Generate spanMetrics BEFORE tailSampling processor

[cameronwaterman]

More of a question than an issue, but is there a way to generate spanMetrics prior to running the tailSampling processor? It's my understanding that with the current configurations available the otlp destination processors will always run before the applicationObservability connectors.

I need pre-sampled spanMetrics for the Grafana Cloud Application Observability functionality to work as expected, as I'm sampling error and non-error traces at different rates.

[rlankfo]

Span metrics generation runs prior to tail sampling. The applicationObservability pipeline will run before the destination pipelines. Are you seeing something that would indicate this isn't the case?