Is your feature request related to a problem?
`PrometheusConfig` exposes `histogram_bucket_overrides` for overriding the Core SDK's default histogram bucket boundaries. `OpenTelemetryConfig` does not, and neither does the bridge (`temporalio.bridge.runtime.OpenTelemetryConfig`). The Core SDK's default duration histogram buckets top out at 10-60s depending on the metric. For longer-running activities or workflows, all samples land in `+Inf` and `histogram_quantile()` returns meaningless values. The OTel Collector cannot rebucket explicit-bucket histograms downstream; boundaries are baked in at the producer.
Describe the solution you'd like
Add `histogram_bucket_overrides: Mapping[str, Sequence[float]] | None = None` to `OpenTelemetryConfig`, matching `PrometheusConfig`.
Additional context
temporalio 1.23.0, Python 3.13.
Is your feature request related to a problem?
`PrometheusConfig` exposes `histogram_bucket_overrides` for overriding the Core SDK's default histogram bucket boundaries. `OpenTelemetryConfig` does not, and neither does the bridge (`temporalio.bridge.runtime.OpenTelemetryConfig`). The Core SDK's default duration histogram buckets top out at 10-60s depending on the metric. For longer-running activities or workflows, all samples land in `+Inf` and `histogram_quantile()` returns meaningless values. The OTel Collector cannot rebucket explicit-bucket histograms downstream; boundaries are baked in at the producer.
Describe the solution you'd like
Add `histogram_bucket_overrides: Mapping[str, Sequence[float]] | None = None` to `OpenTelemetryConfig`, matching `PrometheusConfig`.
Additional context
temporalio 1.23.0, Python 3.13.