Quality

Quality is velocity. Each change should have automated tests and it's important to help team understand on how they can test their changes

Types of test:

• unit

• mutation

• contract

• integration

• snapshot

• shadow

• end-to-end (e2e)

Comparison

Type	Execution speed	Flakiness	Environment setup cost	Systems included	Measurability
Description	How fast tests can be executed	How stable is the test suite	How expensive is to spin up test environment	How many systems are included in test execution	Ability to measure the coverage
Unit	milliseconds			single
Mutation	seconds			single
Contract	seconds			single
Integration	minutes			many
Snapshot
Shadow				many
E2E	minutes / hours			many

Contract testing

Contract testing focuses on interactions between services. It's used to help to ensure that different systems are compatible with each other.

Consumers are the systems which requests the data from the Provider.

Provider is the system which respond with the data to Consumer.

What is contract testing?

Both sides agree on the format of the requests and the responses. The communication is described in form of a contract. All involved parties must follow the contract to prevent breaking changes.

Sample contract:

{
  "consumer": {
    "name": "OrderingService"
  },
  "provider": {
    "name": "PaymentService"
  },
  "interactions": [
    {
      "description": "A request to process a payment",
      "request": {
        "method": "POST",
        "path": "/api/payments",
        "headers": {
          "Content-Type": "application/json"
        },
        "body": {
          "orderId": "abc123",
          "amount": 100.00,
          "currency": "USD",
          "paymentMethod": "credit_card"
        }
      },
      "response": {
        "status": 200,
        "headers": {
          "Content-Type": "application/json"
        },
        "body": {
          "transactionId": "txn789",
          "status": "success",
          "amount": 100.00,
          "currency": "USD"
        }
      }
    }
  ],
  "metadata": {
    "pactSpecification": {
      "version": "2.0.0"
    }
  }
}

During each change either from Consumer or Provider side there is a verification process which confirms that both parties still adhere to the contract.

Consumer driven contract test

Actual Consumer sends the request to a Mock Provider which verifies the request and returns the responses based on the contract.

Mock Consumer sends the request defined in a contract to Actual Provider and verifies the response against the contract.

The beauty of this process is that it removes a dependency between both of the systems and each one can be tested in a separation.

Contract mismatch

Contract testing is especially useful:

• for systems with microservices architecture

• for system developed by multiple teams

• when you maintain the system with multiple clients

FAQ

Question	Answer
How we measure coverage with contract tests?
How in contract test we can make sure that mock is up to date?
Is mock auto generated in contract test?
What is consumer driven contract test?
What is provider driven contract test?
For each type of the test, how to measure the coverage?