org.apache.hadoop.hdfs.server.blockmanagement
Class BlockPlacementPolicyWithNodeGroup

java.lang.Object
  org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicy
      org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicyDefault
          org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicyWithNodeGroup

public class BlockPlacementPolicyWithNodeGroup
extends org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicyDefault

The class is responsible for choosing the desired number of targets for placing block replicas on environment with node-group layer. The replica placement strategy is adjusted to: If the writer is on a datanode, the 1st replica is placed on the local node (or local node-group), otherwise a random datanode. The 2nd replica is placed on a datanode that is on a different rack with 1st replica node. The 3rd replica is placed on a datanode which is on a different node-group but the same rack as the second replica node.

Field Summary

Fields inherited from class org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicyDefault

clusterMap, considerLoad, heartbeatInterval, threadLocalBuilder, tolerateHeartbeatMultiplier

Method Summary

protected int	addToExcludedNodes(org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor localMachine, HashMap<org.apache.hadoop.net.Node,org.apache.hadoop.net.Node> excludedNodes) Find other nodes in the same nodegroup of localMachine and add them into excludeNodes as replica should not be duplicated for nodes within the same nodegroup
protected void	adjustExcludedNodes(HashMap<org.apache.hadoop.net.Node,org.apache.hadoop.net.Node> excludedNodes, org.apache.hadoop.net.Node chosenNode) After choosing a node to place replica, adjust excluded nodes accordingly.
protected org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor	chooseLocalNode(org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor localMachine, HashMap<org.apache.hadoop.net.Node,org.apache.hadoop.net.Node> excludedNodes, long blocksize, int maxNodesPerRack, List<org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor> results, boolean avoidStaleNodes) choose local node of localMachine as the target.
protected org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor	chooseLocalRack(org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor localMachine, HashMap<org.apache.hadoop.net.Node,org.apache.hadoop.net.Node> excludedNodes, long blocksize, int maxNodesPerRack, List<org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor> results, boolean avoidStaleNodes)
protected void	chooseRemoteRack(int numOfReplicas, org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor localMachine, HashMap<org.apache.hadoop.net.Node,org.apache.hadoop.net.Node> excludedNodes, long blocksize, int maxReplicasPerRack, List<org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor> results, boolean avoidStaleNodes)
protected String	getRack(org.apache.hadoop.hdfs.protocol.DatanodeInfo cur) Get rack string from a data node
void	initialize(org.apache.hadoop.conf.Configuration conf, org.apache.hadoop.hdfs.server.namenode.FSClusterStats stats, org.apache.hadoop.net.NetworkTopology clusterMap) Used to setup a BlockPlacementPolicy object.
Iterator<org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor>	pickupReplicaSet(Collection<org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor> first, Collection<org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor> second) Pick up replica node set for deleting replica as over-replicated.

Methods inherited from class org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicyDefault

chooseRandom, chooseRandom, chooseReplicaToDelete, chooseTarget, chooseTarget, isGoodTarget, verifyBlockPlacement

Methods inherited from class org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicy

adjustSetsWithChosenReplica, getInstance, splitNodesWithRack

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Detail

initialize

public void initialize(org.apache.hadoop.conf.Configuration conf,
                       org.apache.hadoop.hdfs.server.namenode.FSClusterStats stats,
                       org.apache.hadoop.net.NetworkTopology clusterMap)

Description copied from class: org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicy

Used to setup a BlockPlacementPolicy object. This should be defined by all implementations of a BlockPlacementPolicy.

Overrides:

initialize in class org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicyDefault

Parameters:

conf - the configuration object

stats - retrieve cluster status from here

clusterMap - cluster topology

chooseLocalNode

protected org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor chooseLocalNode(org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor localMachine,
                                                                                           HashMap<org.apache.hadoop.net.Node,org.apache.hadoop.net.Node> excludedNodes,
                                                                                           long blocksize,
                                                                                           int maxNodesPerRack,
                                                                                           List<org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor> results,
                                                                                           boolean avoidStaleNodes)
                                                                                    throws org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicy.NotEnoughReplicasException

choose local node of localMachine as the target. if localMachine is not available, choose a node on the same nodegroup or rack instead.

Overrides:

chooseLocalNode in class org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicyDefault

Returns:

the chosen node

Throws:

org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicy.NotEnoughReplicasException

adjustExcludedNodes

protected void adjustExcludedNodes(HashMap<org.apache.hadoop.net.Node,org.apache.hadoop.net.Node> excludedNodes,
                                   org.apache.hadoop.net.Node chosenNode)

After choosing a node to place replica, adjust excluded nodes accordingly. It should do nothing here as chosenNode is already put into exlcudeNodes, but it can be overridden in subclass to put more related nodes into excludedNodes.

Overrides:

adjustExcludedNodes in class org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicyDefault

chooseLocalRack

protected org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor chooseLocalRack(org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor localMachine,
                                                                                           HashMap<org.apache.hadoop.net.Node,org.apache.hadoop.net.Node> excludedNodes,
                                                                                           long blocksize,
                                                                                           int maxNodesPerRack,
                                                                                           List<org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor> results,
                                                                                           boolean avoidStaleNodes)
                                                                                    throws org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicy.NotEnoughReplicasException

Overrides:

chooseLocalRack in class org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicyDefault

Throws:

org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicy.NotEnoughReplicasException

chooseRemoteRack

protected void chooseRemoteRack(int numOfReplicas,
                                org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor localMachine,
                                HashMap<org.apache.hadoop.net.Node,org.apache.hadoop.net.Node> excludedNodes,
                                long blocksize,
                                int maxReplicasPerRack,
                                List<org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor> results,
                                boolean avoidStaleNodes)
                         throws org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicy.NotEnoughReplicasException

Overrides:

chooseRemoteRack in class org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicyDefault

Throws:

org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicy.NotEnoughReplicasException

getRack

protected String getRack(org.apache.hadoop.hdfs.protocol.DatanodeInfo cur)

Description copied from class: org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicy

Get rack string from a data node

Overrides:

getRack in class org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicy

Returns:

rack of data node

addToExcludedNodes

protected int addToExcludedNodes(org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor localMachine,
                                 HashMap<org.apache.hadoop.net.Node,org.apache.hadoop.net.Node> excludedNodes)

Find other nodes in the same nodegroup of localMachine and add them into excludeNodes as replica should not be duplicated for nodes within the same nodegroup

Overrides:

addToExcludedNodes in class org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicyDefault

Returns:

number of new excluded nodes

pickupReplicaSet

public Iterator<org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor> pickupReplicaSet(Collection<org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor> first,
                                                                                                   Collection<org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor> second)

Pick up replica node set for deleting replica as over-replicated. First set contains replica nodes on rack with more than one replica while second set contains remaining replica nodes. If first is not empty, divide first set into two subsets: moreThanOne contains nodes on nodegroup with more than one replica exactlyOne contains the remaining nodes in first set then pickup priSet if not empty. If first is empty, then pick second.

Overrides:

pickupReplicaSet in class org.apache.hadoop.hdfs.server.blockmanagement.BlockPlacementPolicyDefault