com.datastax.oss.dsbulk.partitioner

Class TokenRangeReadStatementGenerator

java.lang.Object

com.datastax.oss.dsbulk.partitioner.TokenRangeReadStatementGenerator

public class TokenRangeReadStatementGenerator
extends Object

Generates SELECT statements that read the entire table by token ranges.

Constructor Summary

Constructors

Constructor and Description
TokenRangeReadStatementGenerator(com.datastax.oss.driver.api.core.metadata.schema.RelationMetadata table, com.datastax.oss.driver.api.core.metadata.Metadata metadata)

Method Summary

Modifier and Type	Method and Description
List<com.datastax.oss.driver.api.core.cql.Statement<?>>	generate(int splitCount) Generates default SELECT statements to read the entire table, with a minimum of splitCount statements.
List<com.datastax.oss.driver.api.core.cql.Statement<?>>	generate(int splitCount, Function<com.datastax.oss.driver.api.core.metadata.token.TokenRange,com.datastax.oss.driver.api.core.cql.Statement<?>> statementFactory) Generates SELECT statements to read the entire table, with a minimum of splitCount statements and using the given factory to generate statements.

Methods inherited from class java.lang.Object

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

Constructor Detail

TokenRangeReadStatementGenerator

public TokenRangeReadStatementGenerator(@NonNull
                                        com.datastax.oss.driver.api.core.metadata.schema.RelationMetadata table,
                                        @NonNull
                                        com.datastax.oss.driver.api.core.metadata.Metadata metadata)

Parameters:

table - The table (or materialized view) to scan.

metadata - The cluster metadata to use.

Method Detail

generate

@NonNull
public List<com.datastax.oss.driver.api.core.cql.Statement<?>> generate(int splitCount)

Generates default SELECT statements to read the entire table, with a minimum of splitCount statements.

For a given split / token range, the generated statement is a statement of the form: SELECT col1, col2,... FROM table WHERE token(...) > [range.start()] AND token(...) <= [range.end()]).

Note that the splitting algorithm doesn't guarantee an exact number of splits, but rather a minimum number. The number of resulting statements depends on the set of primary token ranges in the ring and how contiguous token ranges are distributed across the ring. In particular with vnodes, the total number of statements can be much higher than splitCount.

Parameters:

splitCount - The minimum desired number of statements to generate (on a best-effort basis).

Returns:

A list of SELECT statements to read the entire table.

generate

@NonNull
public List<com.datastax.oss.driver.api.core.cql.Statement<?>> generate(int splitCount,
                                                                                 @NonNull
                                                                                 Function<com.datastax.oss.driver.api.core.metadata.token.TokenRange,com.datastax.oss.driver.api.core.cql.Statement<?>> statementFactory)

Generates SELECT statements to read the entire table, with a minimum of splitCount statements and using the given factory to generate statements.

For each split / token range, the generated statement is a statement resulting from applying statementFactory to the token range; statement factories should typically generate a statement of the form: SELECT col1, col2,... FROM table WHERE token(...) > ? AND token(...) <= ?). Please note that this method does not fully validate that the statements created by the factory are valid, and thus should be used with caution.

Note that the splitting algorithm doesn't guarantee an exact number of splits, but rather a minimum number. The number of resulting statements depends on the set of primary token ranges in the ring and how contiguous token ranges are distributed across the ring. In particular with vnodes, the total number of statements can be much higher than splitCount.

Parameters:

splitCount - The minimum desired number of statements to generate (on a best-effort basis).

statementFactory - The factory to use to generate statements for each split.

Returns:

A list of SELECT statements to read the entire table.