⚡️ [Perf] FastAPI 클러스터링 결과 저장 로직 성능 개선

반복적인 DB저장 호출 최소화

Author: hyeonda02

src/main/java/com/going/server/global/temp/service/FastApiService.java

@@ -7,6 +7,7 @@
 import com.going.server.domain.word.entity.Word;
 import com.going.server.domain.word.repository.WordRepository;
 import jakarta.annotation.PostConstruct;
+import java.util.ArrayList;
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
 import org.springframework.beans.factory.annotation.Value;
@@ -54,52 +55,51 @@ public String callFastApi() {
 
     @Profile("!test")
     public void setCluster() {
-        // FastAPI 요청 데이터 (필요시 변경)
         Map<String, Object> requestData = Map.of("input_text", "클러스터링할 데이터 예제");
 
-        // FastAPI 응답 받기
         Map<String, Object> response = webClient.post()
-                .uri(baseUrl + "/api/cluster")  // FastAPI 클러스터링 엔드포인트
+                .uri(baseUrl + "/api/cluster")
                 .bodyValue(requestData)
                 .retrieve()
                 .bodyToMono(Map.class)
                 .block();
 
-        //모든 클러스터링 결과 저장
         List<Map<String, Object>> clusters = (List<Map<String, Object>>) response.get("clusters");
-        //클러스터링 결과 이미지 저장
         String imageUrl = response.get("image_url").toString();
 
+        // 병렬처리를 위한 리스트
+        List<Cluster> clusterEntities = new ArrayList<>();
+        List<Word> wordEntities = new ArrayList<>();
+        List<Sentence> sentenceEntities = new ArrayList<>();
+
         for (Map<String, Object> cluster : clusters) {
             Map<String, List<String>> wordSentences = (Map<String, List<String>>) cluster.get("word_sentences");
 
-            if (wordSentences.isEmpty()) continue; // 빈 클러스터 예외 처리
+            if (wordSentences.isEmpty()) continue;
 
-            //첫 번째 단어를 대표 어휘로 설정
             String representWord = wordSentences.keySet().iterator().next();
-            //엔티티 저장
             Cluster clusterEntity = Cluster.toEntity(representWord, imageUrl);
-            //클러스터 결과 DB에 저장
-            Cluster saveCluster = clusterRepository.save(clusterEntity);
+            clusterEntities.add(clusterEntity);
 
             for (Map.Entry<String, List<String>> entry : wordSentences.entrySet()) {
-                //단어
-                String word = entry.getKey();
-                //문장들
-                List<String> sentences = entry.getValue();
-                //Word 엔티티 생성
-                Word wordEntity = Word.toEntity(word, saveCluster);
-                //DB에 저장
-                Word saveWord = wordRepository.save(wordEntity);
-                for (String sentence : sentences) {
-                    //Sentence 엔티티 생성
-                    Sentence sententEntity = Sentence.toEntity(sentence, saveWord);
-                    sentenceRepository.save(sententEntity);
+                Word wordEntity = Word.toEntity(entry.getKey(), clusterEntity);
+                wordEntities.add(wordEntity);
+
+                for (String sentence : entry.getValue()) {
+                    sentenceEntities.add(Sentence.toEntity(sentence, wordEntity));
                 }
             }
         }
+
+        // 1. 클러스터 저장
+        clusterRepository.saveAll(clusterEntities);
+
+        // 2. 클러스터 저장 후 Word에 연결된 객체들을 저장
+        wordRepository.saveAll(wordEntities);
+        sentenceRepository.saveAll(sentenceEntities);
     }
 
+
     public Word testWord(String word) {
         Word wordEntity = Word.toEntity(word, null);
         return wordRepository.save(wordEntity);