org.apache.druid.segment.incremental

Class IncrementalIndex

java.lang.Object

org.apache.druid.segment.AbstractIndex

org.apache.druid.segment.incremental.IncrementalIndex

All Implemented Interfaces:

Closeable, AutoCloseable, Iterable<Row>, Expr.InputBindingInspector, ColumnInspector

Direct Known Subclasses:

OnheapIncrementalIndex

public abstract class IncrementalIndex
extends AbstractIndex
implements Iterable<Row>, Closeable, ColumnInspector

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	IncrementalIndex.DimensionDesc
static class	IncrementalIndex.MetricDesc

Field Summary

Fields

Modifier and Type	Field and Description
protected boolean	preserveExistingMetrics

Constructor Summary

Constructors

Modifier

Constructor and Description

protected

IncrementalIndex(IncrementalIndexSchema incrementalIndexSchema, boolean deserializeComplexMetrics, boolean concurrentEventAdd, boolean preserveExistingMetrics, boolean useMaxMemoryEstimates) Setting deserializeComplexMetrics to false is necessary for intermediate aggregation such as groupBy that should not deserialize input columns using ComplexMetricSerde for aggregators that return complex metrics.

Method Summary

Modifier and Type	Method and Description
IncrementalIndexAddResult	add(InputRow row) Adds a new row.
IncrementalIndexAddResult	add(InputRow row, boolean skipMaxRowsInMemoryCheck) Adds a new row.
protected abstract org.apache.druid.segment.incremental.IncrementalIndex.AddToFactsResult	addToFacts(InputRow row, IncrementalIndexRow key, ThreadLocal<InputRow> rowContainer, com.google.common.base.Supplier<InputRow> rowSupplier, boolean skipMaxRowsInMemoryCheck)
abstract boolean	canAppendRow()
void	close()
protected Comparator<IncrementalIndexRow>	dimsComparator()
InputRow	formatRow(InputRow row)
AtomicLong	getBytesInMemory()
Map<String,ColumnCapabilities>	getColumnCapabilities()
ColumnCapabilities	getColumnCapabilities(String columnName) Returns capabilities of a particular column.
List<String>	getColumnNames()
static ParseException	getCombinedParseException(InputRow row, List<String> dimParseExceptionMessages, List<String> aggParseExceptionMessages)
IncrementalIndex.DimensionDesc	getDimension(String dimension)
Integer	getDimensionIndex(String dimension)
List<String>	getDimensionNames()
List<String>	getDimensionOrder()
List<IncrementalIndex.DimensionDesc>	getDimensions()
DimensionsSpec	getDimensionsSpec() Returns dimensionsSpec from the ingestionSpec.
abstract org.apache.druid.segment.incremental.IncrementalIndex.FactsHolder	getFacts()
org.joda.time.Interval	getInterval()
abstract int	getLastRowIndex()
org.joda.time.DateTime	getMaxIngestedEventTime()
org.joda.time.DateTime	getMaxTime()
Metadata	getMetadata()
AggregatorFactory[]	getMetricAggs()
protected abstract double	getMetricDoubleValue(int rowOffset, int aggOffset)
protected abstract float	getMetricFloatValue(int rowOffset, int aggOffset)
protected abstract long	getMetricLongValue(int rowOffset, int aggOffset)
List<String>	getMetricNames()
protected abstract Object	getMetricObjectValue(int rowOffset, int aggOffset)
String	getMetricType(String metric)
org.joda.time.DateTime	getMinTime()
abstract String	getOutOfRowsReason()
protected abstract void	initAggs(AggregatorFactory[] metrics, com.google.common.base.Supplier<InputRow> rowSupplier, boolean deserializeComplexMetrics, boolean concurrentEventAdd)
boolean	isEmpty()
protected abstract boolean	isNull(int rowOffset, int aggOffset)
boolean	isRollup()
abstract Iterable<Row>	iterableWithPostAggregations(List<PostAggregator> postAggs, boolean descending)
Iterator<Row>	iterator()
void	loadDimensionIterable(Iterable<String> oldDimensionOrder, Map<String,ColumnCapabilities> oldColumnCapabilities) Currently called to initialize IncrementalIndex dimension order during index creation Index dimension ordering could be changed to initialize from DimensionsSpec after resolution of https://github.com/apache/druid/issues/2011
protected ColumnSelectorFactory	makeColumnSelectorFactory(AggregatorFactory agg, com.google.common.base.Supplier<InputRow> in, boolean deserializeComplexMetrics)
static ColumnSelectorFactory	makeColumnSelectorFactory(VirtualColumns virtualColumns, AggregatorFactory agg, com.google.common.base.Supplier<InputRow> in, boolean deserializeComplexMetrics) Column selector used at ingestion time for inputs to aggregators.
ColumnValueSelector<?>	makeMetricColumnValueSelector(String metric, IncrementalIndexRowHolder currEntry)
int	size()
StorageAdapter	toStorageAdapter()

Methods inherited from class org.apache.druid.segment.AbstractIndex

toString

Methods inherited from class java.lang.Object

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

Methods inherited from interface java.lang.Iterable

forEach, spliterator

Methods inherited from interface org.apache.druid.segment.ColumnInspector

getColumnCapabilitiesWithDefault, getType

Methods inherited from interface org.apache.druid.math.expr.Expr.InputBindingInspector

areNumeric, areNumeric, areSameTypes, areSameTypes, areScalar, areScalar, canVectorize, canVectorize

Field Detail

preserveExistingMetrics

protected final boolean preserveExistingMetrics

Constructor Detail

IncrementalIndex

protected IncrementalIndex(IncrementalIndexSchema incrementalIndexSchema,
                           boolean deserializeComplexMetrics,
                           boolean concurrentEventAdd,
                           boolean preserveExistingMetrics,
                           boolean useMaxMemoryEstimates)

Setting deserializeComplexMetrics to false is necessary for intermediate aggregation such as groupBy that should not deserialize input columns using ComplexMetricSerde for aggregators that return complex metrics.

Set concurrentEventAdd to true to indicate that adding of input row should be thread-safe (for example, groupBy where the multiple threads can add concurrently to the IncrementalIndex).

Parameters:

incrementalIndexSchema - the schema to use for incremental index

deserializeComplexMetrics - flag whether or not to call ComplexMetricExtractor.extractValue() on the input value for aggregators that return metrics other than float.

concurrentEventAdd - flag whether ot not adding of input rows should be thread-safe

preserveExistingMetrics - When set to true, for any row that already has metric (with the same name defined in metricSpec), the metric aggregator in metricSpec is skipped and the existing metric is unchanged. If the row does not already have the metric, then the metric aggregator is applied on the source column as usual. This should only be set for DruidInputSource since that is the only case where we can have existing metrics. This is currently only use by auto compaction and should not be use for anything else.

useMaxMemoryEstimates - true if max values should be used to estimate memory

Method Detail

makeColumnSelectorFactory

public static ColumnSelectorFactory makeColumnSelectorFactory(VirtualColumns virtualColumns,
                                                              AggregatorFactory agg,
                                                              com.google.common.base.Supplier<InputRow> in,
                                                              boolean deserializeComplexMetrics)

Column selector used at ingestion time for inputs to aggregators.

Parameters:

agg - the aggregator

in - ingestion-time input row supplier

deserializeComplexMetrics - whether complex objects should be deserialized by a ComplexMetricExtractor

Returns:

column selector factory

getFacts

public abstract org.apache.druid.segment.incremental.IncrementalIndex.FactsHolder getFacts()

canAppendRow

public abstract boolean canAppendRow()

getOutOfRowsReason

public abstract String getOutOfRowsReason()

initAggs

protected abstract void initAggs(AggregatorFactory[] metrics,
                                 com.google.common.base.Supplier<InputRow> rowSupplier,
                                 boolean deserializeComplexMetrics,
                                 boolean concurrentEventAdd)

addToFacts

protected abstract org.apache.druid.segment.incremental.IncrementalIndex.AddToFactsResult addToFacts(InputRow row,
                                                                                                     IncrementalIndexRow key,
                                                                                                     ThreadLocal<InputRow> rowContainer,
                                                                                                     com.google.common.base.Supplier<InputRow> rowSupplier,
                                                                                                     boolean skipMaxRowsInMemoryCheck)
                                                                                              throws IndexSizeExceededException

Throws:

IndexSizeExceededException

getLastRowIndex

public abstract int getLastRowIndex()

getMetricFloatValue

protected abstract float getMetricFloatValue(int rowOffset,
                                             int aggOffset)

getMetricLongValue

protected abstract long getMetricLongValue(int rowOffset,
                                           int aggOffset)

getMetricObjectValue

protected abstract Object getMetricObjectValue(int rowOffset,
                                               int aggOffset)

getMetricDoubleValue

protected abstract double getMetricDoubleValue(int rowOffset,
                                               int aggOffset)

isNull

protected abstract boolean isNull(int rowOffset,
                                  int aggOffset)

isRollup

public boolean isRollup()

close

public void close()

Specified by:

close in interface Closeable

Specified by:

close in interface AutoCloseable

formatRow

public InputRow formatRow(InputRow row)

getColumnCapabilities

public Map<String,ColumnCapabilities> getColumnCapabilities()

getColumnCapabilities

@Nullable
public ColumnCapabilities getColumnCapabilities(String columnName)

Description copied from interface: ColumnInspector

Returns capabilities of a particular column.

Specified by:

getColumnCapabilities in interface ColumnInspector

Parameters:

columnName - column name

Returns:

capabilities, or null

add

public IncrementalIndexAddResult add(InputRow row)
                              throws IndexSizeExceededException

Adds a new row. The row might correspond with another row that already exists, in which case this will update that row instead of inserting a new one.

Calls to add() are thread safe.

Parameters:

row - the row of data to add

Returns:

the number of rows in the data set after adding the InputRow. If any parse failure occurs, a ParseException is returned in IncrementalIndexAddResult.

Throws:

IndexSizeExceededException

add

public IncrementalIndexAddResult add(InputRow row,
                                     boolean skipMaxRowsInMemoryCheck)
                              throws IndexSizeExceededException

Adds a new row. The row might correspond with another row that already exists, in which case this will update that row instead of inserting a new one.

Calls to add() are thread safe.

Parameters:

row - the row of data to add

skipMaxRowsInMemoryCheck - whether or not to skip the check of rows exceeding the max rows limit

Returns:

the number of rows in the data set after adding the InputRow. If any parse failure occurs, a ParseException is returned in IncrementalIndexAddResult.

Throws:

IndexSizeExceededException - this exception is thrown once it reaches max rows limit and skipMaxRowsInMemoryCheck is set to false.

getCombinedParseException

@Nullable
public static ParseException getCombinedParseException(InputRow row,
                                                                 @Nullable
                                                                 List<String> dimParseExceptionMessages,
                                                                 @Nullable
                                                                 List<String> aggParseExceptionMessages)

isEmpty

public boolean isEmpty()

size

public int size()

getBytesInMemory

public AtomicLong getBytesInMemory()

getMetricAggs

public AggregatorFactory[] getMetricAggs()

getDimensionsSpec

public DimensionsSpec getDimensionsSpec()

Returns dimensionsSpec from the ingestionSpec.

getDimensionNames

public List<String> getDimensionNames()

getDimensions

public List<IncrementalIndex.DimensionDesc> getDimensions()

getDimension

@Nullable
public IncrementalIndex.DimensionDesc getDimension(String dimension)

getMetricType

@Nullable
public String getMetricType(String metric)

makeMetricColumnValueSelector

public ColumnValueSelector<?> makeMetricColumnValueSelector(String metric,
                                                            IncrementalIndexRowHolder currEntry)

getInterval

public org.joda.time.Interval getInterval()

getMinTime

@Nullable
public org.joda.time.DateTime getMinTime()

getMaxTime

@Nullable
public org.joda.time.DateTime getMaxTime()

getDimensionIndex

@Nullable
public Integer getDimensionIndex(String dimension)

getDimensionOrder

public List<String> getDimensionOrder()

loadDimensionIterable

public void loadDimensionIterable(Iterable<String> oldDimensionOrder,
                                  Map<String,ColumnCapabilities> oldColumnCapabilities)

Currently called to initialize IncrementalIndex dimension order during index creation Index dimension ordering could be changed to initialize from DimensionsSpec after resolution of https://github.com/apache/druid/issues/2011

getMetricNames

public List<String> getMetricNames()

getColumnNames

public List<String> getColumnNames()

Specified by:

getColumnNames in class AbstractIndex

toStorageAdapter

public StorageAdapter toStorageAdapter()

Specified by:

toStorageAdapter in class AbstractIndex

getMetadata

public Metadata getMetadata()

iterator

public Iterator<Row> iterator()

Specified by:

iterator in interface Iterable<Row>

iterableWithPostAggregations

public abstract Iterable<Row> iterableWithPostAggregations(@Nullable
                                                           List<PostAggregator> postAggs,
                                                           boolean descending)

getMaxIngestedEventTime

public org.joda.time.DateTime getMaxIngestedEventTime()

makeColumnSelectorFactory

protected ColumnSelectorFactory makeColumnSelectorFactory(AggregatorFactory agg,
                                                          com.google.common.base.Supplier<InputRow> in,
                                                          boolean deserializeComplexMetrics)

dimsComparator

protected final Comparator<IncrementalIndexRow> dimsComparator()