Microservices flow visual

Modern apps often split into many small services, then put one front door in front of them to route requests and hide the internal sprawl. (martinfowler.com) That front door is usually an application programming interface gateway, which sends a mobile app or browser request to the right backend service and can also handle tasks like authentication. Service discovery is the address book behind it: Netflix’s Eureka says it is used for discovery, load balancing, and failover of mid-tier servers. (github.com) (docs.spring.io) Many teams then connect services with event streams or message queues instead of direct calls. Amazon Web Services says event-driven architecture is common in microservices because services can react to state changes without being tightly bound to each other. (aws.amazon.com) (docs.aws.amazon.com) A common rule is “database per service,” meaning each service owns its own data store instead of sharing one central database. Chris Richardson’s microservices patterns site says that design creates the need for sagas, which coordinate a business process across multiple local transactions. (microservices.io) A saga is a step-by-step workflow for work that spans services, such as creating an order, reserving credit, and confirming payment. Microsoft’s Azure Architecture Center says the pattern is used to maintain data consistency in distributed systems without relying on a single distributed transaction. (learn.microsoft.com) That tradeoff comes with a cost: more network hops, more waiting, and more ways for partial failures to surface. Martin Fowler’s microservices guide describes independently deployable services communicating over lightweight mechanisms, but each boundary also adds operational complexity. (martinfowler.com) The architecture is often associated with companies operating at very large scale, but the underlying parts are not exotic. Netflix open-sourced Eureka, and Amazon Web Services has published multiple event-driven reference architectures for teams moving away from tightly coupled systems. (github.com) (aws.amazon.com) What the diagram captures, in plain terms, is a resilience playbook: one entry point, a live directory of services, asynchronous messaging, and local data ownership. What it leaves implicit is the bill for that flexibility—extra latency, harder debugging, and workflows that must be repaired with saga logic when one step fails. (microservices.io) (learn.microsoft.com)