Spring Boot Configuration

The flow

1. 1.Resolve properties
2. 2.Select profile
3. 3.Create beans
4. 4.Validate
5. 5.Start

Starters and versions do different jobs

A starter brings a coherent dependency set; auto-configuration creates conditional beans when prerequisites are present. A BOM aligns dependency versions but does not create objects at runtime. The lab pins Spring Boot 4.0.5 and Spring AI 2.0.2 in Maven.

Keep default properties separate from environment-specific endpoints and credentials. Environment variables and command-line arguments can override packaged configuration; investigate the effective property source before assuming a YAML edit took effect.

Use explicit operating modes

The default demo profile provides deterministic chat and eight-dimensional fixture embeddings. The ollama profile uses real provider adapters with a separate embedding dimension. Never reuse the demo index with real embeddings: matching dimensions alone would not make vector spaces compatible.

The lab disables automatic model configuration by default and enables it in application-ollama.yml. A missing database password or demo user password fails startup. Profiles are a composition mechanism, not an authorization boundary.

Validate before accepting traffic

Use typed configuration and validation for larger settings groups. Keep local schema initialization opt-in; production migrations should run with separate privileges. Do not print secret values to diagnose configuration.

Try it

server:
  address: 127.0.0.1
  port: 8081
spring:
  profiles:
    default: demo
  datasource:
    url: ${DATABASE_URL:jdbc:postgresql://127.0.0.1:5438/interview}
    username: ${DATABASE_USER:interview}
    password: ${DATABASE_PASSWORD}
  ai:
    model:
      chat: none
      embedding: none
lab:
  initialize-schema: ${INITIALIZE_SCHEMA:false}
  dimensions: ${EMBEDDING_DIMENSIONS:8}
  alpha-password: ${DEMO_ALPHA_PASSWORD}
  beta-password: ${DEMO_BETA_PASSWORD}

examples/spring-ai-lab/src/main/resources/application.yml · mvn -f examples/spring-ai-lab/pom.xml test. Java 21+, Spring Boot 4.0.5, Spring AI 2.0.2 where applicable. The demo uses synthetic models; live inference quality is a separate check.

Interview questions

Official references

• Spring Boot — external configuration

Continue learning

• REST, Validation & Errors →
• Spring AI: Getting Started →