plone.observability

Kubernetes-style health probes, a pluggable Prometheus and JSON metrics endpoint, and optional OpenTelemetry tracing for Plone running in containers.

Plone's built-in @@ok view answers OK whether the database is reachable or not, right up until the process is too busy to answer at all. plone.observability gives an orchestrator the real signals it needs: separate liveness, readiness, and startup probes on a dedicated port that stays answerable under load, plus a metrics endpoint your monitoring stack can scrape.

Features

• Liveness, readiness, and startup probes on a separate HTTP port
• A pluggable @@metrics endpoint with Prometheus and JSON output
• Request, ZODB, and content metrics with low-cardinality labels
• Optional OpenTelemetry tracing that honors the standard OTEL_* environment variables
• Extensible via the Zope Component Architecture: custom health checks, metric providers, and formatters

Installation

Add plone.observability to your dependencies, include its ZCML, and wire the WSGI filters. See the installation guide for the full steps.

[project]
dependencies = [
    "plone.observability",
]

<include package="plone.observability" />

Documentation

Full documentation is at https://plone.github.io/plone.observability/:

• How-to guides — install, configure Kubernetes probes, scrape with Prometheus, enable tracing, extend via ZCA
• Reference — configuration, health endpoints, metrics, tracing, interfaces
• Explanation — why three probes, why a separate port, label cardinality, tracing design

For LLM context, see llms.txt.

License

GPL-2.0-only. See LICENSE.