Skip to main content

v0.1.0 is out: zero to a live service map in five minutes

avuru obs team
Maintainers

avuru obs v0.1.0 is tagged: install the Helm chart on a fresh Kubernetes cluster and watch a live service map appear in under five minutes, zero app changes — a promise enforced in CI, not a demo.

  • The wedge, shipped and gated. One helm install deploys the full stack — hub, UI, gateway, ClickHouse and the per-node sensor. eBPF instruments every HTTP/gRPC service with no SDKs and no env vars; CI reruns the cluster-to-service-map path on every change with a hard 300-second budget.
  • Four signal tiers. Traces (Full): explorer, heatmap, six trace views and structural diff. Logs (Basic): zero-config stdout/stderr with trace_id correlation. Continuous CPU profiling (Lite, opt-in): flame graphs at ~20 Hz. Infra metrics (Supporting): node & pod health from kubeletstats.
  • Projects. Config-defined or auto-discovered environments with a sidebar switcher, shareable ?project= links and per-project caches (API).
  • Collection controls. Turn collection off per signal, namespace, pod label or node label — and see per-node sensor freshness in Settings → Collection.
  • Deeper trace inspect. Resizable span detail with copyable attributes, per-span tree view, derived span status, service perspective from inside a trace, span-id lookup and filter autocomplete.
  • Drop-in OTLP. Already instrumented? Point your exporters at the gateway and keep everything.

See the full notes on the Releases page and the GitHub release. The trunk is now 0.2.0-SNAPSHOT — next up on the Roadmap: the runtime collection control plane.