Sum not being reset on AggreagationTempoarlity.DELTA Histograms

[pichlermarc]

Describe your environment

• Python 3.8.10
• Ubuntu 20.04
• OTel Python Metrics SDK v0.29b0

Steps to reproduce

Here's an example that exports to the console:

from opentelemetry import _metrics
from opentelemetry.sdk._metrics import MeterProvider
from opentelemetry.sdk._metrics.export import PeriodicExportingMetricReader, ConsoleMetricExporter
from opentelemetry.sdk._metrics.point import AggregationTemporality


class DeltaConsoleMetricExporter(ConsoleMetricExporter):
    @property
    def preferred_temporality(self) -> AggregationTemporality:
        return AggregationTemporality.DELTA

if __name__ == '__main__':
    reader = PeriodicExportingMetricReader(
            export_interval_millis=3600000,
            exporter=DeltaConsoleMetricExporter())

    _metrics.set_meter_provider(MeterProvider(metric_readers=[reader]))

    meter = _metrics.get_meter("my.meter")

    my_instrument = meter.create_histogram(
        name="my.instrument",
    )

    my_instrument.record(10)
    reader.collect()
    my_instrument.record(10)
    reader.collect()
    my_instrument.record(10)
    reader.collect()

What is the expected behavior?

sum should not carry over from the last collection cycle to the next, should be 10 on every export.

What is the actual behavior?

sum carries over from the previous collection cycle to the next.

Console Output from the example above:

{"attributes": "", "description": "", "instrumentation_info": "InstrumentationInfo(my.meter, , None)", "name": "my.instrument", "resource": "BoundedAttributes({'telemetry.sdk.language': 'python', 'telemetry.sdk.name': 'opentelemetry', 'telemetry.sdk.version': '1.10.0', 'service.name': 'unknown_service'}, maxlen=None)", "unit": "", "point": {"start_time_unix_nano": 1647361426978696864, "time_unix_nano": 1647361426978710755, "bucket_counts": [0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0], "explicit_bounds": [0.0, 5.0, 10.0, 25.0, 50.0, 75.0, 100.0, 250.0, 500.0, 1000.0], "sum": 10, "aggregation_temporality": 1}}
{"attributes": "", "description": "", "instrumentation_info": "InstrumentationInfo(my.meter, , None)", "name": "my.instrument", "resource": "BoundedAttributes({'telemetry.sdk.language': 'python', 'telemetry.sdk.name': 'opentelemetry', 'telemetry.sdk.version': '1.10.0', 'service.name': 'unknown_service'}, maxlen=None)", "unit": "", "point": {"start_time_unix_nano": 1647361426978710756, "time_unix_nano": 1647361427004077424, "bucket_counts": [0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0], "explicit_bounds": [0.0, 5.0, 10.0, 25.0, 50.0, 75.0, 100.0, 250.0, 500.0, 1000.0], "sum": 20, "aggregation_temporality": 1}}
{"attributes": "", "description": "", "instrumentation_info": "InstrumentationInfo(my.meter, , None)", "name": "my.instrument", "resource": "BoundedAttributes({'telemetry.sdk.language': 'python', 'telemetry.sdk.name': 'opentelemetry', 'telemetry.sdk.version': '1.10.0', 'service.name': 'unknown_service'}, maxlen=None)", "unit": "", "point": {"start_time_unix_nano": 1647361427004077425, "time_unix_nano": 1647361427004196751, "bucket_counts": [0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0], "explicit_bounds": [0.0, 5.0, 10.0, 25.0, 50.0, 75.0, 100.0, 250.0, 500.0, 1000.0], "sum": 30, "aggregation_temporality": 1}}
{"attributes": "", "description": "", "instrumentation_info": "InstrumentationInfo(my.meter, , None)", "name": "my.instrument", "resource": "BoundedAttributes({'telemetry.sdk.language': 'python', 'telemetry.sdk.name': 'opentelemetry', 'telemetry.sdk.version': '1.10.0', 'service.name': 'unknown_service'}, maxlen=None)", "unit": "", "point": {"start_time_unix_nano": 1647361427004196752, "time_unix_nano": 1647361427004373526, "bucket_counts": [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], "explicit_bounds": [0.0, 5.0, 10.0, 25.0, 50.0, 75.0, 100.0, 250.0, 500.0, 1000.0], "sum": 30, "aggregation_temporality": 1}}

sum is kept over multiple collection cycles while bucket count is being reset.