chore(benchmark): 누락된 gatling 시나리오 추가

Author: yyubin

src/gatling/java/benchmark/HeavyLoadSimulation.java

@@ -0,0 +1,100 @@
+package benchmark;
+
+import io.gatling.javaapi.core.*;
+import io.gatling.javaapi.http.*;
+
+import static io.gatling.javaapi.core.CoreDsl.*;
+import static io.gatling.javaapi.http.HttpDsl.*;
+
+import java.time.Duration;
+
+/**
+ * Heavy Load Simulation for JIT Profiling
+ * - Warm-up: ~20,000 requests
+ * - Load test: ~80,000 requests
+ * - Total: ~100,000 requests
+ */
+public class HeavyLoadSimulation extends Simulation {
+
+    HttpProtocolBuilder httpProtocol = http
+            .baseUrl("http://localhost:8080")
+            .acceptHeader("text/plain,application/json")
+            .userAgentHeader("Gatling Heavy Load Test")
+            .shareConnections() // 커넥션 재사용
+            .disableFollowRedirect()
+            .disableWarmUp() // Gatling 기본 warm-up 비활성화
+            .maxConnectionsPerHost(500)
+            .connectionHeader("keep-alive")
+            .header("Keep-Alive", "timeout=5, max=1000");
+
+    // Warm-up 시나리오: 약 20,000 요청
+    // 목적: 서버 JVM JIT 컴파일, 스레드풀, 커넥션 풀 예열
+    ScenarioBuilder warmUp = scenario("Warm-up Phase")
+            .exec(
+                    http("Warm-up request")
+                            .get("/benchmark/hello")
+                            .check(status().is(200))
+            );
+
+    // Heavy Load 시나리오: 약 80,000 요청
+    // 목적: JIT 컴파일 완료 후 steady-state 성능 측정
+    ScenarioBuilder heavyLoad = scenario("Heavy Load Test")
+            .exec(
+                    http("GET /benchmark/hello")
+                            .get("/benchmark/hello")
+                            .check(status().is(200))
+            );
+
+    {
+        setUp(
+                // ========================================
+                // Phase 1: Warm-up (~20,000 requests)
+                // ========================================
+                warmUp.injectOpen(
+                        // 점진적 증가: 0 -> 100 RPS (5초)
+                        rampUsersPerSec(0).to(100).during(Duration.ofSeconds(5)),
+                        // 안정적 부하: 100 RPS (10초) = 1,000 requests
+                        constantUsersPerSec(100).during(Duration.ofSeconds(10)),
+                        // 중간 부하: 300 RPS (20초) = 6,000 requests
+                        constantUsersPerSec(300).during(Duration.ofSeconds(20)),
+                        // 고부하: 500 RPS (25초) = 12,500 requests
+                        constantUsersPerSec(500).during(Duration.ofSeconds(25))
+                        // Total: ~19,500 requests
+                ).protocols(httpProtocol),
+
+                // ========================================
+                // Phase 2: Heavy Load (~80,000 requests)
+                // ========================================
+                heavyLoad.injectOpen(
+                        // warm-up 완료 대기
+                        nothingFor(Duration.ofSeconds(60)),
+
+                        // 점진적 램프업: 100 -> 1000 RPS (10초)
+                        rampUsersPerSec(100).to(1000).during(Duration.ofSeconds(10)),
+
+                        // 고부하 구간 1: 1000 RPS (30초) = 30,000 requests
+                        constantUsersPerSec(1000).during(Duration.ofSeconds(30)),
+
+                        // 최대 부하: 1000 -> 1500 RPS (10초)
+                        rampUsersPerSec(1000).to(1500).during(Duration.ofSeconds(10)),
+
+                        // 고부하 구간 2: 1500 RPS (30초) = 45,000 requests
+                        constantUsersPerSec(1500).during(Duration.ofSeconds(30)),
+
+                        // 점진적 감소: 1500 -> 500 RPS (10초)
+                        rampUsersPerSec(1500).to(500).during(Duration.ofSeconds(10))
+                        // Total: ~80,000 requests
+                ).protocols(httpProtocol)
+        )
+                .assertions(
+                        // 최대 응답시간 1초 이하
+                        global().responseTime().max().lt(1000),
+                        // 95 percentile 응답시간 500ms 이하
+                        global().responseTime().percentile3().lt(500),
+                        // 성공률 99% 이상
+                        global().successfulRequests().percent().gt(99.0),
+                        // 초당 요청 처리량 (throughput) 체크
+                        global().requestsPerSec().gt(500.0)
+                );
+    }
+}