Skip to main content
All RunsOn integration capabilities

​
About

RunsOn runs self-hosted GitHub Actions runners in your own AWS account. Its Flex and Fleet stacks export OpenTelemetry over OTLP/HTTP to one endpoint you set on the stack. Set that endpoint to your Sazabi intake URL and the control plane sends its structured operator logs (job_launched, job_summary, operator_snapshot, spot_interruption) and traces. Runners with the otel extra also send their bootstrap log and agent spans. No collector, RunsOn API access, or workflow change is needed, and the CloudWatch logs and GitHub job logs are unchanged. Control-plane records carry the resource service.name runs-on-flex or runs-on-fleet, with the stack name as service.namespace. Runner records carry service.name runs-on-agent.
RunsOn sends logs, traces and metrics to the same endpoint, and has no setting to turn metric export off. Its metrics arrive on this source too, and Sazabi stores each data point as a searchable record, the way the OpenTelemetry Metrics source does. To keep metrics in their own source, run an OpenTelemetry Collector between RunsOn and Sazabi and send its metrics pipeline to the OpenTelemetry Metrics source.

​
Prerequisites

  • A RunsOn Flex or Fleet stack on v3.2.0 or later, deployed with the Terraform module or, for Flex, the CloudFormation template. Runner-side export needs Flex v2.12.0 or Fleet v3.2.0.
  • Outbound HTTPS from the RunsOn control plane and runners to your Sazabi intake host
  • Your Sazabi intake URL, shown above — open the log source and copy the Intake URL from its setup screen (the public key is embedded in the URL)

​
Set up in the dashboard

1

​
Set the stack’s OTLP endpoint

Replace <your intake host> with the intake URL shown above, without https://. Use the base URL with no path: RunsOn appends /v1/logs, /v1/traces and /v1/metrics itself. Leave the exporter headers empty; the public key is in the hostname.Flex or Fleet with Terraform. Add these inputs to your RunsOn module block, then run terraform apply. Both enable flags default to true.
Terraform
Flex with CloudFormation. Update the stack in place with the current template, set OtelExporterEndpoint to https://<your intake host>, and leave every other parameter as it is. Logs and traces are exported as soon as the endpoint is set.
2

​
Add runner bootstrap logs (optional)

Runners export their bootstrap log (output.log) and agent spans only with the otel extra. Keep any extras you already use.An extras job label replaces the runner’s extras list, so list every extra the job needs (extras=s3-cache+otel). Fleet workflows can’t opt in per job.
3

​
Hand off the change

If your RunsOn configuration is version-controlled, commit it on a new branch, push it, and open a pull request. Otherwise apply it directly. In Terraform, pass the intake URL through a variable rather than a literal.

​
Set up with the CLI

You can also register the RunsOn log source with the Sazabi CLI (installed and authenticated — see CLI reference). Registering the source mints the same intake URL the dashboard shows above (the public key is embedded in its hostname):
The command prints the intake URL for the new source — copy the whole URL and use it in the configuration shown above. Run sazabi log-sources get <log-source-id> at any time to reprint the endpoint, or sazabi log-sources list to see every log source in the project.

​
Verify

Run any workflow job on RunsOn. Open the Intake page in the Sazabi dashboard and confirm records from the RunsOn source appear within a minute or two, then search for service.name runs-on-flex or runs-on-fleet, or a job_summary event.

​
Troubleshooting

No records — Confirm the stack update finished and the endpoint has no path. The control-plane logs in CloudWatch show OTLP export errors. Control-plane records, but no runner records — The runner needs the otel extra. Run a new job after the change. Only traces and metrics — Check that otel_logs_enabled isn’t false in Terraform. Expecting GitHub job output — Runner logs are RunsOn’s bootstrap and agent log, not the GitHub Actions job log.

​
Further reading

Intake

See per-source ingestion volume and health.

Log sources

Browse the full log source catalog.