In Red Hat OpenShift 5, observability, powered by Red Hat OpenShift Lightspeed, transforms cluster management through 3 core capabilities designed to accelerate incident response while maintaining enterprise-grade safety and human control. After you configure the Alertmanager adapter, Red Hat OpenShift Lightspeed takes it from there: The moment an alert fires, it automatically launches an investigation, functioning like a virtual site reliability engineer (SRE) without requiring a new alerting language.

The agent inspects the cluster state, drafts an initial diagnosis, and stages an actionable remediation plan in a decision queue for review through the console or command-line interface (CLI). Security guardrails verify all state changes require explicit administrator approval executed within sandboxed pods under dynamic role-based access control (RBAC). Should a proposed fix fail, the agent avoids endless loops or unsafe guesses by escalating directly to a human operator.

This technology preview feature is powered by Red Hat OpenShift Lightspeed autonomous agents, available with Red Hat OpenShift Lightspeed operator 2 and Red Hat OpenShift 5. If you'd rather drive the investigation yourself, Red Hat OpenShift Lightspeed synthesizes telemetry across the entire observability stack, including metrics, events, pod logs, and distributed traces, into a single, structured summary to help you resolve multi-service degradation. It organizes findings to immediately isolate the true root cause while filtering out irrelevant noise.

This context allows engineers to focus on high-value decision-making rather than manual data gathering. To get started, open the Red Hat OpenShift Lightspeed side panel, select troubleshooting mode, and give it a try. Install Red Hat OpenShift 4. 22 or later and OpenShift Lightspeed operator 1. The feature is available and usable in production environments with Red Hat OpenShift 5. Beyond troubleshooting, OpenShift Lightspeed turns cluster metrics into dashboards without PromQL.