Self-hosting website analytics often requires complex orchestration of separate transactional and analytical databases. This configuration deploys Plausible CE on Kubernetes with PostgreSQL for operational data and ClickHouse for analytics events, producing a GDPR-compliant, cookie-free analytics stack accessible at a local domain.

What's Included

Component	Type	Port	Role
plausible	Namespace	-	Isolates all resources under plausible
plausible-pg-pvc	PersistentVolumeClaim	-	10 GB for PostgreSQL data
plausible-ch-pvc	PersistentVolumeClaim	-	20 GB for ClickHouse events
plausible-pg-config	ConfigMap	-	PostgreSQL env vars
plausible-pg-secret	Secret	-	PostgreSQL password
plausible-ch-config	ConfigMap	-	ClickHouse env vars
plausible-ch-secret	Secret	-	ClickHouse password
plausible-ch-files	ConfigMap	-	config.xml and users.xml as mounted files
plausible-app-config	ConfigMap	-	Plausible env vars (BASE_URL, DB URLs)
plausible-app-secret	Secret	-	SECRET_KEY_BASE
plausible-postgres	StatefulSet	5432	PostgreSQL 16: OLTP store for users and config
plausible-postgres-service	Service (ClusterIP)	5432	Internal DNS only
plausible-clickhouse	Deployment	8123/9000	ClickHouse: OLAP store for events
plausible-clickhouse-service	Service (ClusterIP)	8123/9000	Internal DNS only
plausible	Deployment	8000	App server with 3 init containers
plausible-service	Service (ClusterIP)	8000	Internal: Ingress routes here
nginx	IngressClass	-	Declares NGINX as default ingress class
plausible-ingress	Ingress	80	Routes plausible.local to app service
plausible-hpa	HorizontalPodAutoscaler	-	Scales Plausible pods from 1 to 3 replicas
plausible-quota	ResourceQuota	-	Namespace resource cap
plausible-limits	LimitRange	-	Per-container defaults
plausible-postgres-netpol	NetworkPolicy	5432	PostgreSQL: Plausible pods only
plausible-clickhouse-netpol	NetworkPolicy	8123/9000	ClickHouse: Plausible pods only
plausible-app-netpol	NetworkPolicy	8000	Plausible: NGINX ingress only

Architecture Overview

PostgreSQL stores users, sites, and goals, while ClickHouse stores every pageview for fast aggregation. The Plausible pod is the only component that talks to both databases, a constraint enforced by NetworkPolicies. Three sequential init containers ensure both databases are ready and migrations complete before the main application starts. NGINX Ingress forwards real client IPs for accurate geo data parsing.

Prerequisites

• NGINX Ingress Controller deployed in the ingress-nginx namespace
• 6 GB RAM available to the cluster node
• 127.0.0.1 plausible.local added to your hosts file
• KubeKanvas CLI installed and running on your computer (Optional, if you want to use one-click deployment)

How to Deploy

1. Confirm the NGINX Ingress Controller is running in the ingress-nginx namespace.
2. Generate a SECRET_KEY_BASE value using openssl rand -hex 64, base64-encode it, and replace the placeholder in the plausible-app-secret manifest.
3. Update the storage paths in both PVCs to match your operating system filesystem structure.
4. Deploy the template to your cluster via the Play button in the top right bar. If you prefer to deploy manually, download the YAML and apply it with kubectl.
5. Wait for all pods to reach Running status. You can monitor progress in the Release Monitor screen.

How to Test

1. Run kubectl get pods -n plausible and confirm all three pods show a status of Running.
2. Open http://plausible.local in your browser and confirm the Plausible admin setup screen loads.
3. Create your admin account, add a test site, and confirm the analytics dashboard renders.
4. Execute an HTTP GET request against http://plausible.local/api/health and confirm the JSON response shows healthy.

Use Cases

• Privacy-first analytics: Replace Google Analytics with a cookie-free, GDPR-compliant alternative where all data stays on your own infrastructure.
• Air-gapped deployments: Run an analytics platform that requires no external internet access to record or display traffic data once deployed.
• Multi-site tracking: Aggregating multiple tracking scripts into one instance using isolated dashboards and goals.
• High-volume event storage: Scale from a personal blog to a high-traffic site using ClickHouse as an analytical data store without changing the architecture.

Summary

This template configures a 24-resource Kubernetes stack running Plausible CE with PostgreSQL for operational data and ClickHouse for analytics events, exposed via NGINX Ingress with per-component ConfigMap, Secret, and NetworkPolicy isolation.