MAPREDUCE-6273. HistoryFileManager should check whether summaryFile exists

to avoid FileNotFoundException causing HistoryFileInfo into MOVE_FAILED
state. Contributed by zhihai xu.

(cherry picked from commit 1ea90144d5)

hadoop-mapreduce-project/CHANGES.txt

@@ -167,6 +167,10 @@ Release 2.8.0 - UNRELEASED
     MAPREDUCE-5708. Duplicate String.format in YarnOutputFiles.getSpillFileForWrite.
     (Konstantin Weitz via devaraj)
 
+    MAPREDUCE-6273. HistoryFileManager should check whether summaryFile exists to 
+    avoid FileNotFoundException causing HistoryFileInfo into MOVE_FAILED state.
+    (zhihai xu via devaraj)
+
 Release 2.7.1 - UNRELEASED
 
   INCOMPATIBLE CHANGES

hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-hs/src/main/java/org/apache/hadoop/mapreduce/v2/hs/HistoryFileManager.java

@@ -303,8 +303,9 @@ public class HistoryFileManager extends AbstractService {
     private JobIndexInfo jobIndexInfo;
     private HistoryInfoState state;
 
-    private HistoryFileInfo(Path historyFile, Path confFile, Path summaryFile,
+    @VisibleForTesting
-        JobIndexInfo jobIndexInfo, boolean isInDone) {
+    protected HistoryFileInfo(Path historyFile, Path confFile,
+        Path summaryFile, JobIndexInfo jobIndexInfo, boolean isInDone) {
       this.historyFile = historyFile;
       this.confFile = confFile;
       this.summaryFile = summaryFile;
@@ -337,7 +338,8 @@ public class HistoryFileManager extends AbstractService {
              + " historyFile = " + historyFile;
     }
 
-    private synchronized void moveToDone() throws IOException {
+    @VisibleForTesting
+    synchronized void moveToDone() throws IOException {
       if (LOG.isDebugEnabled()) {
         LOG.debug("moveToDone: " + historyFile);
       }
@@ -368,7 +370,8 @@ public class HistoryFileManager extends AbstractService {
           paths.add(confFile);
         }
 
-        if (summaryFile == null) {
+        if (summaryFile == null || !intermediateDoneDirFc.util().exists(
+            summaryFile)) {
           LOG.info("No summary file for job: " + jobId);
         } else {
           String jobSummaryString = getJobSummary(intermediateDoneDirFc,

hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-hs/src/test/java/org/apache/hadoop/mapreduce/v2/hs/TestHistoryFileManager.java

@@ -35,7 +35,12 @@ import org.apache.hadoop.fs.FileContext;
 import org.apache.hadoop.hdfs.HdfsConfiguration;
 import org.apache.hadoop.hdfs.MiniDFSCluster;
 import org.apache.hadoop.hdfs.protocol.HdfsConstants;
+import org.apache.hadoop.mapreduce.JobID;
+import org.apache.hadoop.mapreduce.TypeConverter;
+import org.apache.hadoop.mapreduce.v2.api.records.JobId;
+import org.apache.hadoop.mapreduce.v2.hs.HistoryFileManager.HistoryFileInfo;
 import org.apache.hadoop.mapreduce.v2.jobhistory.JHAdminConfig;
+import org.apache.hadoop.mapreduce.v2.jobhistory.JobIndexInfo;
 import org.apache.hadoop.test.CoreTestDriver;
 import org.apache.hadoop.yarn.conf.YarnConfiguration;
 import org.apache.hadoop.yarn.exceptions.YarnRuntimeException;
@@ -221,4 +226,34 @@ public class TestHistoryFileManager {
 
   }
 
+  @Test
+  public void testHistoryFileInfoSummaryFileNotExist() throws Exception {
+    HistoryFileManagerTest hmTest = new HistoryFileManagerTest();
+    String job = "job_1410889000000_123456";
+    Path summaryFile = new Path(job + ".summary");
+    JobIndexInfo jobIndexInfo = new JobIndexInfo();
+    jobIndexInfo.setJobId(TypeConverter.toYarn(JobID.forName(job)));
+    Configuration conf = dfsCluster.getConfiguration(0);
+    conf.set(JHAdminConfig.MR_HISTORY_DONE_DIR,
+        "/" + UUID.randomUUID());
+    conf.set(JHAdminConfig.MR_HISTORY_INTERMEDIATE_DONE_DIR,
+        "/" + UUID.randomUUID());
+    hmTest.serviceInit(conf);
+    HistoryFileInfo info = hmTest.getHistoryFileInfo(null, null,
+        summaryFile, jobIndexInfo, false);
+    info.moveToDone();
+    Assert.assertFalse(info.didMoveFail());
+  }
+
+  static class HistoryFileManagerTest extends HistoryFileManager {
+    public HistoryFileManagerTest() {
+      super();
+    }
+    public HistoryFileInfo getHistoryFileInfo(Path historyFile,
+        Path confFile, Path summaryFile, JobIndexInfo jobIndexInfo,
+        boolean isInDone) {
+      return new HistoryFileInfo(historyFile, confFile, summaryFile,
+          jobIndexInfo, isInDone);
+    }
+  }
 }