Custom actuator endpoints in Spring Boot - Performance & Optimization

Performance: Custom actuator endpoints

MEDIUM IMPACT

This affects the server response time and the size of monitoring data sent to clients, impacting backend performance and network load.

Creating a custom actuator endpoint to expose application metrics

Spring Boot

import org.springframework.boot.actuate.endpoint.annotation.Endpoint;
import org.springframework.boot.actuate.endpoint.annotation.ReadOperation;
import java.util.concurrent.atomic.AtomicReference;
import java.util.Map;

@Endpoint(id = "cachedMetrics")
public class CachedMetricsEndpoint {
    private final AtomicReference<Map<String, Object>> cache = new AtomicReference<>();

    public CachedMetricsEndpoint() {
        // Precompute or update cache asynchronously
        cache.set(Map.of("data", "precomputed result"));
    }

    @ReadOperation
    public Map<String, Object> metrics() {
        return cache.get();
    }
}

Precomputes data or caches results to avoid expensive operations during requests, ensuring fast response.

📈 Performance GainResponds instantly without blocking, improving server throughput and reducing latency

Creating a custom actuator endpoint to expose application metrics

Spring Boot

import org.springframework.boot.actuate.endpoint.annotation.Endpoint;
import org.springframework.boot.actuate.endpoint.annotation.ReadOperation;
import java.util.Map;

@Endpoint(id = "heavyMetrics")
public class HeavyMetricsEndpoint {
    @ReadOperation
    public Map<String, Object> metrics() throws InterruptedException {
        // Simulate expensive operation
        Thread.sleep(5000);
        return Map.of("data", "heavy computation result");
    }
}

The endpoint performs a blocking 5-second operation on each request, delaying response and increasing server load.

📉 Performance CostBlocks response for 5 seconds per request, increasing server latency and reducing throughput

Performance Comparison

Pattern	Server Processing	Response Delay	Network Payload	Verdict
Blocking heavy computation in endpoint	High CPU and blocking	5 seconds delay	Small JSON	[X] Bad
Cached or precomputed endpoint data	Minimal CPU, non-blocking	Instant response	Small JSON	[OK] Good

Rendering Pipeline

Custom actuator endpoints run on the server and affect backend processing time before the browser receives data. Slow endpoints delay the server response, increasing time to first byte and overall load time.

→Server Processing

→Network Transfer

⚠️ BottleneckServer Processing when endpoint logic is expensive or blocking

Optimization Tips

1Avoid blocking or long-running operations inside custom actuator endpoints.

2Use caching or asynchronous precomputation to serve data quickly.

3Keep JSON payloads small to reduce network transfer time.

Performance Quiz - 3 Questions

What is the main performance risk of a custom actuator endpoint that performs heavy computation on each request?

A It causes layout shifts on the page

B It increases browser rendering time

C It increases server response time and delays client data reception

D It reduces network bandwidth usage

DevTools: Network

How to check: Open browser DevTools, go to Network tab, trigger the actuator endpoint request, and observe the response time and size.

What to look for: Look for long waiting times (TTFB) indicating slow server processing and check payload size for network impact.