svn commit: r1518899 - in /hadoop/common/trunk/hadoop-hdfs-project/hadoop-hdfs: CHANGES.txt src/main/java/org/apache/hadoop/hdfs/server/namenode/FSNamesystem.java

[jing9]

Author: jing9
Date: Fri Aug 30 07:36:45 2013
New Revision: 1518899

URL: http://svn.apache.org/r1518899
Log:
HDFS-5140. Too many safemode monitor threads being created in the standby 
namenode causing it to fail with out of memory error. Contributed by Jing Zhao.

Modified:
    hadoop/common/trunk/hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt
    
hadoop/common/trunk/hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/namenode/FSNamesystem.java

Modified: hadoop/common/trunk/hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt
URL: 
http://svn.apache.org/viewvc/hadoop/common/trunk/hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt?rev=1518899&r1=1518898&r2=1518899&view=diff
==============================================================================
--- hadoop/common/trunk/hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt (original)
+++ hadoop/common/trunk/hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt Fri Aug 30 
07:36:45 2013
@@ -407,6 +407,9 @@ Release 2.1.1-beta - UNRELEASED
     HDFS-5077. NPE in FSNamesystem.commitBlockSynchronization().
     (Plamen Jeliazkov via shv)
 
+    HDFS-5140. Too many safemode monitor threads being created in the standby 
+    namenode causing it to fail with out of memory error. (jing9)
+
 Release 2.1.0-beta - 2013-08-22
 
   INCOMPATIBLE CHANGES

Modified: 
hadoop/common/trunk/hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/namenode/FSNamesystem.java
URL: 
http://svn.apache.org/viewvc/hadoop/common/trunk/hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/namenode/FSNamesystem.java?rev=1518899&r1=1518898&r2=1518899&view=diff
==============================================================================
--- 
hadoop/common/trunk/hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/namenode/FSNamesystem.java
 (original)
+++ 
hadoop/common/trunk/hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/namenode/FSNamesystem.java
 Fri Aug 30 07:36:45 2013
@@ -366,7 +366,7 @@ public class FSNamesystem implements Nam
 
   final LeaseManager leaseManager = new LeaseManager(this); 
 
-  Daemon smmthread = null;  // SafeModeMonitor thread
+  volatile Daemon smmthread = null;  // SafeModeMonitor thread
   
   Daemon nnrmthread = null; // NamenodeResourceMonitor thread
 
@@ -4555,7 +4555,9 @@ public class FSNamesystem implements Nam
       // Have to have write-lock since leaving safemode initializes
       // repl queues, which requires write lock
       assert hasWriteLock();
-      if (needEnter()) {
+      // if smmthread is already running, the block threshold must have been 
+      // reached before, there is no need to enter the safe mode again
+      if (smmthread == null && needEnter()) {
         enter();
         // check if we are ready to initialize replication queues
         if (canInitializeReplQueues() && !isPopulatingReplQueues()) {
@@ -4564,7 +4566,7 @@ public class FSNamesystem implements Nam
         reportStatus("STATE* Safe mode ON.", false);
         return;
       }
-      // the threshold is reached
+      // the threshold is reached or was reached before
       if (!isOn() ||                           // safe mode is off
           extension <= 0 || threshold <= 0) {  // don't need to wait
         this.leave(); // leave safe mode
@@ -4576,9 +4578,11 @@ public class FSNamesystem implements Nam
       }
       // start monitor
       reached = now();
-      smmthread = new Daemon(new SafeModeMonitor());
-      smmthread.start();
-      reportStatus("STATE* Safe mode extension entered.", true);
+      if (smmthread == null) {
+        smmthread = new Daemon(new SafeModeMonitor());
+        smmthread.start();
+        reportStatus("STATE* Safe mode extension entered.", true);
+      }
 
       // check if we are ready to initialize replication queues
       if (canInitializeReplQueues() && !isPopulatingReplQueues()) {
@@ -4814,6 +4818,7 @@ public class FSNamesystem implements Nam
           if (safeMode.canLeave()) {
             // Leave safe mode.
             safeMode.leave();
+            smmthread = null;
             break;
           }
         } finally {
@@ -4829,7 +4834,6 @@ public class FSNamesystem implements Nam
       if (!fsRunning) {
         LOG.info("NameNode is being shutdown, exit SafeModeMonitor thread");
       }
-      smmthread = null;
     }
   }