Fix SumAggregation (open-telemetry#3390)

Co-authored-by: Aaron Abbott <aaronabbott@google.com>

Author: ocelotl

CHANGELOG.md

@@ -7,6 +7,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## Unreleased
 
+- Fix `SumAggregation`
+  ([#3390](https://github.com/open-telemetry/opentelemetry-python/pull/3390))
 - Fix handling of empty metric collection cycles
   ([#3335](https://github.com/open-telemetry/opentelemetry-python/pull/3335))
 - Fix error when no LoggerProvider configured for LoggingHandler
@@ -24,7 +26,6 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Prometheus exporter support for auto instrumentation 
   ([#3413](https://github.com/open-telemetry/opentelemetry-python/pull/3413))
 
-
 ## Version 1.20.0/0.41b0 (2023-09-04)
 
 - Modify Prometheus exporter to translate non-monotonic Sums into Gauges

opentelemetry-sdk/src/opentelemetry/sdk/metrics/_internal/_view_instrument_match.py

@@ -74,8 +74,8 @@ def conflicts(self, other: "_ViewInstrumentMatch") -> bool:
                 result
                 and self._aggregation._instrument_is_monotonic
                 == other._aggregation._instrument_is_monotonic
-                and self._aggregation._instrument_temporality
-                == other._aggregation._instrument_temporality
+                and self._aggregation._instrument_aggregation_temporality
+                == other._aggregation._instrument_aggregation_temporality
             )
 
         return result
@@ -124,15 +124,15 @@ def consume_measurement(self, measurement: Measurement) -> None:
 
     def collect(
         self,
-        aggregation_temporality: AggregationTemporality,
+        collection_aggregation_temporality: AggregationTemporality,
         collection_start_nanos: int,
     ) -> Optional[Sequence[DataPointT]]:
 
         data_points: List[DataPointT] = []
         with self._lock:
             for aggregation in self._attributes_aggregation.values():
                 data_point = aggregation.collect(
-                    aggregation_temporality, collection_start_nanos
+                    collection_aggregation_temporality, collection_start_nanos
                 )
                 if data_point is not None:
                     data_points.append(data_point)