Gaffer Graph Database Framework

GafferPop is a lightweight implementation of the TinkerPop framework designed for Gaffer. It works by delegating TinkerPop methods to standard Gaffer graph operations.

Note: This implementation is currently in development. It is considered a basic implementation, and its performance relative to standard Gaffer OperationChains is currently unknown.

The Accumulo Store is a component of Gaffer designed to work with Apache Accumulo. For detailed administration and implementation guidance, refer to the official Gaffer documentation.

• Administration Guide: For operational information, see the Accumulo Store page.
• Implementation Details: For technical details on how the store is built, see the Accumulo Store implementation page.

Depending on your workflow, you may find these related projects useful:

• gafferpy: A Python shell for executing Gaffer operations.
• gaffer-docker: Code for running Gaffer using Docker or Kubernetes.
• koryphe: An extensible functions library for filtering, aggregating, and transforming data based on the Java Function API (a dependency of Gaffer).

The Road Traffic example models UK road use data as a Gaffer graph to demonstrate data exploration and summarization. The graph consists of several edge types representing different relationships:

• Region to Location: e.g., South West $\rightarrow$ Bristol
• Location to Roads: e.g., Bristol $\rightarrow$ M32
• Roads and their Junctions: e.g., M32 $\rightarrow$ M32:1
• Junctions and their locations: e.g., M32:2 $\rightarrow$ 361150,175250
• Traffic counts between junctions: e.g., M32:1 $\rightarrow$ M32:2 (representing hourly traffic flow)

Traffic Count Edges (RoadUse): Multiple edges exist between the same two junctions, one for each hour of observation. These edges contain properties including:

• Start of the hour
• End of the hour
• Total vehicle count for the hour
• A map of vehicle type to count for the hour

To build and run the Federated Demo, ensure you have Java 8, Maven, and Git installed. The process involves cloning the Gaffer repository and running a provided start script which handles Maven dependencies like Tomcat.

Once deployed, the REST API will be available at localhost:8080/rest.

# Clone the Gaffer repository (master branch, depth 1)
git clone --depth 1 --branch master https://github.com/gchq/Gaffer.git
cd Gaffer

# Run the start script to download dependencies and start the demo
./example/federated-demo/scripts/start.sh

The road-use directory contains examples of operations and scripts for generating graphs. You can use the provided JSON files to manually test operations via Swagger or curl. To automatically generate three graphs and populate them with data, run the createGraphs.sh script from within the road-use directory.

./scripts/createGraphs.sh

Gaffer is hosted on Maven Central. To use the Java API, you must include the core Gaffer graph module and a specific store module corresponding to your chosen database technology.

For example, to use the Accumulo store, add the following dependencies to your pom.xml:

<dependency>
    <groupId>uk.gov.gchq.gaffer</groupId>
    <artifactId>graph</artifactId>
    <version>${gaffer.version}</version>
</dependency>
<dependency>
    <groupId>uk.gov.gchq.gaffer</groupId>
    <artifactId>accumulo-store</artifactId>
    <version>${gaffer.version}</version>
</dependency>