Actuator is like adding a health monitoring system to your Spring Boot app. It gives you real-time metrics, health checks, and other operational info — all via HTTP endpoints like:

http://localhost:8080/actuator/metrics

🎯 Why Is It Useful for Microservices?

Imagine you have 10 microservices running. You need to answer:

  • Is my app healthy?
  • Are the APIs responding fast?
  • Is the memory usage too high?
  • Are the DB connections overloaded?

👉 Actuator gives you these answers out-of-the-box.

🔍 What Metrics Does Actuator Provide?

Let’s explore the main types of metrics you’ll find useful for monitoring:

1. 🧠 JVM Metrics (Java Virtual Machine)

These tell you how your app is using memory, threads, and classes.

Metric What It Tells You Example
jvm.memory.used How much memory is used by your app 128 MB
jvm.memory.max The max memory your app can use 512 MB
jvm.threads.live How many threads are running 100
jvm.classes.loaded Number of classes loaded into memory 10,000

💡 Why care? If memory or thread usage spikes too much, your app could crash.

2. 🖥️ CPU & System Metrics

Metric What It Tells You Example
system.cpu.usage Total CPU usage (system-wide) 0.60 (60%)
process.cpu.usage CPU used by your app only 0.20 (20%)
system.cpu.count Number of processor cores 8
process.uptime How long the app has been running 4 hours

💡 Why care? High CPU usage may slow down your APIs.

3. 🌐 HTTP Request Metrics

Metric What It Tells You Example
http.server.requests Number of API calls + response time 500 calls, avg 120ms
→ Tags: uri, method, status Helps slice metrics per API route /login, POST, 200 OK

💡 Why care? You can find slow endpoints or too many errors (e.g. 500s).

4. 💽 Database (HikariCP) Metrics

If using a database connection pool (like HikariCP):

Metric What It Tells You Example
hikaricp.connections.active Number of DB connections in use 5
hikaricp.connections.idle Available connections in pool 10
hikaricp.connections.max Maximum allowed connections 20

💡 Why care? If all DB connections are used, your app can freeze or slow down.

5. 📦 Cache Metrics

If you use caching (like Caffeine, Redis, etc.):

Metric What It Tells You Example
cache.gets How many times you read from the cache 1000 hits
cache.misses How many times data wasn't in cache 200 misses
cache.puts How many times you added to cache 300 puts

💡 Why care? Too many cache misses = your DB is doing extra work.

6. 💓 Health Endpoint

Shows status of your app and its dependencies (DB, disk, etc.).

📍 Endpoint:

http://localhost:8080/actuator/health

Example output:

{
  "status": "UP",
  "components": {
    "db": { "status": "UP" },
    "diskSpace": { "status": "UP" }
  }
}

💡 Helps monitoring tools like Prometheus, Grafana, or Kubernetes to know if your service is healthy.

7. 📈 Custom Metrics (Optional but Powerful)

You can define your own metrics like:

meterRegistry.counter("orders.placed").increment();
meterRegistry.gauge("temperature", () -> getRoomTemp());

💡 Used to track business logic (e.g., number of signups, order volume).

⚙️ How to Enable Actuator Metrics

  1. Add the dependency: 
org.springframework.boot
  spring-boot-starter-actuator
  1. In application.properties: 
management.endpoints.web.exposure.include=*
  1. For Prometheus/Grafana: 
io.micrometer
  micrometer-registry-prometheus

Then hit:

http://localhost:8080/actuator/prometheus

🧪 Use Case: Example for Microservice Monitoring

Let’s say you have a payment microservice:

  • /actuator/health: Used by Kubernetes for health checks.
  • /actuator/metrics: View API performance, DB usage.
  • http.server.requests: See if /pay API is slow.
  • hikaricp.connections.active: Monitor DB pool overload.
  • orders.placed: Track how many payments succeeded.

Would you like a dashboard setup in Grafana, or a custom Prometheus config example to visualize these metrics?