# Example: Configuring Versioning via Query Parameter
spring.mvc.apiversion.use-query-parameter=version
# Example: Configuring Versioning via Custom HTTP Header
# spring.mvc.apiversion.use-header=X-API-Version
# Example: Configuring Versioning via URI Path Segments
# spring.mvc.apiversion.use-path-segment=true
Circular Dependencies in Spring Boot — and Why @Lazy is NOT
the Real Fix
🔄 What is
a Circular Dependency?
In Spring Boot, a circular dependency occurs when two or more beans depend on
each other for initialization.
⚠️
Why Not @Lazy?
Many developers love solving this quickly by using @Lazy to break the
initialization cycle:
👉 This
works, but it’s just a band-aid solution.
👉It hides
the real design flaw in your code.
👉Makes
debugging harder in large systems.
👉Can lead
to runtime surprises when the lazy bean is accessed later.
💡 Better
Approaches
>Instead of patching with @Lazy, fix the design issue:
>Refactor Dependencies →
Extract common functionality into a third bean that both classes can depend on.
>Event-Driven Communication →
Use Spring Events, Kafka, or RabbitMQ to decouple services.
>For example, instead of OrderService directly calling PaymentService,
publish an event and let PayementService consume it asynchronously.
>This eliminates direct bean-to-bean dependency while improving scalability.
But personally, I prefer using asynchronous communication (Kafka or RabbitMQ)
instead of relying on @Lazy.
It keeps the architecture clean, scalable, and easier to maintain.
Through its flexible registry system, Micrometer can stream telemetry data to dozens of popular monitoring systems, including Prometheus, Datadog, Dynatrace, New Relic, InfluxDB, and Graphana Cloud
