com.thinkaurelius.titan.graphdb.transaction

Interface TransactionConfiguration

All Superinterfaces:

BaseTransactionConfig

All Known Implementing Classes:

StandardTransactionBuilder

public interface TransactionConfiguration
extends BaseTransactionConfig

Provides configuration options for TitanTransaction.

Author:

Matthias Bröcheler (me@matthiasb.com);

See Also:

TitanTransaction

Method Summary

Methods

Modifier and Type	Method and Description
DefaultSchemaMaker	getAutoSchemaMaker()
int	getDirtyVertexSize() The initial size of the dirty (modified) vertex map used by a transaction.
long	getIndexCacheWeight() The maximum weight for the index cache store used in this particular transaction
String	getLogIdentifier() The name of the log to be used for logging the mutations in this transaction.
int[]	getRestrictedPartitions() Querying of partitioned vertices is restricted to the partitions returned by this method.
int	getVertexCacheSize() The maximum number of recently-used vertices to cache in this transaction.
boolean	hasAcquireLocks() Whether the persistence layer should acquire locks for this transaction during persistence.
boolean	hasAssignIDsImmediately()
boolean	hasEnabledBatchLoading() Whether this transaction should be optimized for batch-loading, i.e.
boolean	hasForceIndexUsage() Whether this transaction should throw an exception when a graph query is issued that cannot be answered with any existing index but instead requires a full graph-scan.
boolean	hasPropertyPrefetching() Whether this transaction loads all properties on a vertex when a single property is requested.
boolean	hasRestrictedPartitions() Returns true of the queryied partitions should be restricted in this transaction
boolean	hasVerifyExternalVertexExistence() Whether the graph transaction is configured to verify that a vertex with the id GIVEN BY THE USER actually exists in the database or not.
boolean	hasVerifyInternalVertexExistence() Whether the graph transaction is configured to verify that a vertex with the id actually exists in the database or not on every retrieval.
boolean	hasVerifyUniqueness() Whether the graph transaction is configured to verify that an added key does not yet exist in the database.
boolean	isReadOnly() Checks whether the graph transaction is configured as read-only.
boolean	isSingleThreaded() Whether this transaction is only accessed by a single thread.
boolean	isThreadBound() Whether this transaction is bound to a running thread.

Methods inherited from interface com.thinkaurelius.titan.diskstorage.BaseTransactionConfig

getCommitTime, getCustomOption, getCustomOptions, getGroupName, getTimestampProvider, hasCommitTime, hasGroupName, setCommitTime

Method Detail

isReadOnly

boolean isReadOnly()

Checks whether the graph transaction is configured as read-only.

Returns:

True, if the transaction is configured as read-only, else false.

hasAssignIDsImmediately

boolean hasAssignIDsImmediately()

Returns:

Whether this transaction is configured to assign idAuthorities immediately.

hasEnabledBatchLoading

boolean hasEnabledBatchLoading()

Whether this transaction should be optimized for batch-loading, i.e. ingestion of lots of data.

Returns:

hasVerifyExternalVertexExistence

boolean hasVerifyExternalVertexExistence()

Whether the graph transaction is configured to verify that a vertex with the id GIVEN BY THE USER actually exists in the database or not. In other words, it is verified that user provided vertex ids (through public APIs) actually exist.

Returns:

True, if vertex existence is verified, else false

hasVerifyInternalVertexExistence

boolean hasVerifyInternalVertexExistence()

Whether the graph transaction is configured to verify that a vertex with the id actually exists in the database or not on every retrieval. In other words, it is always verified that a vertex for a given id exists, even if that id is retrieved internally (through private APIs).

Hence, this is a defensive setting against data degradation, where edges and/or index entries might point to no longer existing vertices. Use this setting with caution as it introduces additional overhead entailed by checking the existence.

Unlike hasVerifyExternalVertexExistence() this is about internally verifying ids.

Returns:

True, if vertex existence is verified, else false

hasAcquireLocks

boolean hasAcquireLocks()

Whether the persistence layer should acquire locks for this transaction during persistence.

Returns:

True, if locks should be acquired, else false

getAutoSchemaMaker

DefaultSchemaMaker getAutoSchemaMaker()

Returns:

The default edge type maker used to automatically create not yet existing edge types.

hasVerifyUniqueness

boolean hasVerifyUniqueness()

Whether the graph transaction is configured to verify that an added key does not yet exist in the database.

Returns:

True, if vertex existence is verified, else false

hasPropertyPrefetching

boolean hasPropertyPrefetching()

Whether this transaction loads all properties on a vertex when a single property is requested. This can be highly beneficial when additional properties are requested on the same vertex at a later time. For vertices with very many properties this might increase latencies of property fetching.

Returns:

True, if this transaction pre-fetches all properties

isSingleThreaded

boolean isSingleThreaded()

Whether this transaction is only accessed by a single thread. If so, then certain data structures may be optimized for single threaded access since locking can be avoided.

Returns:

isThreadBound

boolean isThreadBound()

Whether this transaction is bound to a running thread. If so, then elements in this transaction can expand their life cycle to the next transaction in the thread.

Returns:

getVertexCacheSize

int getVertexCacheSize()

The maximum number of recently-used vertices to cache in this transaction. The recently-used vertex cache can include both clean and dirty vertices.

Returns:

getDirtyVertexSize

int getDirtyVertexSize()

The initial size of the dirty (modified) vertex map used by a transaction.

Returns:

getIndexCacheWeight

long getIndexCacheWeight()

The maximum weight for the index cache store used in this particular transaction

Returns:

getLogIdentifier

String getLogIdentifier()

The name of the log to be used for logging the mutations in this transaction. If the identifier is NULL the mutations will not be logged.

Returns:

hasForceIndexUsage

boolean hasForceIndexUsage()

Whether this transaction should throw an exception when a graph query is issued that cannot be answered with any existing index but instead requires a full graph-scan.

Returns:

getRestrictedPartitions

int[] getRestrictedPartitions()

Querying of partitioned vertices is restricted to the partitions returned by this method. If the return value has length 0 all partitions are queried (i.e. unrestricted).

Returns:

hasRestrictedPartitions

boolean hasRestrictedPartitions()

Returns true of the queryied partitions should be restricted in this transaction