metricstarttimeprocessor: Implementation of the subtractinitial strategy (#38594)

<!--Ex. Fixing a bug - Describe the bug and how this fixes the issue.
Ex. Adding a feature - Explain what this achieves.-->
#### Description
Addresses #38379.

Some of this logic already existed in the [GCM
exporter](https://github.com/GoogleCloudPlatform/opentelemetry-operations-go/blob/main/exporter/collector/internal/normalization/standard_normalizer.go).

For each resource, and for each timeseries
- Skip if the point is not a cumulative sum, histogram, exp histogram,
or summary.
- Lookup the timeseries in the cache. If it is not present, drop this
point from the batch, and store the timestamp and value (including
bucket counts, etc) in the cache.
- Subtract cumulative values (e.g. sum, bucket counts) of the cached
timeseries from the current timeseries, and set the start timestamp to
the cached point's timestamp.

<!-- Issue number (e.g. #1234) or full URL to issue, if applicable. -->
#### Link to tracking issue
Fixes #38379.

<!--Describe what testing was performed and which tests were added.-->
#### Testing
Added unit and integration tests.

<!--Describe the documentation added.-->
#### Documentation
Readme already covers this strategy

<!--Please delete paragraphs that you did not use before submitting.-->

---------

Signed-off-by: Ridwan Sharif <[email protected]>

Author: ridwanmsharif

.chloggen/subtract-strategy-v2.yaml

@@ -0,0 +1,27 @@
+# Use this changelog template to create an entry for release notes.
+
+# One of 'breaking', 'deprecation', 'new_component', 'enhancement', 'bug_fix'
+change_type: enhancement
+
+# The name of the component, or a single word describing the area of concern, (e.g. filelogreceiver)
+component: metricstarttimeprocessor
+
+# A brief description of the change.  Surround your text with quotes ("") if it needs to start with a backtick (`).
+note: Implementation of the subtractinitial strategy
+
+# Mandatory: One or more tracking issues related to the change. You can use the PR number here if no issue exists.
+issues: [38379]
+
+# (Optional) One or more lines of additional information to render under the primary note.
+# These lines will be padded with 2 spaces and then inserted directly into the document.
+# Use pipe (|) for multiline entries.
+subtext:
+
+# If your change doesn't affect end users or the exported elements of any package,
+# you should instead start your pull request title with [chore] or use the "Skip Changelog" label.
+# Optional: The change log or logs in which this entry should be included.
+# e.g. '[user]' or '[user, api]'
+# Include 'user' if the change is relevant to end users.
+# Include 'api' if there is a change to a library API.
+# Default: '[user]'
+change_logs: []

processor/metricstarttimeprocessor/internal/datapointstorage/timeseries_map.go

@@ -101,20 +101,20 @@ func (tsm *TimeseriesMap) GC() {
 	tsm.Mark = false
 }
 
-// IsResetHistogram compares the given histogram datapoint h, to tsi.Histogram
+// IsResetHistogram compares the given histogram datapoint h, to ref
 // and determines whether the metric has been reset based on the values.  It is
 // a reset if any of the bucket boundaries have changed, if any of the bucket
 // counts have decreased or if the total sum or count have decreased.
-func (tsi *TimeseriesInfo) IsResetHistogram(h pmetric.HistogramDataPoint) bool {
-	if h.Count() < tsi.Histogram.Count() {
+func IsResetHistogram(h, ref pmetric.HistogramDataPoint) bool {
+	if h.Count() < ref.Count() {
 		return true
 	}
-	if h.Sum() < tsi.Histogram.Sum() {
+	if h.Sum() < ref.Sum() {
 		return true
 	}
 
 	// Guard against bucket boundaries changes.
-	refBounds := tsi.Histogram.ExplicitBounds().AsRaw()
+	refBounds := ref.ExplicitBounds().AsRaw()
 	hBounds := h.ExplicitBounds().AsRaw()
 	if len(refBounds) != len(hBounds) {
 		return true
@@ -126,69 +126,69 @@ func (tsi *TimeseriesInfo) IsResetHistogram(h pmetric.HistogramDataPoint) bool {
 	}
 
 	// We need to check individual buckets to make sure the counts are all increasing.
-	if tsi.Histogram.BucketCounts().Len() != h.BucketCounts().Len() {
+	if ref.BucketCounts().Len() != h.BucketCounts().Len() {
 		return true
 	}
-	for i := range tsi.Histogram.BucketCounts().Len() {
-		if h.BucketCounts().At(i) < tsi.Histogram.BucketCounts().At(i) {
+	for i := range ref.BucketCounts().Len() {
+		if h.BucketCounts().At(i) < ref.BucketCounts().At(i) {
 			return true
 		}
 	}
 	return false
 }
 
 // IsResetExponentialHistogram compares the given exponential histogram
-// datapoint eh, to tsi.ExponentialHistogram and determines whether the metric
+// datapoint eh, to ref and determines whether the metric
 // has been reset based on the values.  It is a reset if any of the bucket
 // boundaries have changed, if any of the bucket counts have decreased or if the
 // total sum or count have decreased.
-func (tsi *TimeseriesInfo) IsResetExponentialHistogram(eh pmetric.ExponentialHistogramDataPoint) bool {
+func IsResetExponentialHistogram(eh, ref pmetric.ExponentialHistogramDataPoint) bool {
 	// Same as the histogram implementation
-	if eh.Count() < tsi.ExponentialHistogram.Count() {
+	if eh.Count() < ref.Count() {
 		return true
 	}
-	if eh.Sum() < tsi.ExponentialHistogram.Sum() {
+	if eh.Sum() < ref.Sum() {
 		return true
 	}
 
 	// Guard against bucket boundaries changes.
-	if tsi.ExponentialHistogram.Scale() != eh.Scale() {
+	if ref.Scale() != eh.Scale() {
 		return true
 	}
 
 	// We need to check individual buckets to make sure the counts are all increasing.
-	if tsi.ExponentialHistogram.Positive().BucketCounts().Len() != eh.Positive().BucketCounts().Len() {
+	if ref.Positive().BucketCounts().Len() != eh.Positive().BucketCounts().Len() {
 		return true
 	}
-	for i := range tsi.ExponentialHistogram.Positive().BucketCounts().Len() {
-		if eh.Positive().BucketCounts().At(i) < tsi.ExponentialHistogram.Positive().BucketCounts().At(i) {
+	for i := range ref.Positive().BucketCounts().Len() {
+		if eh.Positive().BucketCounts().At(i) < ref.Positive().BucketCounts().At(i) {
 			return true
 		}
 	}
-	if tsi.ExponentialHistogram.Negative().BucketCounts().Len() != eh.Negative().BucketCounts().Len() {
+	if ref.Negative().BucketCounts().Len() != eh.Negative().BucketCounts().Len() {
 		return true
 	}
-	for i := range tsi.ExponentialHistogram.Negative().BucketCounts().Len() {
-		if eh.Negative().BucketCounts().At(i) < tsi.ExponentialHistogram.Negative().BucketCounts().At(i) {
+	for i := range ref.Negative().BucketCounts().Len() {
+		if eh.Negative().BucketCounts().At(i) < ref.Negative().BucketCounts().At(i) {
 			return true
 		}
 	}
 
 	return false
 }
 
-// IsResetSummary compares the given summary datapoint s to tsi.Summary and
+// IsResetSummary compares the given summary datapoint s to ref and
 // determines whether the metric has been reset based on the values.  It is a
 // reset if the count or sum has decreased.
-func (tsi *TimeseriesInfo) IsResetSummary(s pmetric.SummaryDataPoint) bool {
-	return s.Count() < tsi.Summary.Count() || s.Sum() < tsi.Summary.Sum()
+func IsResetSummary(s, ref pmetric.SummaryDataPoint) bool {
+	return s.Count() < ref.Count() || s.Sum() < ref.Sum()
 }
 
-// IsResetSum compares the given number datapoint s to tsi.Number and determines
+// IsResetSum compares the given number datapoint s to ref and determines
 // whether the metric has been reset based on the values.  It is a reset if the
 // value has decreased.
-func (tsi *TimeseriesInfo) IsResetSum(s pmetric.NumberDataPoint) bool {
-	return s.DoubleValue() < tsi.Number.DoubleValue()
+func IsResetSum(s, ref pmetric.NumberDataPoint) bool {
+	return s.DoubleValue() < ref.DoubleValue()
 }
 
 func newTimeseriesMap() *TimeseriesMap {

processor/metricstarttimeprocessor/internal/datapointstorage/timeseries_map_test.go

@@ -316,7 +316,7 @@ func TestTimeseriesInfo_IsResetHistogram(t *testing.T) {
 		t.Run(tt.name, func(t *testing.T) {
 			tsi := tt.setupTsi()
 			h := tt.setupH()
-			assert.Equal(t, tt.expectedReset, tsi.IsResetHistogram(h))
+			assert.Equal(t, tt.expectedReset, IsResetHistogram(h, tsi.Histogram))
 		})
 	}
 }
@@ -518,7 +518,7 @@ func TestTimeseriesInfo_IsResetExponentialHistogram(t *testing.T) {
 		t.Run(tt.name, func(t *testing.T) {
 			tsi := tt.setupTsi()
 			eh := tt.setupEh()
-			assert.Equal(t, tt.expectedReset, tsi.IsResetExponentialHistogram(eh))
+			assert.Equal(t, tt.expectedReset, IsResetExponentialHistogram(eh, tsi.ExponentialHistogram))
 		})
 	}
 }
@@ -604,7 +604,7 @@ func TestTimeseriesInfo_IsResetSummary(t *testing.T) {
 		t.Run(tt.name, func(t *testing.T) {
 			tsi := tt.setupTsi()
 			s := tt.setupS()
-			assert.Equal(t, tt.expectedReset, tsi.IsResetSummary(s))
+			assert.Equal(t, tt.expectedReset, IsResetSummary(s, tsi.Summary))
 		})
 	}
 }
@@ -652,7 +652,7 @@ func TestTimeseriesInfo_IsResetSum(t *testing.T) {
 		t.Run(tt.name, func(t *testing.T) {
 			tsi := tt.setupTsi()
 			s := tt.setupS()
-			assert.Equal(t, tt.expectedReset, tsi.IsResetSum(s))
+			assert.Equal(t, tt.expectedReset, IsResetSum(s, tsi.Number))
 		})
 	}
 }