Skip to main content

Manual service onboarding

Last updated on

Use manual service onboarding to add one Kubernetes workload from the discovery inventory as a Resilience Testing service. This path lets you add a workload after automated onboarding completes, without rerunning the infrastructure onboarding wizard.

Feature flag

Service onboarding is currently behind a feature flag (CHAOS_RISK_SERVICES_ENABLED). Contact your Harness sales representative to enable it for your account.

Before you begin​

  • Discovery agent: A discovery agent that has found the Kubernetes workload you want to onboard.
  • Infrastructure: A Resilience Testing infrastructure associated with the discovery agent.
  • Service access: Permission to create Resilience Testing services in the project.

Use automated service onboarding instead when you connect an infrastructure for the first time and want to discover, scan, and onboard multiple workloads in one run.

Onboard a service manually​

  1. Navigate to Resilience Testing, then select Insights > Services.

  2. Click Resilience Testing Service. The Onboard a Resilience Testing Service wizard opens.

  3. In Onboard Service, select the Discovery Agent that found the workload. The picker shows the number of services each agent discovered and the time of its last discovery.

  4. Click Select Discovery Agent.

  5. In Service, select the workload to onboard. Use the Namespace filter to narrow the list. Each workload shows its namespace, IP address, and port number.

  6. Click Select Service, then click Next.

  7. In Configure Metadata, review the service details:

    • Name: Harness prefills the display name from the workload. Edit it when you need a different name in Resilience Testing.
    • Service ID: Harness generates this immutable identifier from the workload, namespace, and discovery agent.
    • Description: Add optional context about the service.
    • Tags: Add optional labels to organize and filter services.

  8. Click Next.

  9. In Associate Probes, click Associate another probe to attach a probe, then configure its inputs. Repeat this step for each probe the service requires.

  10. Click Finish.

The service appears under Insights > Services after onboarding completes.

Attach probes during manual onboarding​

Manual onboarding does not attach the default probes from the automated onboarding flow. If you finish the wizard without step 9, the service starts with no health checks.

Attach probes during onboarding when you already know which service health signals to validate. Some probes, such as Pod Status Check, have no inputs to configure. Other probes prompt you for target or timing values.

Go to Probes to review the available probe types and how their outcomes affect the resilience score.

Use the correct onboarding path​

  • Multiple discovered workloads: Go to Automated service onboarding to discover, scan, and onboard Kubernetes workloads in bulk.
  • One discovered workload: Use the manual flow on this page.
  • Target outside the Kubernetes inventory: Go to Custom Service Agent to define Linux VMs, Windows VMs, AWS resources, and other custom targets.

Next steps​

  • Go to Services to manage the service and review its Resilience Testing activity.
  • Go to Probes to add or edit health checks.
  • Go to Chaos experiments to target the service with a fault.
Last updated on