ├── dumont-connector.jar
    ├── dumont-connector.properties
    ├── libs/
    │   └── web-crawler-plugin.jar
    └── store/
PageDescription
Configuration ReferenceAll application.yaml properties
ArchitectureSystem components and data flow
Connectors OverviewAvailable connectors and how to configure them
ON THIS PAGE