ocp-distributed-tracing

name	ocp-distributed-tracing
metadata	{"author":"rhoai3-demo","version":"1.0.0","platform-family":"ocp","platform-baseline":"repo","ocp-baseline":"repo","skill-group":"OpenShift Platform"}
description	Use when documenting, reviewing, or rebuilding Red Hat OpenShift distributed tracing platform on OpenShift from official Red Hat documentation: Tempo Operator installation, TempoStack, TempoMonolithic, object storage secrets, tenants, read/write RBAC, gateway authentication and authorization, Jaeger UI, distributed tracing UI plugin, TempoStack sizing, retention, receiver TLS, query RBAC, monitoring and alerts, Tempo Operator upgrades, removal, and must-gather troubleshooting. Do NOT use for OpenTelemetry Collector or workload instrumentation configuration; use ocp-opentelemetry. Do NOT use for generic OCP observability, logging, or default monitoring posture; use ocp-observability. Do NOT invent Tempo CR fields, storage settings, tenant configuration, or Technology Preview support posture.

OCP Distributed Tracing

Use this skill to ground Red Hat OpenShift distributed tracing platform work in the official product documentation pinned in docs/PLATFORM_BASELINE.md.

Source Grounding

Read references/source-capture.md before using product behavior. Official Red Hat documentation is product authority. This skill captures Red Hat OpenShift distributed tracing platform 3.9 and adapts it to the repo's OCP observability skill model.

Distributed tracing is adjacent to, but not the same as, OpenTelemetry collection. Use this skill for Tempo Operator, TempoStack, TempoMonolithic, trace storage, tenants, trace query/read/write RBAC, Jaeger UI, distributed tracing UI plugin, and Tempo troubleshooting. Use ocp-opentelemetry for sending telemetry to Tempo.

Demo Distributed Tracing Posture

For this AWS-hosted RHOAI demo:

• Prefer TempoStack for GitOps-managed distributed tracing back ends.
• Treat TempoMonolithic as Technology Preview and use it only for small demonstrations, tests, or proof-of-concept paths where non-production status is explicit.
• Do not deploy a TempoStack without object storage, object storage credentials, tenants, and tenant read/write RBAC.
• Do not claim tracing is available until a backend, Collector/export path, tenant authorization, and validation path are implemented.
• Keep object storage credentials, cloud role details, and trace data samples out of Git.
• Use Red Hat build of OpenTelemetry to forward traces to Tempo; do not duplicate Collector configuration in this skill.
• Treat Jaeger UI and distributed tracing UI plugin exposure as access-control surfaces that require explicit tenant and route review.

Product Model

Use the official docs to frame:

• Distributed tracing records the path of requests through microservices and presents them as traces made of spans.
• A span represents a logical unit of work with operation name, start time, duration, and potentially tags and logs.
• Distributed tracing supports monitoring distributed transactions, optimizing performance and latency, and root cause analysis.
• Red Hat build of OpenTelemetry forwards traces to TempoStack.
• The distributed tracing UI plugin of the Cluster Observability Operator can provide a UI for distributed tracing.
• TempoStack is the scalable Tempo deployment path and is configured to receive Jaeger Thrift over HTTP and OTLP.
• TempoMonolithic runs Tempo components in one container, does not scale horizontally, and is Technology Preview in the captured docs.

Workflow

1. Confirm the active OCP and distributed tracing baselines in docs/PLATFORM_BASELINE.md.
2. Read references/official-doc-extraction.md.
3. Identify whether the task concerns:
   • Tempo Operator installation or upgrade
   • object storage prerequisites and storage secrets
   • tenant definition and read/write RBAC
   • TempoStack installation, sizing, retention, gateway, Jaeger UI, or monitoring
   • TempoMonolithic Technology Preview demo/test usage
   • receiver TLS, query RBAC, taints, tolerations, monitoring, or alerts
   • distributed tracing UI plugin integration
   • Tempo removal or troubleshooting
   • OpenTelemetry Collector forwarding, which belongs in ocp-opentelemetry
4. Verify API versions, CRDs, fields, namespaces, Operator channel, object storage settings, secrets, RBAC, tenants, routes, TLS, and Technology Preview posture before authoring manifests.
5. For live operations, use the repo environment guard and pair this skill with env-troubleshoot, env-manage-resources, or env-deploy-and-evaluate.
6. Validate the output with references/validation-checklist.md.

Related Skills

• Use ocp-opentelemetry for OpenTelemetry Collector, Instrumentation, workload annotations, and telemetry forwarding into Tempo.
• Use ocp-observability for the broad OCP observability model, default monitoring stack, logging overview, and COO boundary.
• Use rhoai-observability for RHOAI observability stack enablement and dashboard behavior.
• Use rhoai-guardrails-safety for guardrails-specific spans, detector metrics, and guardrails telemetry semantics.
• Use project-gitops-authoring for repo-specific GitOps implementation.
• Use project-manifest-review for structural and security review of Kubernetes and OpenShift manifests.

References

• references/source-capture.md
• references/official-doc-extraction.md
• references/validation-checklist.md
• examples/distributed-tracing-review-patterns.md