서비스 검색 및 통신

In a microservices architecture, applications are broken into many small, independent services. These services need to find and talk to each other to work together.

Service discovery is the automatic process by which services locate each other on a network.

• It solves the problem of services needing to know each other's network locations (IP addresses, ports).
• Essential for dynamic, scalable, and resilient systems.

Imagine you have a 'User Service' and an 'Order Service'. If the Order Service needs user data, it must know the User Service's address.

Without service discovery:

• You might hardcode IP addresses and ports.
• If a service scales up or moves, its address changes, breaking communication.
• Manual updates are error-prone and time-consuming.

This approach isn't feasible for dynamic cloud environments.

At the heart of service discovery is the Service Registry. Think of it as a phone book for your services.

• It's a central database that stores the network locations of all active service instances.
• When a service starts, it registers itself with the registry.
• When a service needs to communicate, it queries the registry to find the target service's address.

Popular examples include HashiCorp Consul, Netflix Eureka, and etcd.

Services need a way to tell the registry they exist and where they can be reached. There are two main patterns:

• Self-Registration: The service itself registers and de-registers with the service registry. It also sends periodic heartbeats to prove it's still alive.
• Third-Party Registration: A separate component (often called a 'Registrar' or 'Agent') handles registration for the service. This decouples the service from the discovery mechanism.

Both methods ensure the registry has up-to-date information.

In client-side discovery, the client service is responsible for querying the service registry to find available instances of a target service.

• The client uses a discovery client library (e.g., Spring Cloud Netflix Eureka Client).
• It retrieves a list of service instances from the registry.
• It then uses a load-balancing algorithm (like round-robin) to select an instance and make a direct request.

This approach puts discovery logic into each client service.

With server-side discovery, a dedicated component (often a load balancer, API Gateway, or router) handles service lookup.

• The client makes a request to a well-known address (e.g., the load balancer).
• The load balancer queries the service registry to find an available instance of the target service.
• It then forwards the client's request to that instance.

This pattern simplifies client logic, as clients don't need discovery libraries.

Once a service has discovered the address of another service, they need to communicate. This typically involves making requests and receiving responses.

• Communication can be synchronous (request-response) or asynchronous (event-driven).
• The choice depends on whether the calling service needs an immediate response or can continue processing.

Let's look at common synchronous methods first.

Synchronous communication means the calling service waits for a response from the called service. The most common protocols are HTTP/REST and gRPC.

Here's a conceptual Java example demonstrating how a service might register and a client might find it to make a 'request':

While synchronous communication is direct, asynchronous communication uses message queues or event streams (as discussed in the previous lesson).

• Services don't wait for an immediate response.
• They publish events or messages to a queue, and other services consume them when ready.
• This decouples services, improving resilience and scalability.

Service discovery ensures event producers and consumers can find the message broker.

Service discovery isn't just about finding services; it enables crucial microservice benefits:

• Load Balancing: If multiple instances of a service are registered, the discovery mechanism (client-side or server-side) can distribute requests evenly among them.
• Resilience: If a service instance fails, it stops sending heartbeats or is de-registered. The registry updates, and clients/load balancers automatically stop routing requests to the failed instance.

This dynamic adaptability is key to robust microservices.

Consider a microservices setup where a 'Product Service' needs to call a 'Review Service'. The Review Service has multiple instances running.

Which of the following best describes the role of a Service Registry in this scenario?

In this lesson, we explored the critical concepts of service discovery and communication in microservices.

• Service Discovery allows services to find each other dynamically.
• The Service Registry is the central 'phone book' for service instances.
• We learned about client-side and server-side discovery patterns.
• Services communicate synchronously (e.g., HTTP/REST) or asynchronously (e.g., message queues).
• Discovery enables key benefits like load balancing and resilience.

Understanding these patterns is vital for building scalable and maintainable microservice architectures.