org.gridgain.grid.cache.affinity.consistenthash

Class GridCacheConsistentHashAffinityFunction

java.lang.Object

org.gridgain.grid.cache.affinity.consistenthash.GridCacheConsistentHashAffinityFunction

All Implemented Interfaces:

Serializable, GridCacheAffinityFunction

public class GridCacheConsistentHashAffinityFunction
extends Object
implements GridCacheAffinityFunction

Affinity function for partitioned cache. This function supports the following configuration:

• backups - Use this flag to control how many back up nodes will be assigned to every key. The default value is 0.
• replicas - Generally the more replicas a node gets, the more key assignments it will receive. You can configure different number of replicas for a node by setting user attribute with name getReplicaCountAttributeName() to some number. Default value is 512 defined by DFLT_REPLICA_COUNT constant.
• backupFilter - Optional filter for back up nodes. If provided, then only nodes that pass this filter will be selected as backup nodes. If not provided, then primary and backup nodes will be selected out of all nodes available for this cache.

Cache affinity can be configured for individual caches via GridCacheConfiguration.getAffinity() method.

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
static int	DFLT_PARTITION_COUNT Default number of partitions.
static int	DFLT_REPLICA_COUNT Default replica count for partitioned caches.
static String	DFLT_REPLICA_COUNT_ATTR_NAME Name of node attribute to specify number of replicas for a node.

Constructor Summary

Constructors

Constructor and Description
GridCacheConsistentHashAffinityFunction() Empty constructor with all defaults.
GridCacheConsistentHashAffinityFunction(boolean exclNeighbors) Initializes affinity with flag to exclude same-host-neighbors from being backups of each other and specified number of backups.
GridCacheConsistentHashAffinityFunction(boolean exclNeighbors, int parts) Initializes affinity with flag to exclude same-host-neighbors from being backups of each other, and specified number of backups and partitions.
GridCacheConsistentHashAffinityFunction(int parts, GridBiPredicate<GridNode,GridNode> backupFilter) Initializes optional counts for replicas and backups.

Method Summary

Methods

Modifier and Type	Method and Description
List<List<GridNode>>	assignPartitions(GridCacheAffinityFunctionContext ctx) Gets affinity nodes for a partition.
GridBiPredicate<GridNode,GridNode>	getBackupFilter() Gets optional backup filter.
int	getDefaultReplicas() Gets default count of virtual replicas in consistent hash ring.
GridCacheAffinityNodeHashResolver	getHashIdResolver() Gets hash ID resolver for nodes.
int	getPartitions() Gets total number of key partitions.
String	getReplicaCountAttributeName() Gets optional attribute name for replica count.
boolean	isExcludeNeighbors() Checks flag to exclude same-host-neighbors from being backups of each other (default is false).
Collection<GridNode>	nodes(int part, Collection<GridNode> nodes, int backups) Assigns nodes to one partition.
int	partition(Object key) Gets partition number for a given key starting from 0.
int	partitions() Gets total number of partitions available.
void	removeNode(UUID nodeId) Removes node from affinity.
void	reset() Resets cache affinity to its initial state.
void	setBackupFilter(GridBiPredicate<GridNode,GridNode> backupFilter) Sets optional backup filter.
void	setDefaultReplicas(int replicas) Sets default count of virtual replicas in consistent hash ring.
void	setExcludeNeighbors(boolean exclNeighbors) Sets flag to exclude same-host-neighbors from being backups of each other (default is false).
void	setHashIdResolver(GridCacheAffinityNodeHashResolver hashIdRslvr) Sets hash ID resolver for nodes.
void	setPartitions(int parts) Sets total number of partitions.
void	setReplicaCountAttributeName(String attrName) Sets optional attribute name for replica count.
String	toString()

Methods inherited from class java.lang.Object

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

Field Detail

DFLT_PARTITION_COUNT

public static final int DFLT_PARTITION_COUNT

Default number of partitions.

See Also:

Constant Field Values

DFLT_REPLICA_COUNT

public static final int DFLT_REPLICA_COUNT

Default replica count for partitioned caches.

See Also:

Constant Field Values

DFLT_REPLICA_COUNT_ATTR_NAME

public static final String DFLT_REPLICA_COUNT_ATTR_NAME

Name of node attribute to specify number of replicas for a node. Default value is gg:affinity:node:replicas.

See Also:

Constant Field Values

Constructor Detail

GridCacheConsistentHashAffinityFunction

public GridCacheConsistentHashAffinityFunction()

Empty constructor with all defaults.

GridCacheConsistentHashAffinityFunction

public GridCacheConsistentHashAffinityFunction(boolean exclNeighbors)

Initializes affinity with flag to exclude same-host-neighbors from being backups of each other and specified number of backups.

Note that excludeNeighbors parameter is ignored if #getBackupFilter() is set.

Parameters:

exclNeighbors - True if nodes residing on the same host may not act as backups of each other.

GridCacheConsistentHashAffinityFunction

public GridCacheConsistentHashAffinityFunction(boolean exclNeighbors,
                                       int parts)

Initializes affinity with flag to exclude same-host-neighbors from being backups of each other, and specified number of backups and partitions.

Note that excludeNeighbors parameter is ignored if #getBackupFilter() is set.

Parameters:

exclNeighbors - True if nodes residing on the same host may not act as backups of each other.

parts - Total number of partitions.

GridCacheConsistentHashAffinityFunction

public GridCacheConsistentHashAffinityFunction(int parts,
                                       @Nullable
                                       GridBiPredicate<GridNode,GridNode> backupFilter)

Initializes optional counts for replicas and backups.

Note that excludeNeighbors parameter is ignored if backupFilter is set.

Parameters:

parts - Total number of partitions.

backupFilter - Optional back up filter for nodes. If provided, backups will be selected from all nodes that pass this filter. First argument for this filter is primary node, and second argument is node being tested.

Note that excludeNeighbors parameter is ignored if backupFilter is set.

Method Detail

getDefaultReplicas

public int getDefaultReplicas()

Gets default count of virtual replicas in consistent hash ring.

To determine node replicas, node attribute with getReplicaCountAttributeName() name will be checked first. If it is absent, then this value will be used.

Returns:

Count of virtual replicas in consistent hash ring.

setDefaultReplicas

public void setDefaultReplicas(int replicas)

Sets default count of virtual replicas in consistent hash ring.

To determine node replicas, node attribute with getReplicaCountAttributeName() name will be checked first. If it is absent, then this value will be used.

Parameters:

replicas - Count of virtual replicas in consistent hash ring.s

getPartitions

public int getPartitions()

Gets total number of key partitions. To ensure that all partitions are equally distributed across all nodes, please make sure that this number is significantly larger than a number of nodes. Also, partition size should be relatively small. Try to avoid having partitions with more than quarter million keys.

Note that for fully replicated caches this method should always return 1.

Returns:

Total partition count.

setPartitions

public void setPartitions(int parts)

Sets total number of partitions.

Parameters:

parts - Total number of partitions.

getHashIdResolver

public GridCacheAffinityNodeHashResolver getHashIdResolver()

Gets hash ID resolver for nodes. This resolver is used to provide alternate hash ID, other than node ID.

Node IDs constantly change when nodes get restarted, which causes them to be placed on different locations in the hash ring, and hence causing repartitioning. Providing an alternate hash ID, which survives node restarts, puts node on the same location on the hash ring, hence minimizing required repartitioning.

Returns:

Hash ID resolver.

setHashIdResolver

public void setHashIdResolver(GridCacheAffinityNodeHashResolver hashIdRslvr)

Sets hash ID resolver for nodes. This resolver is used to provide alternate hash ID, other than node ID.

Node IDs constantly change when nodes get restarted, which causes them to be placed on different locations in the hash ring, and hence causing repartitioning. Providing an alternate hash ID, which survives node restarts, puts node on the same location on the hash ring, hence minimizing required repartitioning.

Parameters:

hashIdRslvr - Hash ID resolver.

getBackupFilter

@Nullable
public GridBiPredicate<GridNode,GridNode> getBackupFilter()

Gets optional backup filter. If not null, backups will be selected from all nodes that pass this filter. First node passed to this filter is primary node, and second node is a node being tested.

Note that excludeNeighbors parameter is ignored if backupFilter is set.

Returns:

Optional backup filter.

setBackupFilter

public void setBackupFilter(@Nullable
                   GridBiPredicate<GridNode,GridNode> backupFilter)

Sets optional backup filter. If provided, then backups will be selected from all nodes that pass this filter. First node being passed to this filter is primary node, and second node is a node being tested.

Note that excludeNeighbors parameter is ignored if backupFilter is set.

Parameters:

backupFilter - Optional backup filter.

getReplicaCountAttributeName

public String getReplicaCountAttributeName()

Gets optional attribute name for replica count. If not provided, the default is DFLT_REPLICA_COUNT_ATTR_NAME.

Returns:

User attribute name for replica count for a node.

setReplicaCountAttributeName

public void setReplicaCountAttributeName(String attrName)

Sets optional attribute name for replica count. If not provided, the default is DFLT_REPLICA_COUNT_ATTR_NAME.

Parameters:

attrName - User attribute name for replica count for a node.

isExcludeNeighbors

public boolean isExcludeNeighbors()

Checks flag to exclude same-host-neighbors from being backups of each other (default is false).

Note that excludeNeighbors parameter is ignored if #getBackupFilter() is set.

Returns:

True if nodes residing on the same host may not act as backups of each other.

setExcludeNeighbors

public void setExcludeNeighbors(boolean exclNeighbors)

Sets flag to exclude same-host-neighbors from being backups of each other (default is false).

Note that excludeNeighbors parameter is ignored if #getBackupFilter() is set.

Parameters:

exclNeighbors - True if nodes residing on the same host may not act as backups of each other.

assignPartitions

public List<List<GridNode>> assignPartitions(GridCacheAffinityFunctionContext ctx)

Gets affinity nodes for a partition. In case of replicated cache, all returned nodes are updated in the same manner. In case of partitioned cache, the returned list should contain only the primary and back up nodes with primary node being always first.

Note that partitioned affinity must obey the following contract: given that node N is primary for some key K, if any other node(s) leave grid and no node joins grid, node N will remain primary for key K.

Specified by:

assignPartitions in interface GridCacheAffinityFunction

Parameters:

ctx - Affinity function context. Will provide all required information to calculate new partition assignments.

Returns:

Unmodifiable list indexed by partition number. Each element of array is a collection in which first node is a primary node and other nodes are backup nodes.

nodes

public Collection<GridNode> nodes(int part,
                         Collection<GridNode> nodes,
                         int backups)

Assigns nodes to one partition.

Parameters:

part - Partition to assign nodes for.

nodes - Cache topology nodes.

Returns:

Assigned nodes, first node is primary, others are backups.

partition

public int partition(Object key)

Gets partition number for a given key starting from 0. Partitioned caches should make sure that keys are about evenly distributed across all partitions from 0 to partition count for best performance.

Note that for fully replicated caches it is possible to segment key sets among different grid node groups. In that case each node group should return a unique partition number. However, unlike partitioned cache, mappings of keys to nodes in replicated caches are constant and a node cannot migrate from one partition to another.

Specified by:

partition in interface GridCacheAffinityFunction

Parameters:

key - Key to get partition for.

Returns:

Partition number for a given key.

partitions

public int partitions()

Gets total number of partitions available. All caches should always provide correct partition count which should be the same on all participating nodes. Note that partitions should always be numbered from 0 inclusively to N exclusively without any gaps.

Specified by:

partitions in interface GridCacheAffinityFunction

Returns:

Total partition count.

reset

public void reset()

Resets cache affinity to its initial state. This method will be called by the system any time the affinity has been sent to remote node where it has to be reinitialized. If your implementation of affinity function has no initialization logic, leave this method empty.

Specified by:

reset in interface GridCacheAffinityFunction

removeNode

public void removeNode(UUID nodeId)

Removes node from affinity. This method is called when it is safe to remove left node from affinity mapping.

Specified by:

removeNode in interface GridCacheAffinityFunction

Parameters:

nodeId - ID of node to remove.

toString

public String toString()

Overrides:

toString in class Object