Exponential Histogram not exposed by the Collector

[fmhwong]

Component(s)

exporter/prometheus

What happened?

Description

Steps to Reproduce

        AutoConfiguredOpenTelemetrySdk sdk = AutoConfiguredOpenTelemetrySdk
                .builder()
                .build();

        DoubleHistogram carSpeedSummary = meter
                .histogramBuilder("car.speedSummary")
                .setDescription("Summary of car speed")
                .setUnit("km/h")
                .build();

I can see the histogram output http://localhost:8889/metrics if I use the default EXPLICIT_BUCKET_HISTOGRAM

If I set environment variable
OTEL_EXPORTER_OTLP_METRICS_DEFAULT_HISTOGRAM_AGGREGATION=BASE2_EXPONENTIAL_BUCKET_HISTOGRAM, I didn't see any histogram output at http://localhost:8889/metrics

The collector-contrib log did show the histogram.

Metric #1
Descriptor:
     -> Name: car.speedSummary
     -> Description: Summary of car speed
     -> Unit: km/h
     -> DataType: ExponentialHistogram
     -> AggregationTemporality: Cumulative
ExponentialHistogramDataPoints #0
Data point attributes:
     -> driver: Str(Don)
StartTimestamp: 2023-08-10 15:49:54.862805415 +0000 UTC
Timestamp: 2023-08-10 15:50:19.84214473 +0000 UTC
Count: 24
Sum: 1959.000000
Min: 0.000000
Max: 129.000000
Bucket [0, 0], Count: 5
Bucket (22.873813, 23.122893], Count: 1
Bucket (23.122893, 23.374685], Count: 0
Bucket (23.374685, 23.629218], Count: 0
Bucket (23.629218, 23.886524], Count: 0
Bucket (23.886524, 24.146631], Count: 0
Bucket (24.146631, 24.409570], Count: 0
Bucket (24.409570, 24.675373], Count: 0
Bucket (24.675373, 24.944070], Count: 0
Bucket (24.944070, 25.215694], Count: 0
Bucket (25.215694, 25.490274], Count: 0
Bucket (25.490274, 25.767845], Count: 0

Expected Result

Expected to see the exponential histogram in the Prometheus endpoint from the Collector.

Actual Result

The histogram was not printed.

Collector version

prometheusexporter@v0.82.0

Environment information

Environment

OS: Docker on Mac OS
Compiler(if manually compiled): (e.g., "go 14.2")

OpenTelemetry Collector configuration

receivers:
  otlp:
    protocols:
        grpc:

exporters:
  prometheus:
    endpoint: "0.0.0.0:8889"
  logging:
    loglevel: debug

processors:
  batch:

service:
  pipelines:
    metrics:
      receivers: [otlp]
      processors: [batch]
      exporters: [prometheus, logging]

Log output

StartTimestamp: 2023-08-10 15:49:54.704889957 +0000 UTC
Timestamp: 2023-08-10 15:50:04.716748635 +0000 UTC
Value: 0
	{"kind": "exporter", "data_type": "metrics", "name": "logging"}
2023-08-10T15:50:04.991Z	error	prometheusexporter@v0.82.0/accumulator.go:94	failed to translate metric	{"kind": "exporter", "data_type": "metrics", "name": "prometheus", "data_type": "\u0004", "metric_name": "car.speedSummary"}
github.com/open-telemetry/opentelemetry-collector-contrib/exporter/prometheusexporter.(*lastValueAccumulator).addMetric
	github.com/open-telemetry/opentelemetry-collector-contrib/exporter/prometheusexporter@v0.82.0/accumulator.go:94
github.com/open-telemetry/opentelemetry-collector-contrib/exporter/prometheusexporter.(*lastValueAccumulator).Accumulate
	github.com/open-telemetry/opentelemetry-collector-contrib/exporter/prometheusexporter@v0.82.0/accumulator.go:71
github.com/open-telemetry/opentelemetry-collector-contrib/exporter/prometheusexporter.(*collector).processMetrics
	github.com/open-telemetry/opentelemetry-collector-contrib/exporter/prometheusexporter@v0.82.0/collector.go:92
github.com/open-telemetry/opentelemetry-collector-contrib/exporter/prometheusexporter.(*prometheusExporter).ConsumeMetrics
	github.com/open-telemetry/opentelemetry-collector-contrib/exporter/prometheusexporter@v0.82.0/prometheus.go:85
go.opentelemetry.io/collector/exporter/exporterhelper.(*metricsRequest).Export
	go.opentelemetry.io/collector/exporter@v0.82.0/exporterhelper/metrics.go:54
go.opentelemetry.io/collector/exporter/exporterhelper.(*timeoutSender).send
	go.opentelemetry.io/collector/exporter@v0.82.0/exporterhelper/common.go:197
go.opentelemetry.io/collector/exporter/exporterhelper.(*retrySender).send
	go.opentelemetry.io/collector/exporter@v0.82.0/exporterhelper/queued_retry.go:355
go.opentelemetry.io/collector/exporter/exporterhelper.(*metricsSenderWithObservability).send
	go.opentelemetry.io/collector/exporter@v0.82.0/exporterhelper/metrics.go:125
go.opentelemetry.io/collector/exporter/exporterhelper.(*queuedRetrySender).send
	go.opentelemetry.io/collector/exporter@v0.82.0/exporterhelper/queued_retry.go:291
go.opentelemetry.io/collector/exporter/exporterhelper.NewMetricsExporter.func2
	go.opentelemetry.io/collector/exporter@v0.82.0/exporterhelper/metrics.go:105
go.opentelemetry.io/collector/consumer.ConsumeMetricsFunc.ConsumeMetrics
	go.opentelemetry.io/collector/consumer@v0.82.0/metrics.go:25
go.opentelemetry.io/collector/internal/fanoutconsumer.(*metricsConsumer).ConsumeMetrics
	go.opentelemetry.io/collector@v0.82.0/internal/fanoutconsumer/metrics.go:69
go.opentelemetry.io/collector/processor/batchprocessor.(*batchMetrics).export
	go.opentelemetry.io/collector/processor/batchprocessor@v0.82.0/batch_processor.go:442
go.opentelemetry.io/collector/processor/batchprocessor.(*shard).sendItems
	go.opentelemetry.io/collector/processor/batchprocessor@v0.82.0/batch_processor.go:256
go.opentelemetry.io/collector/processor/batchprocessor.(*shard).start
	go.opentelemetry.io/collector/processor/batchprocessor@v0.82.0/batch_processor.go:218

Additional context

No response

[github-actions]

Pinging code owners:

• exporter/prometheus: @Aneurysm9

See Adding Labels via Comments if you do not have permissions to add labels yourself.

[mx-psi]

In case it is useful, the PRW exporter supports exponential histograms, since #17370

[crobert-1]

@fmhwong Did the prometheus remote write exporter accomplish what you were looking for here, or is there something else needed?

[fmhwong]

@crobert-1 The problem is in the Prometheus exporter. Will that be fixed too?

[crobert-1]

You're right, this should be investigated further for this exporter, thanks!

[crobert-1]

@fmhwong It looks like this is a frequency of #13443. There's more discussion there so I'm going to close this issue and comment there as well.