Kubernetes Deployment

pg_ripple ships a Helm chart (charts/pg_ripple/) that deploys the batteries-included image — PostgreSQL 18 with pg_ripple, PostGIS, and pgvector pre-installed — on any Kubernetes cluster.

Prerequisites

Kubernetes ≥ 1.25
Helm ≥ 3.10
Persistent volume provisioner (any cloud provider or local-path-provisioner)

Installation

Add the Helm repository

helm repo add pg-ripple https://trickle-labs.github.io/pg-ripple/charts
helm repo update

Install with defaults

helm install pg-ripple pg-ripple/pg-ripple \
  --set postgres.password=mysecretpassword

Install from source

helm install pg-ripple ./charts/pg_ripple \
  --set postgres.password=mysecretpassword

Verify deployment

kubectl get pods -l app.kubernetes.io/name=pg-ripple
kubectl exec -it <pod-name> -- psql -U postgres \
  -c "SELECT extname, extversion FROM pg_extension WHERE extname = 'pg_ripple';"

Values Reference

Key	Default	Description
`replicaCount`	`1`	Number of PostgreSQL Pods
`image.repository`	`ghcr.io/trickle-labs/pg-ripple`	Image repository
`image.tag`	`0.54.0`	Image tag
`image.pullPolicy`	`IfNotPresent`	Image pull policy
`postgres.password`	`ripple`	Superuser password (use a Secret in production)
`postgres.database`	`postgres`	Database to create
`persistence.enabled`	`true`	Enable persistent storage
`persistence.size`	`10Gi`	PVC size
`persistence.storageClass`	`""`	StorageClass (empty = cluster default)
`service.type`	`ClusterIP`	PostgreSQL service type
`service.port`	`5432`	PostgreSQL port
`http.enabled`	`true`	Enable SPARQL HTTP sidecar
`http.service.type`	`ClusterIP`	SPARQL HTTP service type
`http.service.port`	`7878`	SPARQL HTTP port
`ripple.federationEndpoints`	`[]`	SPARQL federation endpoints
`ripple.shacl.shapesConfigMap`	`""`	SHACL shapes ConfigMap
`ripple.llm.apiKeySecret`	`""`	LLM API key Secret name

Common Configurations

Expose SPARQL HTTP externally (LoadBalancer)

helm upgrade pg-ripple ./charts/pg_ripple \
  --set http.service.type=LoadBalancer \
  --set http.service.port=7878

Increase storage

helm upgrade pg-ripple ./charts/pg_ripple \
  --set persistence.size=100Gi \
  --set persistence.storageClass=premium-ssd

Configure federation endpoints

helm upgrade pg-ripple ./charts/pg_ripple \
  --set 'ripple.federationEndpoints[0].name=wikidata' \
  --set 'ripple.federationEndpoints[0].url=https://query.wikidata.org/sparql'

Enable Prometheus monitoring

helm upgrade pg-ripple ./charts/pg_ripple \
  --set metrics.enabled=true \
  --set metrics.serviceMonitor.enabled=true

Health Probes

The chart configures both liveness and readiness probes using pg_isready:

livenessProbe:
  exec:
    command: ["pg_isready", "-U", "postgres"]
  initialDelaySeconds: 30
  periodSeconds: 10

readinessProbe:
  exec:
    command: ["pg_isready", "-U", "postgres"]
  initialDelaySeconds: 5
  periodSeconds: 5

When metrics.serviceMonitor.enabled = true, the chart creates a ServiceMonitor resource for the Prometheus Operator. pg_ripple exposes query stats via pg_stat_statements and OTEL tracing via the pg_ripple.tracing_otlp_endpoint GUC.

Configure the OTEL endpoint:

kubectl exec -it <pod-name> -- psql -U postgres \
  -c "SET pg_ripple.tracing_otlp_endpoint = 'http://otel-collector:4318';"

Future: Kubernetes Operator

A future release will provide a Go operator built with controller-runtime that manages pg_ripple clusters as first-class Kubernetes resources — similar to CloudNativePG but tailored to the RDF workload lifecycle (bulk load, VP merge scheduling, SHACL validation pipelines).

The operator will provide:

RDFTripleStore custom resource with declarative schema management
Automated rolling upgrades with zero downtime
Built-in Prometheus metrics CRDs
Automated SHACL shape deployment via ConfigMap reference

Pre-Installed Extensions

The batteries-included image pre-installs:

Extension	Version	Activate with
pg_ripple	0.54.0	`CREATE EXTENSION pg_ripple;`
PostGIS	3.4.3	`CREATE EXTENSION postgis;`
pgvector	0.7.4	`CREATE EXTENSION vector;`