Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[processor/cumulativetodelta] Convert cumulative Histograms to delta temporality (#12423) #12563

Merged
merged 15 commits into from
Aug 3, 2022
Merged
10 changes: 10 additions & 0 deletions processor/cumulativetodeltaprocessor/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -11,6 +11,8 @@

The cumulative to delta processor (`cumulativetodeltaprocessor`) converts monotonic, cumulative sum and histogram metrics to monotonic, delta metrics. Non-monotonic sums and exponential histograms are excluded.

Histogram conversion is currently behind a [feature gate](#feature-gate-configurations) and will only be converted if the feature flag is set.

## Configuration

Configuration is specified through a list of metrics. The processor uses metric names to identify a set of cumulative metrics and converts them from cumulative to delta.
Expand Down Expand Up @@ -76,6 +78,14 @@ processors:
# convert all cumulative sum or histogram metrics to delta
```

## Feature gate configurations

The **processor.cumulativetodeltaprocessor.EnableHistogramSupport** feature flag controls whether cumulative histograms will be converted to delta temporality or not. It is disabled by default, meaning histograms will not be modified by the processor.
mistodon marked this conversation as resolved.
Show resolved Hide resolved

Pass `--feature-gates processor.cumulativetodeltaprocessor.EnableHistogramSupport` to enable this feature.

This feature flag will be removed, and histograms will be enabled by default in release v0.60.0, September 2022.

## Warnings

- [Statefulness](https://github.com/open-telemetry/opentelemetry-collector/blob/main/docs/standard-warnings.md#statefulness): The cumulativetodelta processor's calculates delta by remembering the previous value of a metric. For this reason, the calculation is only accurate if the metric is continuously sent to the same instance of the collector. As a result, the cumulativetodelta processor may not work as expected if used in a deployment of multiple collectors. When using this processor it is best for the data source to being sending data to a single collector.
Expand Down