<aside> 💡 Unit tests with static mocks

</aside>

Mocks are not trustworthy, especially those that developers write themselves. Static mocks that are not updated to account for changing responses could still miss the error in our example because;

• mocks that mock dependencies need to be updated regularly to be in-sync with changing response

<aside> 💡 E2E tests

</aside>

They’re extremely difficult to write, maintain, and update. An E2E test that actually invokes inter-service communication like a real user would catch this issue. But the cost of catching this issue with a test that could involve many services would be very high, given the time and effort spent creating it.

• unreliable, un-targeted and extremely expensive
• needs the entire system up & running, making it slower and difficult to identify the error initiation point