How are GitHub Actions metrics computed in insights

[ooooo-create]

Why are you starting this discussion?

Question

What GitHub Actions topic or product is this about?

Metrics & Insights

Discussion Details

Is there a detailed calculation process for Actions Usage Metrics and Actions Performance Metrics in the project's insights?

• total_minutes
• job_runs
• failure_rate
• avg_run_time
• avg_queue_time

[abre0101]

Good question! The metrics can be a bit confusing at first. Let me break down what I know about how GitHub calculates these.

Actions Usage Metrics

These track your consumption and are calculated based on:

Billable minutes: Actual runtime of jobs, multiplied by the runner type multiplier
Linux runners: 1x multiplier
Windows runners: 2x multiplier
macOS runners: 10x multiplier
Storage: Artifacts and caches stored in your repo
Data transfer: Usually not charged for public repos
So if a job runs for 10 minutes on macOS, that counts as 100 billable minutes.

Actions Performance Metrics

These show workflow efficiency:

Workflow run duration: Total time from trigger to completion (including queue time)
Job duration: Actual execution time per job
Success/failure rates: Percentage of runs that pass vs fail
Queue time: How long jobs wait before a runner picks them up
Where the Data Comes From

GitHub pulls this from:

Workflow run logs
Job execution timestamps
Runner allocation records
Artifact/cache storage APIs
What's Not Always Clear

The insights don't always show:

Exact queue time breakdowns
Per-step timing (you need to check individual run logs)
Historical trends beyond the displayed time range
Concurrent job limits impact
How to See More Detail

For deeper analysis:

Check individual workflow runs for step-by-step timing
Use the GitHub API: GET /repos/{owner}/{repo}/actions/workflows/{workflow_id}/timing
Export data using Actions usage reports (Settings → Billing → Usage this month → Export)
Is There Official Documentation?

GitHub's docs cover the basics here: https://docs.github.com/en/organizations/collaborating-with-groups-in-organizations/viewing-insights-for-your-organization

But honestly, they could be more detailed about the exact calculation formulas. If you're trying to optimize costs or performance, I'd recommend tracking specific workflows over time and comparing the metrics.

What are you trying to figure out specifically? If you share more about your use case, I might be able to point you in a better direction!

[ooooo-create]

Thanks. Because I want to collect the action metric for a repo, may be I will create some metric different from the metric in insights, I will use rest api to get jobs's info and compute it, so I want to a unify way with existing metric. If can't know, the existing metric value also should use rest API to compute it.