HDFS-3664. BlockManager race when stopping active services. Contributed by Colin Patrick McCabe

git-svn-id: https://svn.apache.org/repos/asf/hadoop/common/trunk@1383753 13f79535-47bb-0310-9956-ffa450edef68

hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt

@@ -755,6 +755,9 @@ Release 2.0.2-alpha - 2012-09-07
     HDFS-2757. Cannot read a local block that's being written to when
     using the local read short circuit. (Jean-Daniel Cryans via eli)
 
+    HDFS-3664. BlockManager race when stopping active services.
+    (Colin Patrick McCabe via eli)
+    
   BREAKDOWN OF HDFS-3042 SUBTASKS
 
     HDFS-2185. HDFS portion of ZK-based FailoverController (todd)

hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/blockmanagement/BlockManager.java

@@ -363,11 +363,10 @@ public class BlockManager {
         replicationThread.join(3000);
       }
     } catch (InterruptedException ie) {
-    } finally {
-      if (pendingReplications != null) pendingReplications.stop();
-      blocksMap.close();
-      datanodeManager.close();
     }
+    datanodeManager.close();
+    pendingReplications.stop();
+    blocksMap.close();
   }
 
   /** @return the datanodeManager */

hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/blockmanagement/DatanodeManager.java

@@ -192,7 +192,13 @@ public class DatanodeManager {
   }
 
   void close() {
-    if (decommissionthread != null) decommissionthread.interrupt();
+    if (decommissionthread != null) {
+      decommissionthread.interrupt();
+      try {
+        decommissionthread.join(3000);
+      } catch (InterruptedException e) {
+      }
+    }
     heartbeatManager.close();
   }
 

hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/blockmanagement/HeartbeatManager.java

@@ -74,6 +74,11 @@ class HeartbeatManager implements DatanodeStatistics {
 
   void close() {
     heartbeatThread.interrupt();
+    try {
+      // This will no effect if the thread hasn't yet been started.
+      heartbeatThread.join(3000);
+    } catch (InterruptedException e) {
+    }
   }
   
   synchronized int getLiveDatanodeCount() {