svn merge -c 1292419 from trunk for HDFS-2907.

git-svn-id: https://svn.apache.org/repos/asf/hadoop/common/branches/branch-0.23@1292428 13f79535-47bb-0310-9956-ffa450edef68

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

@@ -25,6 +25,9 @@ Release 0.23.2 - UNRELEASED
     HDFS-2725. hdfs script usage information is missing the information 
     about "dfs" command (Prashant Sharma via stevel)
 
+    HDFS-2907.  Add a conf property dfs.datanode.fsdataset.factory to make
+    FSDataset in Datanode pluggable.  (szetszwo)
+
   OPTIMIZATIONS
 
   BUG FIXES

hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/DFSConfigKeys.java

@@ -167,7 +167,6 @@ public class DFSConfigKeys extends CommonConfigurationKeys {
   public static final String  DFS_CLIENT_RETRY_WINDOW_BASE= "dfs.client.retry.window.base";
   public static final String  DFS_METRICS_SESSION_ID_KEY = "dfs.metrics.session-id";
   public static final String  DFS_DATANODE_HOST_NAME_KEY = "dfs.datanode.hostname";
-  public static final String  DFS_DATANODE_STORAGEID_KEY = "dfs.datanode.StorageId";
   public static final String  DFS_NAMENODE_HOSTS_KEY = "dfs.namenode.hosts";
   public static final String  DFS_NAMENODE_HOSTS_EXCLUDE_KEY = "dfs.namenode.hosts.exclude";
   public static final String  DFS_CLIENT_SOCKET_TIMEOUT_KEY = "dfs.client.socket-timeout";
@@ -211,10 +210,6 @@ public class DFSConfigKeys extends CommonConfigurationKeys {
   public static final int     DFS_DATANODE_NUMBLOCKS_DEFAULT = 64;
   public static final String  DFS_DATANODE_SCAN_PERIOD_HOURS_KEY = "dfs.datanode.scan.period.hours";
   public static final int     DFS_DATANODE_SCAN_PERIOD_HOURS_DEFAULT = 0;
-  public static final String  DFS_DATANODE_SIMULATEDDATASTORAGE_KEY = "dfs.datanode.simulateddatastorage";
-  public static final boolean DFS_DATANODE_SIMULATEDDATASTORAGE_DEFAULT = false;
-  public static final String  DFS_DATANODE_SIMULATEDDATASTORAGE_CAPACITY_KEY = "dfs.datanode.simulateddatastorage.capacity";
-  public static final long    DFS_DATANODE_SIMULATEDDATASTORAGE_CAPACITY_DEFAULT = 2L<<40;
   public static final String  DFS_DATANODE_TRANSFERTO_ALLOWED_KEY = "dfs.datanode.transferTo.allowed";
   public static final boolean DFS_DATANODE_TRANSFERTO_ALLOWED_DEFAULT = true;
   public static final String  DFS_DATANODE_BLOCKVOLUMECHOICEPOLICY = "dfs.datanode.block.volume.choice.policy";
@@ -282,6 +277,7 @@ public class DFSConfigKeys extends CommonConfigurationKeys {
 
   //Keys with no defaults
   public static final String  DFS_DATANODE_PLUGINS_KEY = "dfs.datanode.plugins";
+  public static final String  DFS_DATANODE_FSDATASET_FACTORY_KEY = "dfs.datanode.fsdataset.factory";
   public static final String  DFS_DATANODE_SOCKET_WRITE_TIMEOUT_KEY = "dfs.datanode.socket.write.timeout";
   public static final String  DFS_DATANODE_STARTUP_KEY = "dfs.datanode.startup";
   public static final String  DFS_NAMENODE_PLUGINS_KEY = "dfs.namenode.plugins";

hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/HdfsConfiguration.java

@@ -88,7 +88,6 @@ public class HdfsConfiguration extends Configuration {
     deprecate("fs.checkpoint.period", DFSConfigKeys.DFS_NAMENODE_CHECKPOINT_PERIOD_KEY);
     deprecate("dfs.upgrade.permission", DFSConfigKeys.DFS_NAMENODE_UPGRADE_PERMISSION_KEY);
     deprecate("heartbeat.recheck.interval", DFSConfigKeys.DFS_NAMENODE_HEARTBEAT_RECHECK_INTERVAL_KEY);
-    deprecate("StorageId", DFSConfigKeys.DFS_DATANODE_STORAGEID_KEY);
     deprecate("dfs.https.client.keystore.resource", DFSConfigKeys.DFS_CLIENT_HTTPS_KEYSTORE_RESOURCE_KEY);
     deprecate("dfs.https.need.client.auth", DFSConfigKeys.DFS_CLIENT_HTTPS_NEED_AUTH_KEY);
     deprecate("slave.host.name", DFSConfigKeys.DFS_DATANODE_HOST_NAME_KEY);

hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/DataNode.java

@@ -43,10 +43,7 @@ import static org.apache.hadoop.hdfs.DFSConfigKeys.DFS_DATANODE_KEYTAB_FILE_KEY;
 import static org.apache.hadoop.hdfs.DFSConfigKeys.DFS_DATANODE_PLUGINS_KEY;
 import static org.apache.hadoop.hdfs.DFSConfigKeys.DFS_DATANODE_SCAN_PERIOD_HOURS_DEFAULT;
 import static org.apache.hadoop.hdfs.DFSConfigKeys.DFS_DATANODE_SCAN_PERIOD_HOURS_KEY;
-import static org.apache.hadoop.hdfs.DFSConfigKeys.DFS_DATANODE_SIMULATEDDATASTORAGE_DEFAULT;
-import static org.apache.hadoop.hdfs.DFSConfigKeys.DFS_DATANODE_SIMULATEDDATASTORAGE_KEY;
 import static org.apache.hadoop.hdfs.DFSConfigKeys.DFS_DATANODE_STARTUP_KEY;
-import static org.apache.hadoop.hdfs.DFSConfigKeys.DFS_DATANODE_STORAGEID_KEY;
 import static org.apache.hadoop.hdfs.DFSConfigKeys.DFS_DATANODE_USER_NAME_KEY;
 import static org.apache.hadoop.hdfs.DFSConfigKeys.DFS_FEDERATION_NAMESERVICES;
 
@@ -152,13 +149,11 @@ import org.apache.hadoop.util.DiskChecker;
 import org.apache.hadoop.util.DiskChecker.DiskErrorException;
 import org.apache.hadoop.util.DiskChecker.DiskOutOfSpaceException;
 import org.apache.hadoop.util.GenericOptionsParser;
-import org.apache.hadoop.util.ReflectionUtils;
 import org.apache.hadoop.util.ServicePlugin;
 import org.apache.hadoop.util.StringUtils;
 import org.apache.hadoop.util.VersionInfo;
 import org.mortbay.util.ajax.JSON;
 
-import com.google.common.base.Preconditions;
 
 
 /**********************************************************
@@ -427,13 +422,14 @@ public class DataNode extends Configured
     }
   }
 
-  private synchronized void setClusterId(String cid) throws IOException {
-    if(clusterId != null && !clusterId.equals(cid)) {
-      throw new IOException ("cluster id doesn't match. old cid=" + clusterId 
-          + " new cid="+ cid);
+  private synchronized void setClusterId(final String nsCid, final String bpid
+      ) throws IOException {
+    if(clusterId != null && !clusterId.equals(nsCid)) {
+      throw new IOException ("Cluster IDs not matched: dn cid=" + clusterId 
+          + " but ns cid="+ nsCid + "; bpid=" + bpid);
     }
     // else
-    clusterId = cid;    
+    clusterId = nsCid;
   }
 
   private static String getHostName(Configuration config)
@@ -810,51 +806,22 @@ public class DataNode extends Configured
    */
   void initBlockPool(BPOfferService bpos) throws IOException {
     NamespaceInfo nsInfo = bpos.getNamespaceInfo();
-    Preconditions.checkState(nsInfo != null,
-        "Block pool " + bpos + " should have retrieved " +
-        "its namespace info before calling initBlockPool.");
+    if (nsInfo == null) {
+      throw new IOException("NamespaceInfo not found: Block pool " + bpos
+          + " should have retrieved namespace info before initBlockPool.");
+    }
     
-    String blockPoolId = nsInfo.getBlockPoolID();
-
     // Register the new block pool with the BP manager.
     blockPoolManager.addBlockPool(bpos);
 
-    synchronized (this) {
-      // we do not allow namenode from different cluster to register
-      if(clusterId != null && !clusterId.equals(nsInfo.clusterID)) {
-        throw new IOException(
-            "cannot register with the namenode because clusterid do not match:"
-            + " nn=" + nsInfo.getBlockPoolID() + "; nn cid=" + nsInfo.clusterID + 
-            ";dn cid=" + clusterId);
-      }
-
-      setClusterId(nsInfo.clusterID);
-    }
-    
-    StartupOption startOpt = getStartupOption(conf);
-    assert startOpt != null : "Startup option must be set.";
-
-    boolean simulatedFSDataset = conf.getBoolean(
-        DFS_DATANODE_SIMULATEDDATASTORAGE_KEY,
-        DFS_DATANODE_SIMULATEDDATASTORAGE_DEFAULT);
-    
-    if (!simulatedFSDataset) {
-      // read storage info, lock data dirs and transition fs state if necessary          
-      storage.recoverTransitionRead(DataNode.this, blockPoolId, nsInfo,
-          dataDirs, startOpt);
-      StorageInfo bpStorage = storage.getBPStorage(blockPoolId);
-      LOG.info("setting up storage: nsid=" +
-          bpStorage.getNamespaceID() + ";bpid="
-          + blockPoolId + ";lv=" + storage.getLayoutVersion() +
-          ";nsInfo=" + nsInfo);
-    }
+    setClusterId(nsInfo.clusterID, nsInfo.getBlockPoolID());
     
     // In the case that this is the first block pool to connect, initialize
     // the dataset, block scanners, etc.
-    initFsDataSet();
+    initStorage(nsInfo);
     initPeriodicScanners(conf);
     
-    data.addBlockPool(blockPoolId, conf);
+    data.addBlockPool(nsInfo.getBlockPoolID(), conf);
   }
 
   /**
@@ -881,31 +848,28 @@ public class DataNode extends Configured
    * Initializes the {@link #data}. The initialization is done only once, when
    * handshake with the the first namenode is completed.
    */
-  private synchronized void initFsDataSet() throws IOException {
-    if (data != null) { // Already initialized
-      return;
-    }
-
-    // get version and id info from the name-node
-    boolean simulatedFSDataset = conf.getBoolean(
-        DFS_DATANODE_SIMULATEDDATASTORAGE_KEY,
-        DFS_DATANODE_SIMULATEDDATASTORAGE_DEFAULT);
-
-    if (simulatedFSDataset) {
-      storage.createStorageID(getPort());
-      // it would have been better to pass storage as a parameter to
-      // constructor below - need to augment ReflectionUtils used below.
-      conf.set(DFS_DATANODE_STORAGEID_KEY, getStorageId());
-      try {
-        data = (FSDatasetInterface) ReflectionUtils.newInstance(
-            Class.forName(
-            "org.apache.hadoop.hdfs.server.datanode.SimulatedFSDataset"),
-            conf);
-      } catch (ClassNotFoundException e) {
-        throw new IOException(StringUtils.stringifyException(e));
+  private void initStorage(final NamespaceInfo nsInfo) throws IOException {
+    final FSDatasetInterface.Factory factory
+        = FSDatasetInterface.Factory.getFactory(conf);
+    
+    if (!factory.isSimulated()) {
+      final StartupOption startOpt = getStartupOption(conf);
+      if (startOpt == null) {
+        throw new IOException("Startup option not set.");
+      }
+      final String bpid = nsInfo.getBlockPoolID();
+      //read storage info, lock data dirs and transition fs state if necessary
+      storage.recoverTransitionRead(this, bpid, nsInfo, dataDirs, startOpt);
+      final StorageInfo bpStorage = storage.getBPStorage(bpid);
+      LOG.info("Setting up storage: nsid=" + bpStorage.getNamespaceID()
+          + ";bpid=" + bpid + ";lv=" + storage.getLayoutVersion()
+          + ";nsInfo=" + nsInfo);
+    }
+
+    synchronized(this)  {
+      if (data == null) {
+        data = factory.createFSDatasetInterface(this, storage, conf);
       }
-    } else {
-      data = new FSDataset(this, storage, conf);
     }
   }
 

hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/FSDataset.java

@@ -76,6 +76,16 @@ import org.apache.hadoop.util.ReflectionUtils;
  ***************************************************/
 @InterfaceAudience.Private
 class FSDataset implements FSDatasetInterface {
+  /**
+   * A factory for creating FSDataset objects.
+   */
+  static class Factory extends FSDatasetInterface.Factory {
+    @Override
+    public FSDatasetInterface createFSDatasetInterface(DataNode datanode,
+        DataStorage storage, Configuration conf) throws IOException {
+      return new FSDataset(datanode, storage, conf);
+    }
+  }
 
   /**
    * A node type that can be built into a tree reflecting the
@@ -1057,8 +1067,8 @@ class FSDataset implements FSDatasetInterface {
   /**
    * An FSDataset has a directory where it loads its data files.
    */
-  FSDataset(DataNode datanode, DataStorage storage, Configuration conf)
-      throws IOException {
+  private FSDataset(DataNode datanode, DataStorage storage, Configuration conf
+      ) throws IOException {
     this.datanode = datanode;
     this.maxBlocksPerDir = 
       conf.getInt(DFSConfigKeys.DFS_DATANODE_NUMBLOCKS_KEY,

hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/FSDatasetInterface.java

@@ -29,6 +29,7 @@ import java.util.Map;
 
 import org.apache.hadoop.classification.InterfaceAudience;
 import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.hdfs.DFSConfigKeys;
 import org.apache.hadoop.hdfs.protocol.Block;
 import org.apache.hadoop.hdfs.protocol.BlockListAsLongs;
 import org.apache.hadoop.hdfs.protocol.BlockLocalPathInfo;
@@ -38,6 +39,7 @@ import org.apache.hadoop.hdfs.server.protocol.BlockRecoveryCommand.RecoveringBlo
 import org.apache.hadoop.hdfs.server.protocol.ReplicaRecoveryInfo;
 import org.apache.hadoop.io.IOUtils;
 import org.apache.hadoop.util.DataChecksum;
+import org.apache.hadoop.util.ReflectionUtils;
 import org.apache.hadoop.util.DiskChecker.DiskErrorException;
 
 /**
@@ -49,6 +51,30 @@ import org.apache.hadoop.util.DiskChecker.DiskErrorException;
  */
 @InterfaceAudience.Private
 public interface FSDatasetInterface extends FSDatasetMBean {
+  /**
+   * A factory for creating FSDatasetInterface objects.
+   */
+  public abstract class Factory {
+    /** @return the configured factory. */
+    public static Factory getFactory(Configuration conf) {
+      final Class<? extends Factory> clazz = conf.getClass(
+          DFSConfigKeys.DFS_DATANODE_FSDATASET_FACTORY_KEY,
+          FSDataset.Factory.class,
+          Factory.class);
+      return ReflectionUtils.newInstance(clazz, conf);
+    }
+
+    /** Create a FSDatasetInterface object. */
+    public abstract FSDatasetInterface createFSDatasetInterface(
+        DataNode datanode, DataStorage storage, Configuration conf
+        ) throws IOException;
+
+    /** Does the factory create simulated objects? */
+    public boolean isSimulated() {
+      return false;
+    }
+  }
+
   /**
    * This is an interface for the underlying volume.
    * @see org.apache.hadoop.hdfs.server.datanode.FSDataset.FSVolume

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/DataNodeCluster.java

@@ -25,6 +25,7 @@ import org.apache.hadoop.conf.Configuration;
 import org.apache.hadoop.fs.FileSystem;
 import org.apache.hadoop.hdfs.protocol.Block;
 import org.apache.hadoop.hdfs.server.common.HdfsServerConstants.StartupOption;
+import org.apache.hadoop.hdfs.server.datanode.FSDatasetInterface;
 import org.apache.hadoop.hdfs.server.datanode.SimulatedFSDataset;
 import org.apache.hadoop.hdfs.server.namenode.CreateEditsLog;
 import org.apache.hadoop.net.DNS;
@@ -122,10 +123,9 @@ public class DataNodeCluster {
         }
         dataNodeDirs = args[i];
       } else if (args[i].equals("-simulated")) {
-        conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+        SimulatedFSDataset.setFactory(conf);
       } else if (args[i].equals("-inject")) {
-        if (!conf.getBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED,
-                                                                false) ) {
+        if (!FSDatasetInterface.Factory.getFactory(conf).isSimulated()) {
           System.out.print("-inject is valid only for simulated");
           printUsageExit(); 
         }
@@ -158,7 +158,7 @@ public class DataNodeCluster {
       System.exit(-1);
     }
     boolean simulated = 
-      conf.getBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, false);
+        FSDatasetInterface.Factory.getFactory(conf).isSimulated();
     System.out.println("Starting " + numDataNodes + 
           (simulated ? " Simulated " : " ") +
           " Data Nodes that will connect to Name Node at " + nameNodeAdr);

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/MiniDFSCluster.java

@@ -875,7 +875,7 @@ public class MiniDFSCluster {
         conf.set(DFSConfigKeys.DFS_DATANODE_DATA_DIR_KEY, dirs);
       }
       if (simulatedCapacities != null) {
-        dnConf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+        SimulatedFSDataset.setFactory(dnConf);
         dnConf.setLong(SimulatedFSDataset.CONFIG_PROPERTY_CAPACITY,
             simulatedCapacities[i-curDatanodesNum]);
       }

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/TestFileAppend.java

@@ -107,7 +107,7 @@ public class TestFileAppend{
   public void testCopyOnWrite() throws IOException {
     Configuration conf = new HdfsConfiguration();
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).build();
     FileSystem fs = cluster.getFileSystem();
@@ -178,7 +178,7 @@ public class TestFileAppend{
   public void testSimpleFlush() throws IOException {
     Configuration conf = new HdfsConfiguration();
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     fileContents = AppendTestUtil.initBuffer(AppendTestUtil.FILE_SIZE);
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).build();
@@ -234,7 +234,7 @@ public class TestFileAppend{
   public void testComplexFlush() throws IOException {
     Configuration conf = new HdfsConfiguration();
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     fileContents = AppendTestUtil.initBuffer(AppendTestUtil.FILE_SIZE);
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).build();
@@ -283,7 +283,7 @@ public class TestFileAppend{
   public void testFileNotFound() throws IOException {
     Configuration conf = new HdfsConfiguration();
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).build();
     FileSystem fs = cluster.getFileSystem();

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/TestFileAppend2.java

@@ -82,7 +82,7 @@ public class TestFileAppend2 extends TestCase {
   public void testSimpleAppend() throws IOException {
     final Configuration conf = new HdfsConfiguration();
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     conf.setInt(DFSConfigKeys.DFS_DATANODE_HANDLER_COUNT_KEY, 50);
     conf.setBoolean("dfs.support.append", true);

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/TestFileAppend4.java

@@ -77,7 +77,7 @@ public class TestFileAppend4 {
   public void setUp() throws Exception {
     this.conf = new Configuration();
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     conf.setBoolean(DFSConfigKeys.DFS_SUPPORT_APPEND_KEY, true);
 

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/TestFileCreation.java

@@ -144,7 +144,7 @@ public class TestFileCreation extends junit.framework.TestCase {
   public void testFileCreation() throws IOException {
     Configuration conf = new HdfsConfiguration();
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).build();
     FileSystem fs = cluster.getFileSystem();
@@ -223,7 +223,7 @@ public class TestFileCreation extends junit.framework.TestCase {
   public void testDeleteOnExit() throws IOException {
     Configuration conf = new HdfsConfiguration();
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).build();
     FileSystem fs = cluster.getFileSystem();
@@ -287,7 +287,7 @@ public class TestFileCreation extends junit.framework.TestCase {
     conf.setInt(DFS_NAMENODE_HEARTBEAT_RECHECK_INTERVAL_KEY, 1000);
     conf.setInt(DFS_HEARTBEAT_INTERVAL_KEY, 1);
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     // create cluster
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).build();
@@ -361,7 +361,7 @@ public class TestFileCreation extends junit.framework.TestCase {
     conf.setInt(DFS_NAMENODE_HEARTBEAT_RECHECK_INTERVAL_KEY, 1000);
     conf.setInt(DFS_HEARTBEAT_INTERVAL_KEY, 1);
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     // create cluster
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).build();
@@ -460,7 +460,7 @@ public class TestFileCreation extends junit.framework.TestCase {
     conf.setInt(DFS_NAMENODE_HEARTBEAT_RECHECK_INTERVAL_KEY, 1000);
     conf.setInt(DFS_HEARTBEAT_INTERVAL_KEY, 1);
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
 
     // create cluster
@@ -599,7 +599,7 @@ public class TestFileCreation extends junit.framework.TestCase {
     Configuration conf = new HdfsConfiguration();
     System.out.println("Testing adbornal client death.");
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).build();
     FileSystem fs = cluster.getFileSystem();
@@ -634,7 +634,7 @@ public class TestFileCreation extends junit.framework.TestCase {
   public void testFileCreationNonRecursive() throws IOException {
     Configuration conf = new HdfsConfiguration();
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).build();
     FileSystem fs = cluster.getFileSystem();

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/TestInjectionForSimulatedStorage.java

@@ -136,7 +136,7 @@ public class TestInjectionForSimulatedStorage extends TestCase {
       Configuration conf = new HdfsConfiguration();
       conf.set(DFSConfigKeys.DFS_REPLICATION_KEY, Integer.toString(numDataNodes));
       conf.setInt(DFSConfigKeys.DFS_BYTES_PER_CHECKSUM_KEY, checksumSize);
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
       //first time format
       cluster = new MiniDFSCluster.Builder(conf).numDataNodes(numDataNodes).build();
       cluster.waitActive();
@@ -159,7 +159,7 @@ public class TestInjectionForSimulatedStorage extends TestCase {
       
       LOG.info("Restarting minicluster");
       conf = new HdfsConfiguration();
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
       conf.set(DFSConfigKeys.DFS_NAMENODE_SAFEMODE_THRESHOLD_PCT_KEY, "0.0f"); 
       
       cluster = new MiniDFSCluster.Builder(conf)

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/TestLargeBlock.java

@@ -174,7 +174,7 @@ public class TestLargeBlock {
 
     Configuration conf = new Configuration();
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).build();
     FileSystem fs = cluster.getFileSystem();

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/TestPread.java

@@ -206,7 +206,7 @@ public class TestPread extends TestCase {
     conf.setLong(DFSConfigKeys.DFS_BLOCK_SIZE_KEY, 4096);
     conf.setLong(DFSConfigKeys.DFS_CLIENT_READ_PREFETCH_SIZE_KEY, 4096);
     if (simulatedStorage) {
-      conf.setBoolean("dfs.datanode.simulateddatastorage", true);
+      SimulatedFSDataset.setFactory(conf);
     }
     if (disableTransferTo) {
       conf.setBoolean("dfs.datanode.transferTo.allowed", false);

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/TestReplication.java

@@ -199,7 +199,7 @@ public class TestReplication extends TestCase {
     Configuration conf = new HdfsConfiguration();
     conf.setBoolean(DFSConfigKeys.DFS_NAMENODE_REPLICATION_CONSIDERLOAD_KEY, false);
     if (simulated) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf)
                                                .numDataNodes(numDatanodes)

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/TestSetrepIncreasing.java

@@ -28,7 +28,7 @@ public class TestSetrepIncreasing extends TestCase {
   static void setrep(int fromREP, int toREP, boolean simulatedStorage) throws IOException {
     Configuration conf = new HdfsConfiguration();
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     conf.set(DFSConfigKeys.DFS_REPLICATION_KEY, "" + fromREP);
     conf.setLong(DFSConfigKeys.DFS_BLOCKREPORT_INTERVAL_MSEC_KEY, 1000L);

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/TestShortCircuitLocalRead.java

@@ -124,7 +124,7 @@ public class TestShortCircuitLocalRead {
     conf.set(DFSConfigKeys.DFS_BLOCK_LOCAL_PATH_ACCESS_USER_KEY,
         UserGroupInformation.getCurrentUser().getShortUserName());
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).numDataNodes(1)
         .format(true).build();
@@ -248,7 +248,7 @@ public class TestShortCircuitLocalRead {
     conf.set(DFSConfigKeys.DFS_BLOCK_LOCAL_PATH_ACCESS_USER_KEY,
         UserGroupInformation.getCurrentUser().getShortUserName());
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).numDataNodes(1)
         .format(true).build();

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/TestSmallBlock.java

@@ -92,7 +92,7 @@ public class TestSmallBlock extends TestCase {
   public void testSmallBlock() throws IOException {
     Configuration conf = new HdfsConfiguration();
     if (simulatedStorage) {
-      conf.setBoolean("dfs.datanode.simulateddatastorage", true);
+      SimulatedFSDataset.setFactory(conf);
     }
     conf.set(DFSConfigKeys.DFS_BYTES_PER_CHECKSUM_KEY, "1");
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).build();

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/balancer/TestBalancer.java

@@ -77,7 +77,7 @@ public class TestBalancer extends TestCase {
     conf.setLong(DFSConfigKeys.DFS_BLOCK_SIZE_KEY, DEFAULT_BLOCK_SIZE);
     conf.setInt(DFSConfigKeys.DFS_BYTES_PER_CHECKSUM_KEY, DEFAULT_BLOCK_SIZE);
     conf.setLong(DFSConfigKeys.DFS_HEARTBEAT_INTERVAL_KEY, 1L);
-    conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+    SimulatedFSDataset.setFactory(conf);
     conf.setLong(DFSConfigKeys.DFS_BALANCER_MOVEDWINWIDTH_KEY, 2000L);
   }
 

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/datanode/SimulatedFSDataset.java

@@ -22,7 +22,6 @@ import java.io.IOException;
 import java.io.InputStream;
 import java.io.OutputStream;
 import java.util.ArrayList;
-import java.util.Arrays;
 import java.util.HashMap;
 import java.util.List;
 import java.util.Map;
@@ -32,7 +31,6 @@ import javax.management.NotCompliantMBeanException;
 import javax.management.ObjectName;
 import javax.management.StandardMBean;
 
-import org.apache.hadoop.conf.Configurable;
 import org.apache.hadoop.conf.Configuration;
 import org.apache.hadoop.hdfs.DFSConfigKeys;
 import org.apache.hadoop.hdfs.protocol.Block;
@@ -63,21 +61,33 @@ import org.apache.hadoop.util.DiskChecker.DiskErrorException;
  * 
  * Note the synchronization is coarse grained - it is at each method. 
  */
+public class SimulatedFSDataset implements FSDatasetInterface {
+  static class Factory extends FSDatasetInterface.Factory {
+    @Override
+    public FSDatasetInterface createFSDatasetInterface(DataNode datanode,
+        DataStorage storage, Configuration conf) throws IOException {
+      return new SimulatedFSDataset(datanode, storage, conf);
+    }
 
-public class SimulatedFSDataset  implements FSDatasetInterface, Configurable{
+    @Override
+    public boolean isSimulated() {
+      return true;
+    }
+  }
+  
+  public static void setFactory(Configuration conf) {
+    conf.set(DFSConfigKeys.DFS_DATANODE_FSDATASET_FACTORY_KEY,
+        Factory.class.getName());
+  }
   
-  public static final String CONFIG_PROPERTY_SIMULATED =
-                                    "dfs.datanode.simulateddatastorage";
   public static final String CONFIG_PROPERTY_CAPACITY =
-                            "dfs.datanode.simulateddatastorage.capacity";
+      "dfs.datanode.simulateddatastorage.capacity";
   
   public static final long DEFAULT_CAPACITY = 2L<<40; // 1 terabyte
-  public static final byte DEFAULT_DATABYTE = 9; // 1 terabyte
-  byte simulatedDataByte = DEFAULT_DATABYTE;
-  Configuration conf = null;
+  public static final byte DEFAULT_DATABYTE = 9;
   
-  static byte[] nullCrcFileData;
-  {
+  static final byte[] nullCrcFileData;
+  static {
     DataChecksum checksum = DataChecksum.newDataChecksum( DataChecksum.
                               CHECKSUM_NULL, 16*1024 );
     byte[] nullCrcHeader = checksum.getHeader();
@@ -360,31 +370,22 @@ public class SimulatedFSDataset  implements FSDatasetInterface, Configurable{
     }
   }
   
-  private Map<String, Map<Block, BInfo>> blockMap = null;
-  private SimulatedStorage storage = null;
-  private String storageId;
-  
-  public SimulatedFSDataset(Configuration conf) throws IOException {
-    setConf(conf);
-  }
+  private final Map<String, Map<Block, BInfo>> blockMap
+      = new HashMap<String, Map<Block,BInfo>>();
+  private final SimulatedStorage storage;
+  private final String storageId;
   
-  // Constructor used for constructing the object using reflection
-  @SuppressWarnings("unused")
-  private SimulatedFSDataset() { // real construction when setConf called..
-  }
-  
-  public Configuration getConf() {
-    return conf;
-  }
-
-  public void setConf(Configuration iconf)  {
-    conf = iconf;
-    storageId = conf.get(DFSConfigKeys.DFS_DATANODE_STORAGEID_KEY, "unknownStorageId" +
-                                        new Random().nextInt());
+  public SimulatedFSDataset(DataNode datanode, DataStorage storage,
+      Configuration conf) {
+    if (storage != null) {
+      storage.createStorageID(datanode.getPort());
+      this.storageId = storage.getStorageID();
+    } else {
+      this.storageId = "unknownStorageId" + new Random().nextInt();
+    }
     registerMBean(storageId);
-    storage = new SimulatedStorage(
+    this.storage = new SimulatedStorage(
         conf.getLong(CONFIG_PROPERTY_CAPACITY, DEFAULT_CAPACITY));
-    blockMap = new HashMap<String, Map<Block,BInfo>>(); 
   }
 
   public synchronized void injectBlocks(String bpid,
@@ -441,23 +442,16 @@ public class SimulatedFSDataset  implements FSDatasetInterface, Configurable{
 
   @Override
   public synchronized BlockListAsLongs getBlockReport(String bpid) {
+    final List<Block> blocks = new ArrayList<Block>();
     final Map<Block, BInfo> map = blockMap.get(bpid);
-    Block[] blockTable = new Block[map.size()];
     if (map != null) {
-      int count = 0;
       for (BInfo b : map.values()) {
         if (b.isFinalized()) {
-          blockTable[count++] = b.theBlock;
+          blocks.add(b.theBlock);
         }
       }
-      if (count != blockTable.length) {
-        blockTable = Arrays.copyOf(blockTable, count);
-      }
-    } else {
-      blockTable = new Block[0];
     }
-    return new BlockListAsLongs(
-        new ArrayList<Block>(Arrays.asList(blockTable)), null);
+    return new BlockListAsLongs(blocks, null);
   }
 
   @Override // FSDatasetMBean

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/datanode/TestDataNodeMetrics.java

@@ -34,7 +34,7 @@ public class TestDataNodeMetrics extends TestCase {
   
   public void testDataNodeMetrics() throws Exception {
     Configuration conf = new HdfsConfiguration();
-    conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+    SimulatedFSDataset.setFactory(conf);
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).build();
     try {
       FileSystem fs = cluster.getFileSystem();

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/datanode/TestSimulatedFSDataset.java

@@ -44,8 +44,8 @@ public class TestSimulatedFSDataset extends TestCase {
 
   protected void setUp() throws Exception {
     super.setUp();
-      conf = new HdfsConfiguration();
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+    conf = new HdfsConfiguration();
+    SimulatedFSDataset.setFactory(conf);
   }
 
   protected void tearDown() throws Exception {
@@ -86,6 +86,18 @@ public class TestSimulatedFSDataset extends TestCase {
   int addSomeBlocks(FSDatasetInterface fsdataset ) throws IOException {
     return addSomeBlocks(fsdataset, 1);
   }
+  
+  public void testFSDatasetFactory() {
+    final Configuration conf = new Configuration();
+    FSDatasetInterface.Factory f = FSDatasetInterface.Factory.getFactory(conf);
+    assertEquals(FSDataset.Factory.class, f.getClass());
+    assertFalse(f.isSimulated());
+
+    SimulatedFSDataset.setFactory(conf);
+    FSDatasetInterface.Factory s = FSDatasetInterface.Factory.getFactory(conf);
+    assertEquals(SimulatedFSDataset.Factory.class, s.getClass());
+    assertTrue(s.isSimulated());
+  }
 
   public void testGetMetaData() throws IOException {
     FSDatasetInterface fsdataset = getSimulatedFSDataset(); 
@@ -287,8 +299,8 @@ public class TestSimulatedFSDataset extends TestCase {
     }
   }
   
-  private SimulatedFSDataset getSimulatedFSDataset() throws IOException {
-    SimulatedFSDataset fsdataset = new SimulatedFSDataset(conf); 
+  private SimulatedFSDataset getSimulatedFSDataset() {
+    SimulatedFSDataset fsdataset = new SimulatedFSDataset(null, null, conf); 
     fsdataset.addBlockPool(bpid, conf);
     return fsdataset;
   }

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/namenode/TestFileLimit.java

@@ -83,7 +83,7 @@ public class TestFileLimit extends TestCase {
     int currentNodes = 0;
     
     if (simulatedStorage) {
-      conf.setBoolean(SimulatedFSDataset.CONFIG_PROPERTY_SIMULATED, true);
+      SimulatedFSDataset.setFactory(conf);
     }
     MiniDFSCluster cluster = new MiniDFSCluster.Builder(conf).build();
     FileSystem fs = cluster.getFileSystem();