select-dataset-list api solarPanelCnt 추가, spotless 적용

src/main/java/com/kamco/cd/training/train/service/DockerTrainService.java

@@ -262,9 +262,9 @@ public class DockerTrainService {
     c.add("-v");
     c.add(basePath + ":" + basePath); // 심볼릭 링크와 연결되는 실제 파일 경로도 마운트를 해줘야 함
     c.add("-v");
-    c.add(symbolicDir + ":/data");    //요청할경로
+    c.add(symbolicDir + ":/data"); // 요청할경로
     c.add("-v");
-    c.add(responseDir + ":/checkpoints"); //저장될경로
+    c.add(responseDir + ":/checkpoints"); // 저장될경로
 
     // 표준입력 유지 (-it 대신 -i만 사용)
     c.add("-i");

src/main/java/com/kamco/cd/training/train/service/JobRecoveryOnStartupService.java

@@ -384,7 +384,20 @@ public class JobRecoveryOnStartupService {
         return new OutputResult(false, "total-epoch-missing");
       }
 
-      log.info("[RECOVERY] totalEpoch={}. jobId={}", totalEpoch, job.getId());
+      Integer valInterval = extractValInterval(job).orElse(null);
+      if (valInterval == null || valInterval <= 0) {
+        log.warn(
+            "[RECOVERY] valInterval missing or invalid. jobId={}, valInterval={}",
+            job.getId(),
+            valInterval);
+        return new OutputResult(false, "val-interval-missing");
+      }
+
+      log.info(
+          "[RECOVERY] totalEpoch={}. valInterval={}. jobId={}",
+          totalEpoch,
+          valInterval,
+          job.getId());
 
       // 3) val.csv 존재 확인
       Path valCsv = outDir.resolve("val.csv");
@@ -396,14 +409,17 @@ public class JobRecoveryOnStartupService {
       // 4) val.csv 라인 수 확인
       long lines = countNonHeaderLines(valCsv);
 
+      // expected = 실제 val 실행 횟수
+      int expectedLines = totalEpoch / valInterval;
+
       log.info(
           "[RECOVERY] val.csv lines counted. jobId={}, lines={}, expected={}",
           job.getId(),
           lines,
-          totalEpoch);
+          expectedLines);
 
       // 5) 완료 판정
-      if (lines == totalEpoch) {
+      if (lines >= expectedLines) {
         log.info("[RECOVERY] outputs look COMPLETE. jobId={}", job.getId());
         return new OutputResult(true, "ok");
       }
@@ -412,7 +428,7 @@ public class JobRecoveryOnStartupService {
           "[RECOVERY] val.csv line mismatch. jobId={}, lines={}, expected={}",
           job.getId(),
           lines,
-          totalEpoch);
+          expectedLines);
 
       return new OutputResult(
           false, "val.csv-lines-mismatch lines=" + lines + " expected=" + totalEpoch);
@@ -530,4 +546,19 @@ public class JobRecoveryOnStartupService {
       return reason;
     }
   }
+
+  /** paramsJson에서 valInterval 추출 */
+  private Optional<Integer> extractValInterval(ModelTrainJobDto job) {
+    Map<String, Object> params = job.getParamsJson();
+    if (params == null) return Optional.empty();
+
+    Object v = params.get("valInterval");
+    if (v == null) return Optional.empty();
+
+    try {
+      return Optional.of(Integer.parseInt(String.valueOf(v)));
+    } catch (Exception ignore) {
+      return Optional.empty();
+    }
+  }
 }

src/main/java/com/kamco/cd/training/train/service/TmpDatasetService.java

@@ -19,6 +19,7 @@ public class TmpDatasetService {
 
   @Value("${train.docker.symbolic_link_dir}")
   private String symbolicDir;
+
   /**
    * train, val, test 폴더별로 link
    *

src/main/java/com/kamco/cd/training/train/service/TrainJobWorker.java

@@ -132,7 +132,9 @@ public class TrainJobWorker {
         String failMsg = result.getStatus() + "\n" + result.getLogs();
         log.info("training fail exitCode={} Msg ={}", result.getExitCode(), failMsg);
 
-        if (result.getExitCode() == -1 || result.getExitCode() == 143) {
+        if (result.getExitCode() == -1
+            || result.getExitCode() == 143
+            || result.getExitCode() == 137) {
           // 실패 처리
           modelTrainJobCoreService.markPaused(
               jobId, result.getExitCode(), result.getStatus() + "\n" + result.getLogs());