HDFS-6710. Change BlockPlacementPolicy to consider block storage policy in replica deletion.

git-svn-id: https://svn.apache.org/repos/asf/hadoop/common/branches/HDFS-6584@1612265 13f79535-47bb-0310-9956-ffa450edef68

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

@@ -11,6 +11,9 @@ HDFS-6584: Archival Storage
     HDFS-6671. Change BlockPlacementPolicy to consider block storage policy
     in replicaiton.  (szetszwo)
 
+    HDFS-6710. Change BlockPlacementPolicy to consider block storage policy
+    in replica deletion.  (szetszwo)
+
 Trunk (Unreleased)
 
   INCOMPATIBLE CHANGES

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

@@ -118,15 +118,40 @@ public class BlockStoragePolicy {
   public List<StorageType> chooseStorageTypes(final short replication,
       final Iterable<StorageType> chosen) {
     final List<StorageType> types = chooseStorageTypes(replication);
+    diff(types, chosen, null);
+    return types;
+  }
 
-    //remove the chosen storage types
+  /**
-    for(StorageType c : chosen) {
+   * Compute the list difference t = t - c.
-      final int i = types.indexOf(c);
+   * Further, if e is not null, set e = e + c - t;
+   */
+  private static void diff(List<StorageType> t, Iterable<StorageType> c,
+      List<StorageType> e) {
+    for(StorageType storagetype : c) {
+      final int i = t.indexOf(storagetype);
       if (i >= 0) {
-        types.remove(i);
+        t.remove(i);
+      } else if (e != null) {
+        e.add(storagetype);
       }
     }
-    return types;
+  }
+
+  /**
+   * Choose excess storage types for deletion, given the
+   * replication number and the storage types of the chosen replicas.
+   *
+   * @param replication the replication number.
+   * @param chosen the storage types of the chosen replicas.
+   * @return a list of {@link StorageType}s for deletion.
+   */
+  public List<StorageType> chooseExcess(final short replication,
+      final Iterable<StorageType> chosen) {
+    final List<StorageType> types = chooseStorageTypes(replication);
+    final List<StorageType> excess = new LinkedList<StorageType>();
+    diff(types, chosen, excess);
+    return excess;
   }
 
   /** @return the fallback {@link StorageType} for creation. */
@@ -264,5 +289,5 @@ public class BlockStoragePolicy {
     throw new IllegalArgumentException(message + " in "
         + DFS_BLOCK_STORAGE_POLICIES_KEY + " \""
         + conf.get(DFS_BLOCK_STORAGE_POLICIES_KEY) + "\".");
- }
+  }
 }

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

@@ -46,6 +46,7 @@ import org.apache.hadoop.hdfs.BlockStoragePolicy;
 import org.apache.hadoop.hdfs.DFSConfigKeys;
 import org.apache.hadoop.hdfs.DFSUtil;
 import org.apache.hadoop.hdfs.HAUtil;
+import org.apache.hadoop.hdfs.StorageType;
 import org.apache.hadoop.hdfs.protocol.Block;
 import org.apache.hadoop.hdfs.protocol.BlockListAsLongs;
 import org.apache.hadoop.hdfs.protocol.BlockListAsLongs.BlockReportIterator;
@@ -2712,6 +2713,10 @@ public class BlockManager {
     assert namesystem.hasWriteLock();
     // first form a rack to datanodes map and
     BlockCollection bc = getBlockCollection(b);
+    final BlockStoragePolicy storagePolicy = storagePolicySuite.getPolicy(bc.getStoragePolicyID());
+    final List<StorageType> excessTypes = storagePolicy.chooseExcess(
+        replication, DatanodeStorageInfo.toStorageTypes(nonExcess));
+
 
     final Map<String, List<DatanodeStorageInfo>> rackMap
         = new HashMap<String, List<DatanodeStorageInfo>>();
@@ -2741,7 +2746,7 @@ public class BlockManager {
         cur = delNodeHintStorage;
       } else { // regular excessive replica removal
         cur = replicator.chooseReplicaToDelete(bc, b, replication,
-        		moreThanOne, exactlyOne);
+            moreThanOne, exactlyOne, excessTypes);
       }
       firstOne = false;
 

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

@@ -119,18 +119,21 @@ public abstract class BlockPlacementPolicy {
    * @param srcBC block collection of file to which block-to-be-deleted belongs
    * @param block The block to be deleted
    * @param replicationFactor The required number of replicas for this block
-   * @param existingReplicas The replica locations of this block that are present
+   * @param moreThanOne The replica locations of this block that are present
-                  on at least two unique racks. 
+   *                    on more than one unique racks.
-   * @param moreExistingReplicas Replica locations of this block that are not
+   * @param exactlyOne Replica locations of this block that  are present
-                   listed in the previous parameter.
+   *                    on exactly one unique racks.
+   * @param excessTypes The excess {@link StorageType}s according to the
+   *                    {@link BlockStoragePolicy}.
    * @return the replica that is the best candidate for deletion
    */
   abstract public DatanodeStorageInfo chooseReplicaToDelete(
       BlockCollection srcBC,
       Block block, 
       short replicationFactor,
-      Collection<DatanodeStorageInfo> existingReplicas,
+      Collection<DatanodeStorageInfo> moreThanOne,
-      Collection<DatanodeStorageInfo> moreExistingReplicas);
+      Collection<DatanodeStorageInfo> exactlyOne,
+      List<StorageType> excessTypes);
 
   /**
    * Used to setup a BlockPlacementPolicy object. This should be defined by 

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

@@ -22,7 +22,6 @@ import static org.apache.hadoop.util.Time.now;
 import java.util.ArrayList;
 import java.util.Collection;
 import java.util.HashSet;
-import java.util.Iterator;
 import java.util.List;
 import java.util.Set;
 import java.util.TreeSet;
@@ -273,8 +272,8 @@ public class BlockPlacementPolicyDefault extends BlockPlacementPolicy {
     // Keep a copy of original excludedNodes
     final Set<Node> oldExcludedNodes = avoidStaleNodes ? 
         new HashSet<Node>(excludedNodes) : null;
-    final List<StorageType> storageTypes = chooseStorageTypes(storagePolicy,
+    final List<StorageType> storageTypes = storagePolicy.chooseStorageTypes(
-        (short)totalReplicasExpected, results); 
+        (short)totalReplicasExpected, DatanodeStorageInfo.toStorageTypes(results));
     try {
       if (numOfResults == 0) {
         writer = chooseLocalStorage(writer, excludedNodes, blocksize,
@@ -671,28 +670,6 @@ public class BlockPlacementPolicyDefault extends BlockPlacementPolicy {
     }
     return true;
   }
-  
-  private static List<StorageType> chooseStorageTypes(
-      final BlockStoragePolicy storagePolicy, final short replication,
-      final Iterable<DatanodeStorageInfo> chosen) {
-    return storagePolicy.chooseStorageTypes(
-        replication, new Iterable<StorageType>() {
-          @Override
-          public Iterator<StorageType> iterator() {
-            return new Iterator<StorageType>() {
-              final Iterator<DatanodeStorageInfo> i = chosen.iterator();
-              @Override
-              public boolean hasNext() {return i.hasNext();}
-              @Override
-              public StorageType next() {return i.next().getStorageType();}
-              @Override
-              public void remove() {
-                throw new UnsupportedOperationException();
-              }
-            };
-          }
-        });
-  }
 
   /**
    * Return a pipeline of nodes.
@@ -759,7 +736,8 @@ public class BlockPlacementPolicyDefault extends BlockPlacementPolicy {
   public DatanodeStorageInfo chooseReplicaToDelete(BlockCollection bc,
       Block block, short replicationFactor,
       Collection<DatanodeStorageInfo> first,
-      Collection<DatanodeStorageInfo> second) {
+      Collection<DatanodeStorageInfo> second,
+      final List<StorageType> excessTypes) {
     long oldestHeartbeat =
       now() - heartbeatInterval * tolerateHeartbeatMultiplier;
     DatanodeStorageInfo oldestHeartbeatStorage = null;
@@ -769,6 +747,10 @@ public class BlockPlacementPolicyDefault extends BlockPlacementPolicy {
     // Pick the node with the oldest heartbeat or with the least free space,
     // if all hearbeats are within the tolerable heartbeat interval
     for(DatanodeStorageInfo storage : pickupReplicaSet(first, second)) {
+      if (!excessTypes.contains(storage.getStorageType())) {
+        continue;
+      }
+
       final DatanodeDescriptor node = storage.getDatanodeDescriptor();
       long free = node.getRemaining();
       long lastHeartbeat = node.getLastUpdate();
@@ -781,9 +763,10 @@ public class BlockPlacementPolicyDefault extends BlockPlacementPolicy {
         minSpaceStorage = storage;
       }
     }
-
+    final DatanodeStorageInfo storage = oldestHeartbeatStorage != null?
-    return oldestHeartbeatStorage != null? oldestHeartbeatStorage
+        oldestHeartbeatStorage : minSpaceStorage;
-        : minSpaceStorage;
+    excessTypes.remove(storage.getStorageType());
+    return storage;
   }
 
   /**

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

@@ -292,6 +292,26 @@ public class DatanodeStorageInfo {
     return "[" + storageType + "]" + storageID + ":" + state;
   }
 
+  static Iterable<StorageType> toStorageTypes(
+      final Iterable<DatanodeStorageInfo> infos) {
+    return new Iterable<StorageType>() {
+        @Override
+        public Iterator<StorageType> iterator() {
+          return new Iterator<StorageType>() {
+            final Iterator<DatanodeStorageInfo> i = infos.iterator();
+            @Override
+            public boolean hasNext() {return i.hasNext();}
+            @Override
+            public StorageType next() {return i.next().getStorageType();}
+            @Override
+            public void remove() {
+              throw new UnsupportedOperationException();
+            }
+          };
+        }
+      };
+  }
+
   /** @return the first {@link DatanodeStorageInfo} corresponding to
    *          the given datanode
    */

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/blockmanagement/TestReplicationPolicy.java

@@ -46,6 +46,7 @@ import org.apache.hadoop.hdfs.DFSUtil;
 import org.apache.hadoop.hdfs.HdfsConfiguration;
 import org.apache.hadoop.hdfs.LogVerificationAppender;
 import org.apache.hadoop.hdfs.MiniDFSCluster;
+import org.apache.hadoop.hdfs.StorageType;
 import org.apache.hadoop.hdfs.TestBlockStoragePolicy;
 import org.apache.hadoop.hdfs.protocol.Block;
 import org.apache.hadoop.hdfs.protocol.HdfsConstants;
@@ -933,8 +934,10 @@ public class TestReplicationPolicy {
     // replica nodes, while storages[2] and dataNodes[5] are in second set.
     assertEquals(2, first.size());
     assertEquals(2, second.size());
+    List<StorageType> excessTypes = new ArrayList<StorageType>();
+    excessTypes.add(StorageType.DEFAULT);
     DatanodeStorageInfo chosen = replicator.chooseReplicaToDelete(
-        null, null, (short)3, first, second);
+        null, null, (short)3, first, second, excessTypes);
     // Within first set, storages[1] with less free space
     assertEquals(chosen, storages[1]);
 
@@ -942,8 +945,9 @@ public class TestReplicationPolicy {
     assertEquals(0, first.size());
     assertEquals(3, second.size());
     // Within second set, storages[5] with less free space
+    excessTypes.add(StorageType.DEFAULT);
     chosen = replicator.chooseReplicaToDelete(
-        null, null, (short)2, first, second);
+        null, null, (short)2, first, second, excessTypes);
     assertEquals(chosen, storages[5]);
   }
   

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/blockmanagement/TestReplicationPolicyWithNodeGroup.java

@@ -36,6 +36,7 @@ import org.apache.hadoop.fs.FileSystem;
 import org.apache.hadoop.hdfs.DFSConfigKeys;
 import org.apache.hadoop.hdfs.DFSTestUtil;
 import org.apache.hadoop.hdfs.HdfsConfiguration;
+import org.apache.hadoop.hdfs.StorageType;
 import org.apache.hadoop.hdfs.TestBlockStoragePolicy;
 import org.apache.hadoop.hdfs.protocol.HdfsConstants;
 import org.apache.hadoop.hdfs.server.namenode.NameNode;
@@ -612,8 +613,10 @@ public class TestReplicationPolicyWithNodeGroup {
         replicaList, rackMap, first, second);
     assertEquals(3, first.size());
     assertEquals(1, second.size());
+    List<StorageType> excessTypes = new ArrayList<StorageType>();
+    excessTypes.add(StorageType.DEFAULT);
     DatanodeStorageInfo chosen = replicator.chooseReplicaToDelete(
-        null, null, (short)3, first, second);
+        null, null, (short)3, first, second, excessTypes);
     // Within first set {dataNodes[0], dataNodes[1], dataNodes[2]}, 
     // dataNodes[0] and dataNodes[1] are in the same nodegroup, 
     // but dataNodes[1] is chosen as less free space
@@ -624,16 +627,18 @@ public class TestReplicationPolicyWithNodeGroup {
     assertEquals(1, second.size());
     // Within first set {dataNodes[0], dataNodes[2]}, dataNodes[2] is chosen
     // as less free space
+    excessTypes.add(StorageType.DEFAULT);
     chosen = replicator.chooseReplicaToDelete(
-        null, null, (short)2, first, second);
+        null, null, (short)2, first, second, excessTypes);
     assertEquals(chosen, storages[2]);
 
     replicator.adjustSetsWithChosenReplica(rackMap, first, second, chosen);
     assertEquals(0, first.size());
     assertEquals(2, second.size());
     // Within second set, dataNodes[5] with less free space
+    excessTypes.add(StorageType.DEFAULT);
     chosen = replicator.chooseReplicaToDelete(
-        null, null, (short)1, first, second);
+        null, null, (short)1, first, second, excessTypes);
     assertEquals(chosen, storages[5]);
   }
   

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

@@ -38,6 +38,7 @@ import org.apache.hadoop.hdfs.DFSTestUtil;
 import org.apache.hadoop.hdfs.DFSUtil;
 import org.apache.hadoop.hdfs.MiniDFSCluster;
 import org.apache.hadoop.hdfs.MiniDFSNNTopology;
+import org.apache.hadoop.hdfs.StorageType;
 import org.apache.hadoop.hdfs.protocol.Block;
 import org.apache.hadoop.hdfs.protocol.ExtendedBlock;
 import org.apache.hadoop.hdfs.protocolPB.DatanodeProtocolClientSideTranslatorPB;
@@ -588,7 +589,8 @@ public class TestDNFencing {
     public DatanodeStorageInfo chooseReplicaToDelete(BlockCollection inode,
         Block block, short replicationFactor,
         Collection<DatanodeStorageInfo> first,
-        Collection<DatanodeStorageInfo> second) {
+        Collection<DatanodeStorageInfo> second,
+        List<StorageType> excessTypes) {
       
       Collection<DatanodeStorageInfo> chooseFrom = !first.isEmpty() ? first : second;