Introduction
The sample microservice application
In this module, you configure a sample Microservice architecture to be ready for autoscaling. Later, you view the application autoscaling details and learn how to trigger it to scale.
To start, you clone a Git repository and run a script that sets up Azure Spring Apps microservice applications that connect to an Azure Database for MySQL.
Your Azure Spring Apps is set up with autoscaling. This feature allows Azure Spring Apps to respond to changes in the environment by adding or removing instances and balancing the load between them. Autoscaling doesn't have any effect on the CPU power, memory, or storage capacity of the application instances powering the app. It only changes the number of application instances.
The script deploys a well-known PetClinic microservice application and is built around small independent services, communicating over HTTP via a REST API. The sample is decomposed into four core microservices. All of them are independently deployable, organized by business domains.
- Customers service: Contains general user input logic and validation including pets and owners information (Name, Address, City, Telephone).
- Visits service: Stores and shows visits information for each pet.
- Vets service: Stores and shows Veterinarians' information, including names and specialties.
- API Gateway: A single entry point into the system, used to handle requests and route them to an appropriate service, and aggregate the results.