In this section, we publish a springcloud/spring-cloud-contract Docker image that contains a project that generates tests and runs them in EXPLICIT mode against a running application.

The EXPLICIT mode means that the tests generated from contracts send real requests and not mocked ones.

We also publish a spring-cloud/spring-cloud-contract-stub-runner Docker image that starts the standalone version of Stub Runner.

1. A Short Introduction to Maven, JARs, and Binary Storage

Since non-JVM projects can use the Docker image, it is good to explain the basic terms behind Spring Cloud Contract packaging defaults.

Parts of the following definitions were taken from the Maven Glossary:

  • Project: Maven thinks in terms of projects. Projects are all you build. Those projects follow a well defined “Project Object Model”. Projects can depend on other projects — in that case, the latter are called “dependencies”. A project may consistent of several subprojects. However, these subprojects are still treated equally as projects.

  • Artifact: An artifact is something that is either produced or used by a project. Examples of artifacts produced by Maven for a project include JAR files and source and binary distributions. Each artifact is uniquely identified by a group ID and an artifact ID that is unique within a group.

  • JAR: JAR stands for Java ARchive. Its format is based on the ZIP file format. Spring Cloud Contract packages the contracts and generated stubs in a JAR file.

  • GroupId: A group ID is a universally unique identifier for a project. While this is often just the project name (for example, commons-collections), it is helpful to use a fully-qualified package name to distinguish it from other projects with a similar name (for example, org.apache.maven). Typically, when published to the Artifact Manager, the GroupId gets slash separated and forms part of the URL. For example, for a group ID of com.example and an artifact ID of application, the result would be /com/example/application/.

  • Classifier: The Maven dependency notation looks as follows: groupId:artifactId:version:classifier. The classifier is an additional suffix passed to the dependency — for example, stubs or sources. The same dependency (for example, com.example:application) can produce multiple artifacts that differ from each other with the classifier.

  • Artifact manager: When you generate binaries, sources, or packages, you would like them to be available for others to download, reference, or reuse. In the case of the JVM world, those artifacts are generally JARs. For Ruby, those artifacts are gems. For Docker, those artifacts are Docker images. You can store those artifacts in a manager. Examples of such managers include Artifactory and Nexus.

2. Generating Tests on the Producer Side

The image searches for contracts under the /contracts folder. The output from running the tests is available in the /spring-cloud-contract/build folder (useful for debugging purposes).

You can mount your contracts and pass the environment variables. The image then:

  • Generates the contract tests

  • Runs the tests against the provided URL

  • Generates the WireMock stubs

  • Publishes the stubs to a Artifact Manager (optional — turned on by default)

2.1. Environment Variables

The Docker image requires some environment variables to point to your running application, to the Artifact manager instance, and so on. The following list describes the environment variables:

Table 1. Docker environment variables





Artifact ID of the project with contracts


Classifier of the project with contracts


Group ID of the project with contracts



Path to contracts for the given project, inside the project with contracts. Defaults to slash-separated EXTERNAL_CONTRACTS_GROUP_ID concatenated with / and EXTERNAL_CONTRACTS_ARTIFACT_ID. For example, for group id ` and artifact ID fish, would result in cat/dog/fish for the contracts path.


(optional) Password if the EXTERNAL_CONTRACTS_REPO_WITH_BINARIES_URL requires authentication. It defaults to `REPO_WITH_BINARIES_PASSWORD, If that is not set, it defaults to `password


URL of your Artifact Manager. It defaults to the value of REPO_WITH_BINARIES_URL environment variable and if that is not set, it defaults to localhost:8081/artifactory/libs-release-local


(optional) Username if the EXTERNAL_CONTRACTS_REPO_WITH_BINARIES_URL requires authentication. It defaults to REPO_WITH_BINARIES_USERNAME. If that is not set, it defaults to `admin


Version of the project with contracts. Defautls to an equivalent of picking the latest



If set to true, retrieves the artifact with contracts from the container’s .m2. Mount your local .m2 as a volume available at the container’s /root/.m2 path



Should the build fail if there are no contracts present?



Type of messaging. Can be either [rabbit] or [kafka].


Archive classifier used for generated producer stubs



Your project’s group ID



Your project’s artifact id



Your project’s version



If set to true, publishes the artifact to binary storage



If set to true, publishes the artifacts to local m2



If set to true will run the task to publish stubs to scm



(optional) Password when the Artifact Manager is secured



URL of your Artifact Manager (defaults to the default URL of Artifactory when running locally)



(optional) Username when the Artifact Manager is secured



For standalone version, which additional protocol should be added

The following environment variables are used when tests are run:

Table 2. Docker environment variables - read at runtime





URL at which the application is running.


Optional password to access the application.


Optional username to access the application.


Timeout to connect to the application to trigger a message.



Timeout to read the response from the application to trigger a message.



Defines the messaging type when dealing with message based contracts.


Type of messaging. Can be either [rabbit] or [kafka].


For Kafka - brokers addresses.


For RabbitMQ - brokers addresses.

2.2. Example of Usage via HTTP

In this section, we explore a simple MVC application. To get started, clone the following git repository and cd to the resulting directory, by running the following commands:

$ git clone
$ cd bookstore

The contracts are available in the /contracts folder.

Since we want to run tests, we can run the following command:

$ npm test

However, for learning purposes, we split it into pieces, as follows:

# Stop docker infra (nodejs, artifactory)
$ ./
# Start docker infra (nodejs, artifactory)
$ ./

# Kill & Run app
$ pkill -f "node app"
$ nohup node app &

# Prepare environment variables
$ APP_IP=""
$ APP_PORT="3000"
$ ARTIFACTORY_URL="http://${APP_IP}:${ARTIFACTORY_PORT}/artifactory/libs-release-local"
$ CURRENT_DIR="$( pwd )"

# Run contract tests
$ docker run  --rm -e "APPLICATION_BASE_URL=${APPLICATION_BASE_URL}" -e "PUBLISH_ARTIFACTS=true" -e "PROJECT_NAME=${CURRENT_FOLDER_NAME}" -e "REPO_WITH_BINARIES_URL=${ARTIFACTORY_URL}" -e "PROJECT_VERSION=${PROJECT_VERSION}" -v "${CURRENT_DIR}/contracts/:/contracts:ro" -v "${CURRENT_DIR}/node_modules/spring-cloud-contract/output:/spring-cloud-contract-output/" springcloud/spring-cloud-contract:"${SC_CONTRACT_DOCKER_VERSION}"

# Kill app
$ pkill -f "node app"

Through bash scripts, the following happens:

  • The infrastructure (MongoDb and Artifactory) is set up. In a real-life scenario, you would run the NodeJS application with a mocked database. In this example, we want to show how we can benefit from Spring Cloud Contract in very little time.

  • Due to those constraints, the contracts also represent the stateful situation.

    • The first request is a POST that causes data to get inserted into the database.

    • The second request is a GET that returns a list of data with 1 previously inserted element.

  • The NodeJS application is started (on port 3000).

  • The contract tests are generated through Docker, and tests are run against the running application.

    • The contracts are taken from /contracts folder.

    • The output of the test is available under node_modules/spring-cloud-contract/output.

  • The stubs are uploaded to Artifactory. You can find them in localhost:8081/artifactory/libs-release-local/com/example/bookstore/0.0.1.RELEASE/. The stubs are at localhost:8081/artifactory/libs-release-local/com/example/bookstore/0.0.1.RELEASE/bookstore-0.0.1.RELEASE-stubs.jar.

2.3. Example of Usage via Messaging

If you want to use Spring Cloud Contract with messaging via the Docker images (e.g. in case of polyglot applications) then you’ll have to have the following prerequisites met:

  • Middleware (e.g. RabbitMQ or Kafka) must be running before generating tests

  • Your contract needs to call a method triggerMessage(…​) with a String parameter that is equal to the contract’s label.

  • Your application needs to have a HTTP endpoint via which we can trigger a message

    • That endpoint should not be available on production (could be enabled via an environment variable)

2.3.1. Example of a Messaging Contract

The contract needs to call a triggerMessage(…​) method. That method is already provided in the base class for all tests in the docker image and will send out a request to the HTTP endpoint on the producer side. Below you can find examples of such contracts.


Contract.make {
    description 'Send a pong message in response to a ping message'
    label 'ping_pong'
    input {
        // You have to provide the `triggerMessage` method with the `label`
        // as a String parameter of the method
    outputMessage {
            message: 'pong'
           outputMessage: [
               connectToBroker: [
                   declareQueueWithName: "queue"
                messageProperties: [
                    receivedRoutingKey: '#'
description: 'Send a pong message in response to a ping message'
label: 'ping_pong'
    # You have to provide the `triggerMessage` method with the `label`
    # as a String parameter of the method
    triggeredBy: 'triggerMessage("ping_pong")'
    sentTo: 'output'
        message: 'pong'
                declareQueueWithName: "queue"
                receivedRoutingKey: '#'

2.3.2. HTTP Endpoint to Trigger a Message

Why is there need to develop such an endpoint? Spring Cloud Contract would have to generate code in various languages (as it does in Java) to make it possible to trigger production code that sends a message to a broker. If such code is not generated then we need to be able to trigger the message anyways, and the way to do it is to provide an HTTP endpoint that the user will prepare in the language of their choosing.

The endpoint must have the following configuration:

  • URL: /springcloudcontract/{label} where label can be any text

  • Method: POST

  • Basing on the label will generate a message that will be sent to a given destination according to the contract definition

Below you have an example of such an endpoint. If you’re interested in providing an example in your language don’t hesitate to file an issue in the Spring Cloud Contract repository at Github.

#!/usr/bin/env python

from flask import Flask
from flask import jsonify
import pika
import os

app = Flask(__name__)

# Production code that sends a message to RabbitMQ
def send_message(cmd):
    connection = pika.BlockingConnection(pika.ConnectionParameters(host='localhost'))
    channel =
            delivery_mode=2,  # make message persistent
    return " [x] Sent via Rabbit: %s" % cmd

# This should be ran in tests (shouldn't be publicly available)
if 'CONTRACT_TEST' in os.environ:
    @app.route('/springcloudcontract/<label>', methods=['POST'])
    def springcloudcontract(label):
        if label == "ping_pong":
            return send_message('{"message":"pong"}')
            raise ValueError('No such label expected.')

2.3.3. Running Message Tests on the Producer Side

Now, let’s generate tests from contracts to test the producer side. We will run bash code to start the Docker image with attached contracts, however we will also add variables for the messaging code to work. In this case let’s assume that the contracts are being stored in a Git repository.

set -x

CURRENT_DIR="$( pwd )"

export APP_IP="$( ./ )"
export APP_PORT="${APP_PORT:-8000}"
export PROJECT_NAME="${PROJECT_NAME:-application}"
# In our Python app we want to enable the HTTP endpoint
export CONTRACT_TEST="true"
# In the Verifier docker container we want to add support for RabbitMQ
export MESSAGING_TYPE="rabbit"

# Let's start the infrastructure (e.g. via Docker Compose)
yes | docker-compose kill || echo "Nothing running"
docker-compose up -d

echo "SC Contract Version [${SC_CONTRACT_DOCKER_VERSION}]"
echo "Application URL [${APPLICATION_BASE_URL}]"
echo "Project Version [${PROJECT_VERSION}]"

# Let's run python app
gunicorn -w 4 --bind main:app &

# Generate and run tests
docker run  --rm \
                --name verifier \
                # For the image to find the RabbitMQ running in another container
                -e "SPRING_RABBITMQ_ADDRESSES=${APP_IP}:5672" \
                # We need to tell the container what messaging middleware we will use
                -e "MESSAGING_TYPE=${MESSAGING_TYPE}" \
                -e "PUBLISH_STUBS_TO_SCM=false" \
                -e "PUBLISH_ARTIFACTS=false" \
                -e "PROJECT_NAME=${PROJECT_NAME}" \
                -e "PROJECT_GROUP=${PROJECT_GROUP}" \
                -e "PROJECT_VERSION=${PROJECT_VERSION}" \
                -v "${CURRENT_DIR}/build/spring-cloud-contract/output:/spring-cloud-contract-output/" \

kill $APP_PID

yes | docker-compose kill

What will happen is:

  • Tests will be generated from contracts taken from Git

  • In the contract we’ve provided an entry in metadata called declareQueueWithName that will lead to creation of a queue in RabbitMQ with the given name before the request to trigger the message is sent

  • Via the triggerMessage("ping_pong") method call a POST request to the Python application to the /springcloudcontract/ping_pong endpoint will be made

  • The Python application will generate and send a '{"message":"pong"}' JSON via RabbitMQ to an exchange called output

  • The generated test will poll for a message sent to the output exchange

  • Once the message was received will assert its contents

After the tests have passed we know that the message was properly sent from the Python app to RabbitMQ.

3. Running Stubs on the Consumer Side

This section describes how to use Docker on the consumer side to fetch and run stubs.

We publish a spring-cloud/spring-cloud-contract-stub-runner Docker image that starts the standalone version of Stub Runner.

3.1. Environment Variables

You can run the docker image and pass any of the common properties for JUnit and Spring as environment variables. The convention is that all the letters should be upper case. The dot (.) should be replaced with underscore (_) characters. For example, the stubrunner.repositoryRoot property should be represented as a STUBRUNNER_REPOSITORY_ROOT environment variable.

In addition to those variables you can set the following ones:

  • MESSAGING_TYPE - what type of messaging system are you using (currently supported are rabbit, kafka)

  • ADDITIONAL_OPTS - any additional properties that you would like to pass to the application

3.2. Example of Usage

We want to use the stubs created in this [docker-server-side] step. Assume that we want to run the stubs on port 9876. You can see the NodeJS code by cloning the repository and changing to the directory indicated in the following commands:

$ git clone
$ cd bookstore

Now we can run the Stub Runner Boot application with the stubs, by running the following commands:

# Provide the Spring Cloud Contract Docker version
# The IP at which the app is running and Docker container can reach it
$ APP_IP=""
# Spring Cloud Contract Stub Runner properties
# Stub coordinates 'groupId:artifactId:version:classifier:port'
$ STUBRUNNER_IDS="com.example:bookstore:0.0.1.RELEASE:stubs:9876"
$ STUBRUNNER_REPOSITORY_ROOT="http://${APP_IP}:8081/artifactory/libs-release-local"
# Run the docker with Stub Runner Boot
$ docker run  --rm \
    -p "9876:9876" \

When the preceding commands run,

  • A standalone Stub Runner application gets started.

  • It downloads the stub with coordinates com.example:bookstore:0.0.1.RELEASE:stubs on port 9876.

  • It gets downloads from Artifactory running at

  • After a while, Stub Runner is running on port 8083.

  • The stubs are running at port 9876.

On the server side, we built a stateful stub. We can use curl to assert that the stubs are setup properly. To do so, run the following commands:

# let's run the first request (no response is returned)
$ curl -H "Content-Type:application/json" -X POST --data '{ "title" : "Title", "genre" : "Genre", "description" : "Description", "author" : "Author", "publisher" : "Publisher", "pages" : 100, "image_url" : "", "buy_url" : "" }' http://localhost:9876/api/books
# Now time for the second request
$ curl -X GET http://localhost:9876/api/books
# You will receive contents of the JSON
If you want use the stubs that you have built locally, on your host, you should set the -e STUBRUNNER_STUBS_MODE=LOCAL environment variable and mount the volume of your local m2 (-v "${HOME}/.m2/:/root/.m2:ro").

3.3. Example of Usage with Messaging

In order to make messaging work it’s enough to pass the MESSAGING_TYPE environment variable with kafka or rabbit values. This will lead to setting up the Stub Runner Boot Docker image with dependencies required to connect to the broker.

In order to set the connection properties you can check out Spring Cloud Stream properties page to set proper environment variables.

The most common property you would set is the location of the running middlewara. If a property to set it is called spring.rabbitmq.addresses or spring.kafka.bootstrap-servers then you should name the environment variable SPRING_RABBITMQ_ADDRESSES and SPRING_KAFKA_BOOTSTRAP_SERVERS respectively.

4. Running Contract Tests against Existing Middleware

There is legitimate reason to run your contract tests against existing middleware. Some testing frameworks might give you false positive results - the test within your build passes whereas on production the communication fails.

In Spring Cloud Contract docker images we give an option to connect to existing middleware. As presented in previous subsections we do support Kafka and RabbitMQ out of the box. However, via Apache Camel Components we can support other middleware too. Let’s take a look at the following examples of usage.

4.1. Spring Cloud Contract Docker and running Middleware

In order to connect to arbitrary middleware, we’ll leverage the standalone metadata entry in the contract section.

description: 'Send a pong message in response to a ping message'
label: 'standalone_ping_pong' (1)
  triggeredBy: 'triggerMessage("ping_pong")' (2)
  sentTo: 'rabbitmq:output' (3)
  body: (4)
    message: 'pong'
  standalone: (5)
    setup: (6)
      options: rabbitmq:output?queue=output&routingKey=(7)
    outputMessage: (8)
      additionalOptions: routingKey=#&queue=output (9)
1 Label by which we’ll be able to trigger the message via Stub Runner
2 As in the previous messaging examples we’ll need to trigger the HTTP endpoint in the running application to make it send a message according to the provided protocol
3 protocol:destination as requested by Apache Camel
4 Output message body
5 Standalone metadata entry
6 Setup part will contain information about how to prepare for running contract tests before the actual call to HTTP endpoint of the running application is made
7 Apache Camel URI to be called in the setup phase. In this case we will try to poll for a message at the output exchange and due to to having the queue=output and routingKey= a queue with name output will be set and bound to the output exchange with routing key
8 Additional options (more technical ones) to be appended to the protocol:destination from point (3) - together will be combined in the following format rabbitmq:output?routingKey=#&queue=output.

For the contract tests to pass we will need as usual in case of messaging in polyglot environment a running application and running middleware. This time we will have different environment variables set for the Spring Cloud Contract Docker image.

set -x

# Setup
# Run the middleware
docker-compose up -d rabbitmq (1)

# Run the python application
gunicorn -w 4 --bind main:app & (2)

docker run  --rm \
                --name verifier \
                -e "STANDALONE_PROTOCOL=rabbitmq" \ (3)
                -e "CAMEL_COMPONENT_RABBITMQ_ADDRESSES=" \ (4)
                -e "PUBLISH_STUBS_TO_SCM=false" \
                -e "PUBLISH_ARTIFACTS=false" \
                -e "APPLICATION_BASE_URL=" \
                -e "PROJECT_NAME=application" \
                -e "PROJECT_GROUP=group" \
                -e "EXTERNAL_CONTRACTS_ARTIFACT_ID=application" \
                -e "EXTERNAL_CONTRACTS_GROUP_ID=group" \
                -e "EXTERNAL_CONTRACTS_VERSION=0.0.1-SNAPSHOT" \
                -v "${CURRENT_DIR}/build/spring-cloud-contract/output:/spring-cloud-contract-output/" \

# Teardown
kill $APP_PID
yes | docker-compose kill
1 We need to have the middleware running first
2 The application needs to be up and running
3 Via the STANDALONE_PROTOCOL environment variable we will fetch a Apache Camel Component. The artifact that we will fetch is org.apache.camel.springboot:camel-${STANDALONE_PROTOCOL}-starter. In other words STANDALONE_PROTOCOL is matching Camel’s component.
4 We’re setting addresses (we could be setting credentials) via Camel’s Spring Boot Starter mechanisms. Example for Apache Camel’s RabbitMQ Spring Boot Auto-Configuration

4.2. Stub Runner Docker and running Middleware

In order to trigger a stub message against running middleware, we can run Stub Runner Docker image in the following manner.

Example of usage

$ docker run \
    -e "STUBRUNNER_IDS=group:application:0.0.1-SNAPSHOT" \ (2)
    -e "STUBRUNNER_REPOSITORY_ROOT=git://" \ (3)
    -e ADDITIONAL_OPTS="" \ (4)
    -v "${HOME}/.m2/:/root/.m2:ro" \ (6)
    -p 8750:8750 \ (7)
    springcloud/spring-cloud-contract-stub-runner:3.0.0-SNAPSHOT (8)
1 We’re injecting the address of RabbitMQ via Apache Camel’s Spring Boot Auto-Configuration
2 We’re telling Stub Runner which stubs to download
3 We’re providing an external location for our stubs (Git repository)
4 Via the property we’re telling Stub Runner which additional dependency to fetch at runtime. In this case we want to fetch camel-rabbitmq-starter so XXX is a random string and we want to fetch org.apache.camel.springboot:camel-rabbitmq-starter artifact in version 3.4.0.
5 Since we’re using Git, the remote option of fetching stubs needs to be set
6 So that we speed up launching of Stub Runner, we’re attaching our local Maven repository .m2 as a volume. If you don’t have it populated you can consider setting the write permissions via :rw instead read only :ro.
7 We expose the port 8750 at which Stub Runner is running.
8 Coordinates of the Stub Runner Docker image.

After a while you’ll notice the following text in your console, which means that Stub Runner is ready to accept requests.

o.a.c.impl.engine.AbstractCamelContext   : Apache Camel 3.4.3 (camel-1) started in 0.007 seconds
o.s.c.c.s.server.StubRunnerBoot          : Started StubRunnerBoot in 14.483 seconds (JVM running for 18.666)
o.a.c.c.C.[Tomcat].[localhost].[/]       : Initializing Spring DispatcherServlet 'dispatcherServlet'
o.s.web.servlet.DispatcherServlet        : Initializing Servlet 'dispatcherServlet'
o.s.web.servlet.DispatcherServlet        : Completed initialization in 2 ms

To get the list of triggers you can send an HTTP GET request to localhost:8750/triggers endpoint. To trigger a stub message, you can send a HTTP POST request to localhost:8750/triggers/standalone_ping_pong. In the console you’ll see:

o.s.c.c.v.m.camel.CamelStubMessages      : Will send a message to URI [rabbitmq:output?routingKey=#&queue=output]

If you check the RabbitMQ management console, you’ll see that there’s 1 message available in the output queue.