This project demonstrates a microservices-based system that processes user requests, performs calculations, and communicates asynchronously using Kafka, while ensuring observability and reliability using patterns like the Outbox Pattern, and tools like Prometheus, and Grafana
- A user, or a simulated user via K6, initiates a request to Service A through gRPC.
- The request contains two numbers, and the system's job is to add them together quickly and reliably.
-
Listens for incoming gRPC requests.
-
Performs the simple yet essential task of adding the two numbers.
-
But here’s the twist: instead of immediately pushing the result to Kafka, it:
- Safely stores the result in an "outbox" table in its database—this follows the Outbox Pattern, which ensures messages aren’t lost.
- A dedicated background job then picks up the saved results and sends them to Kafka, ensuring resilience and consistency.
- Serves as the message backbone, connecting services seamlessly.
- Queues the result messages so they can be processed asynchronously by Service B, enabling a decoupled and scalable system.
- Listens to the Kafka topic like a loyal subscriber.
- Retrieves each message, adds the result to its running total, and can optionally log or persist each value.
- It keeps track of everything, ensuring nothing is missed.
- Prometheus acts as the metrics timekeeper, storing rich time-series data.
- Grafana gives life to the metrics with beautiful dashboards, showing real-time insights and helping engineers make informed decisions.
- K6 simulates realistic load scenarios, helping test how the system performs under stress.
- This ensures the architecture is ready for anything—from a few users to thousands.
