Backend, storage & operations
View All TagsUntil now, a gateway believed whatever project a sender claimed. Now you can issue per-project ingest API keys and have the key decide instead — with a default that changes nothing until you are ready.
Projects are no longer only a deployment setting: admins can create, rename
and delete them directly in Settings → General, and the switcher updates
to match — while the built-in default and config-defined projects stay
read-only.
You can now explore a live avuru obs without installing anything: one click on "Try the demo" signs you in as a read-only viewer against a running multi-service app, with real data across every module.
avuru obs now ships secure by default: the hub requires login, users carry fixed roles — Admin, Editor, Viewer — granted per project, and enterprise SSO works with any OpenID Connect identity provider (Keycloak, Microsoft Entra ID, Okta, Google, Dex) — in open source, not behind an enterprise tier, and with no auth proxy to deploy.
avuru obs is AGPL-3.0, and the whole licensing story now lives in one place — what it asks of you (almost nothing), why the project chose it, and how it stays sustainable without gating features behind a paid tier.
avuru obs v0.2.0 is tagged. v0.1 proved the wedge — a live service map in under five minutes, zero app changes. v0.2 makes that install one a real team can run every day: secure by default, modular by design, and four new modules that turn the data you already collect into triage, status and accountability.
The new green module turns the pod→service map you already run into per-service energy (Wh) and carbon (gCO2e) — the numbers ESRS E1 / CSRD and internal sustainability targets ask for, without a spreadsheet estimate or a SaaS that wants your telemetry. Energy is measured by CNCF Kepler (from the CPU's RAPL counters); avuru obs already knows which pod is which service, so it attributes joules to the services on your map. Self-hosted, zero app changes, no external API.
Zero-code tracing only counts if it does no harm. The wedge gate now proves it: CI keeps a deliberately fragile workload — tight CPU limit, aggressive liveness probe — Ready under the sensor, and a new opt-in mode lets cautious fleets adopt uprobes label by label.
A new Alerting module closes the loop on service health: when a group, service, or tier crosses into a bad state and stays there, avuru-obs fires a webhook — into Slack, PagerDuty, Opsgenie, Alertmanager, or any endpoint that accepts one. It's derived from the health you already compute; no probes, no new signal.
The service map now shows connection-level health on its edges: per-edge TCP RTT and failed/reset connections, measured in the kernel by OBI — no traces, no SDKs, no app changes. A link that's slow or refusing connections is now visible on the same topology you already watch.