Package io.opentelemetry.sdk.metrics.internal.exemplar

Interface ExemplarReservoir<T extends ExemplarData>

All Known Implementing Classes:

FilteredExemplarReservoir, FixedSizeExemplarReservoir, HistogramExemplarReservoir, LongToDoubleExemplarReservoir, NoopExemplarReservoir, RandomFixedSizeExemplarReservoir

public interface ExemplarReservoir<T extends ExemplarData>

An interface for an exemplar reservoir of samples.

This represents a reservoir for a specific "point" of metric data.

This class is internal and is hence not for public use. Its APIs are unstable and can change at any time.

Method Summary

Modifier and Type	Method	Description
List<T>	collectAndReset(Attributes pointAttributes)	Returns an immutable list of Exemplars for exporting from the current reservoir.
static ExemplarReservoir<DoubleExemplarData>	doubleFixedSizeReservoir(Clock clock, int size, Supplier<Random> randomSupplier)	A double reservoir with fixed size that stores the given number of exemplars.
static ExemplarReservoir<DoubleExemplarData>	doubleNoSamples()	A double exemplar reservoir that stores no exemplars.
static <T extends ExemplarData> ExemplarReservoir<T>	filtered(ExemplarFilter filter, ExemplarReservoir<T> original)	Wraps a ExemplarReservoir with a measurement pre-filter.
static ExemplarReservoir<DoubleExemplarData>	histogramBucketReservoir(Clock clock, List<Double> boundaries)	A Reservoir sampler that preserves the latest seen measurement per-histogram bucket.
static ExemplarReservoir<LongExemplarData>	longFixedSizeReservoir(Clock clock, int size, Supplier<Random> randomSupplier)	A long reservoir with fixed size that stores the given number of exemplars.
static ExemplarReservoir<LongExemplarData>	longNoSamples()	A long exemplar reservoir that stores no exemplars.
static <T extends ExemplarData> ExemplarReservoir<T>	longToDouble(ExemplarReservoir<T> delegate)	Wraps an ExemplarReservoir, casting calls from offerLongMeasurement(long, Attributes, Context) to offerDoubleMeasurement(double, Attributes, Context) such that collectAndReset(Attributes) only returns DoubleExemplarData.
void	offerDoubleMeasurement(double value, Attributes attributes, Context context)	Offers a double measurement to be sampled.
void	offerLongMeasurement(long value, Attributes attributes, Context context)	Offers a long measurement to be sampled.

Method Details

longToDouble

static <T extends ExemplarData>
ExemplarReservoir<T> longToDouble(ExemplarReservoir<T> delegate)

Wraps an ExemplarReservoir, casting calls from offerLongMeasurement(long, Attributes, Context) to offerDoubleMeasurement(double, Attributes, Context) such that collectAndReset(Attributes) only returns DoubleExemplarData.

This is used for Aggregation.explicitBucketHistogram() and Aggregation.base2ExponentialBucketHistogram() which only support double measurements.

filtered

static <T extends ExemplarData>
ExemplarReservoir<T> filtered(ExemplarFilter filter,
 ExemplarReservoir<T> original)

Wraps a ExemplarReservoir with a measurement pre-filter.

doubleNoSamples

static ExemplarReservoir<DoubleExemplarData> doubleNoSamples()

A double exemplar reservoir that stores no exemplars.

longNoSamples

static ExemplarReservoir<LongExemplarData> longNoSamples()

A long exemplar reservoir that stores no exemplars.

doubleFixedSizeReservoir

static ExemplarReservoir<DoubleExemplarData> doubleFixedSizeReservoir(Clock clock,
 int size,
 Supplier<Random> randomSupplier)

A double reservoir with fixed size that stores the given number of exemplars.

Parameters:

clock - The clock to use when annotating measurements with time.

size - The maximum number of exemplars to preserve.

randomSupplier - The random number generator to use for sampling.

longFixedSizeReservoir

static ExemplarReservoir<LongExemplarData> longFixedSizeReservoir(Clock clock,
 int size,
 Supplier<Random> randomSupplier)

A long reservoir with fixed size that stores the given number of exemplars.

Parameters:

clock - The clock to use when annotating measurements with time.

size - The maximum number of exemplars to preserve.

randomSupplier - The random number generator to use for sampling.

histogramBucketReservoir

static ExemplarReservoir<DoubleExemplarData> histogramBucketReservoir(Clock clock,
 List<Double> boundaries)

A Reservoir sampler that preserves the latest seen measurement per-histogram bucket.

Parameters:

clock - The clock to use when annotating measurements with time.

boundaries - A list of (inclusive) upper bounds for the histogram. Should be in order from lowest to highest.

offerDoubleMeasurement

void offerDoubleMeasurement(double value,
 Attributes attributes,
 Context context)

Offers a double measurement to be sampled.

offerLongMeasurement

void offerLongMeasurement(long value,
 Attributes attributes,
 Context context)

Offers a long measurement to be sampled.

collectAndReset

List<T> collectAndReset(Attributes pointAttributes)

Returns an immutable list of Exemplars for exporting from the current reservoir.

Additionally, clears the reservoir for the next sampling period.

Parameters:

pointAttributes - the Attributes associated with the metric point. ExemplarDatas should filter these out of their final data state.

Returns:

An (immutable) list of sampled exemplars for this point. Implementers are expected to filter out pointAttributes from the original recorded attributes.