Running otel-action
at the end of a GitHub workflow will upload OpenTelemetry spans for each job (and job step) to an OTel compatible endpoint. This provides observability into CI pipelines, and provides a quick method to find offending commits that are the root cause of a CI job becoming slower.
Below is an example trace from this repository's test workflow being viewed in Honeycomb.io's UI (they have an awesome free-forever tier)!
- name: Upload OTEL traces
uses: MNThomson/otel-action@master
with:
endpoint: ${{ secrets.ENDPOINT }}
headers: ${{ secrets.HEADERS }}
service_name: "MyDatasetName"
Checkout the test.yml
worflow as well.