Évora

Évora is a research framework for service composition and orchestration in SDN-enabled environments.

Docker Environment

For research reproducibility, Évora provides a Dockerized environment for its core dependencies (ActiveMQ, Axis2, CXF).

Automated Startup

chmod +x run-evora.sh
./run-evora.sh

Manual Orchestration

docker-compose up -d

The services will be available at:

• ActiveMQ (AMQP): localhost:61616
• Axis2 (Legacy): localhost:8080
• CXF (Modern): localhost:8081

Messaging4Transport Integration

The framework is configured to use the Messaging4Transport middleware for robust MD-SAL to AMQP bindings.

Local Installation of Messaging4Transport

Since Messaging4Transport is a separate project, you must install it to your local Maven repository (~/.m2/repository) before building Évora.

Automated Installation

chmod +x install_m4t.sh
./install_m4t.sh

Manual Installation

1. Clone the repository:

   git clone https://github.com/KathiraveluLab/messaging4transport.git

2. Install to local Maven repository:

   cd messaging4transport
   mvn clean install -DskipTests

Note

If you do not have Maven installed, you can install it on Ubuntu/Linux using: sudo apt install maven

Project Structure

• src/main/java/org/evora/core: Building block, composition, and Orchestration logic.
• src/main/java/org/evora/registry: Service Registry implementation.
• evora_topology.py: 12-node Mininet edge topology script.
• pom.xml: Configured with ODL repositories and messaging4transport-impl dependency.

Empirical Evaluation

To reproduce the performance results (Speedup and Complexity) from the ETT 2018 paper:

1. Build the project: mvn clean compile
2. Run the Benchmarker:

   java -cp target/classes org.evora.core.Benchmarker

SDN Deployment (Mininet)

To spin up the 12-node edge topology in a Mininet environment managed by OpenDaylight:

sudo mn --custom evora_topology.py --topo evora --controller remote,ip=<odl_ip>

Citing Évora

If you use Évora in your research, please cite the following papers:

• Kathiravelu, P., Van Roy, P. and Veiga, L., 2018. Composing network service chains at the edge: A Resilient and adaptive software‐defined approach. Transactions on Emerging Telecommunications Technologies, 29(11), p.e3489.

• Kathiravelu, P., Grbac, T.G. and Veiga, L., 2016, June. Building blocks of Mayan: Componentizing the escience workflows through software-defined service composition. In 2016 IEEE International Conference on Web Services (ICWS) (pp. 372-379). IEEE.