feat: Deploy WrenAI v0.19.7 to Kubernetes with Helm and Kustomize

[rayudu-drivex]

Changes Made

New Files Added

• deployment/kustomizations/kustomization.yaml - Main kustomize configuration
• deployment/kustomizations/base/cm.yaml - ConfigMaps for Wren services
• deployment/kustomizations/base/deploy-wren-ai-service.yaml - AI service deployment
• deployment/kustomizations/base/deploy-wren-ui.yaml - UI deployment
• deployment/kustomizations/base/deploy-wren-engine.yaml - Engine deployment
• deployment/kustomizations/base/deploy-wren-ibis-server.yaml - Ibis server deployment
• deployment/kustomizations/base/svc.yaml - Kubernetes services
• deployment/kustomizations/base/pvc.yaml - Persistent volume claims
• deployment/kustomizations/examples/ingress-wren_example.yaml - Ingress configuration
• deployment/kustomizations/helm-values_postgresql_15.yaml - PostgreSQL Helm values
• deployment/kustomizations/helm-values-qdrant_1.11.0.yaml - Qdrant Helm values

Key Features

• ✅ WrenAI v0.19.7 deployment with all components
• ✅ PostgreSQL 16.3.0 (bitnamilegacy) via Helm chart
• ✅ Qdrant 1.11.0 vector database via Helm chart
• ✅ Persistent storage configuration with AWS gp2 storage class
• ✅ Node selectors and tolerations for pod placement
• ✅ Nginx Ingress with TLS/SSL support
• ✅ cert-manager integration for automatic certificate management
• ✅ ConfigMap updates with v0.19.7 pipeline configurations including:
  • sql_summary
  • sql_breakdown
  • sql_expansion
  • question_recommendation_db_schema_retrieval

Configuration Highlights

• Namespace: wren
• Storage Class: AWS gp2
• Image Versions:
  • wren-bootstrap: 0.1.5
  • wren-engine: 0.14.8
  • wren-ui: 0.24.1
  • wren-ai-service: 0.19.7
  • wren-engine-ibis: 0.14.8
• Node Labels Required: if-wren-ai-pod=true
• TLS: Configured with cert-manager ClusterIssuer

Secrets Required (Not Included - Deploy Separately)

wrenai-postgresql:
  - postgres-password

wrenai-secrets:
  - OPENAI_API_KEY
  - LANGFUSE_PUBLIC_KEY
  - LANGFUSE_SECRET_KEY
  - PG_USERNAME
  - PG_URL
  - USER_UUID
  - POSTHOG_API_KEY
  - POSTHOG_HOST

Deployment Instructions

Prerequisites

1. Kubernetes cluster with storage provisioner
2. kubectl and kustomize installed
3. Helm v4 compatible
4. cert-manager and nginx-ingress-controller installed
5. Node labels: if-wren-ai-pod=true

Deploy to Production

# 1. Create secrets first (manually)
kubectl create namespace wren
kubectl create secret generic wrenai-postgresql -n wren --from-literal=postgres-password=''
kubectl create secret generic wrenai-secrets -n wren --from-literal=OPENAI_API_KEY='' ...

# 2. Label nodes
kubectl label nodes  if-wren-ai-pod=true

# 3. Update values for production
# - Edit examples/ingress-wren_example.yaml with production domain
# - Update helm-values files with production storage sizes
# - Update ClusterIssuer name in ingress

# 4. Deploy
cd deployment/kustomizations
kustomize build . --enable-helm | kubectl apply -n wren -f -

# 5. Verify
kubectl get pods -n wren
kubectl get ingress -n wren
kubectl get certificate -n wren

Configuration Changes for Production

• Update domain in examples/ingress-wren_example.yaml
• Update storage sizes in Helm values files
• Configure appropriate resource limits
• Update ClusterIssuer name for production certificates
• Create secrets with production credentials
• Configure backup strategy for PostgreSQL and Qdrant data

Testing

• ✅ All pods running successfully
• ✅ Services accessible via ingress
• ✅ TLS certificate issued successfully
• ✅ Application UI loads correctly
• ✅ AI service connects to Qdrant and PostgreSQL

Summary by CodeRabbit

• New Features

  • Added new SQL analysis pipelines: query summarization, breakdown, expansion, and database schema recommendations.

• Infrastructure Updates

  • Deployed comprehensive Kubernetes infrastructure with optimized pod scheduling and resource management.
  • Configured storage classes and TLS/SSL certificate handling for secure deployments.

_{✏️ Tip: You can customize this high-level summary in your review settings.}

[coderabbitai]

Caution

Review failed

The pull request is closed.

Walkthrough

This pull request establishes comprehensive Kubernetes infrastructure for the Wren AI system, including submodule URL updates, new SQL-related service pipelines, pod scheduling constraints across deployments, persistent volume configuration, and extensive manifest definitions for multi-component deployment with services, storage, and networking.

Changes

Cohort / File(s)	Summary
Submodule Configuration .gitmodules	Updated submodule URL from SSH (git@...) to HTTPS scheme with minor formatting adjustments.
Service Pipeline Configuration deployment/kustomizations/base/cm.yaml	Added four new SQL-related pipelines: sql_summary, sql_breakdown, sql_expansion, and question_recommendation_db_schema_retrieval (with embedder and document_store references).
Deployment Pod Scheduling deployment/kustomizations/base/deploy-wren-*.yaml (wren-ai-service, wren-engine, wren-ibis-server, wren-ui)	Added nodeSelector (if-wren-ai-pod: 'true') and tolerations for if-wren-ai-pod and if-monitoring-pod taints across all four deployment manifests to constrain pod placement.
Persistent Volume Configuration deployment/kustomizations/base/pvc.yaml	Set explicit storageClassName to gp2, replacing unspecified/commented vsphere configuration.
Example Configurations deployment/kustomizations/examples/certificate-wren_example.yaml, deployment/kustomizations/examples/ingress-wren_example.yaml	Updated certificate dnsNames and issuerRef from myhost.net example to wren.drivex.dev; updated ingress host and TLS config, added cert-manager and nginx annotations for SSL/proxy timeouts.
Helm Values deployment/kustomizations/helm-values-qdrant_1.11.0.yaml, deployment/kustomizations/helm-values_postgresql_15.yaml	Added resource limits/requests, nodeSelector, tolerations, storageClassName (gp2), security context fields, topology constraints for Qdrant; added image config, persistence, nodeSelector, and tolerations for PostgreSQL.
Kustomization Resources deployment/kustomizations/kustomization.yaml	Uncommented example ingress resource reference to enable it in kustomization.
Kubernetes Manifests deployment/kustomizations/manifests.yaml, deployment/kustomizations/output.yaml	Added 1261+ lines defining complete deployment topology: ServiceAccounts (wren-postgresql, wren-qdrant), ConfigMaps (wren-ai-service-config with multi-model pipelines, wren-config, wren-qdrant), Services (wren-ai-service-svc, wren-engine-svc, wren-ibis-server-svc, wren-ui-svc, PostgreSQL, Qdrant variants), StatefulSets for PostgreSQL and Qdrant, Deployments for AI service/engine/IBIS/UI, PersistentVolumeClaims, Ingress (wren-ui-ingress routing to wren.drivex.dev), NetworkPolicy for PostgreSQL, and PodDisruptionBudget.

Estimated code review effort

🎯 4 (Complex) | ⏱️ ~45 minutes

Possibly related PRs

• chore(wren-ai-service): Add Dedicated Pipes for Question Recommendation Components #1114 — Directly overlaps with question_recommendation_db_schema_retrieval pipeline configuration in the same manifest.
• feat(wren-ai-service): Add SQL Functions Support in Query Generation Pipeline (ai-env-changed) #1410 — Modifies the same deployment/kustomizations/base/cm.yaml with additional SQL-related pipeline configurations.
• chore(wren-ai-service): improve ai service #1009 — Introduces SQL pipeline postprocessors that correspond to new pipelines added in this manifest configuration.

Suggested labels

deployment, wren-ai-service, ai-env-changed

Suggested reviewers

• imAsterSun
• yichieh-lu

Poem

🐰 A cluster takes shape, so grand and so wide,
With pipelines for SQL and nowhere to hide,
StatefulSets dance with tolerations so true,
Nodes bow to selectors—this orchestration's new!
From Postgres to Qdrant, all services align,
The Wren AI stack rises—infrastructure divine! ✨

✨ Finishing touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}