Files

Urtzi Alfaro aeff6b1537 Add new infra architecture 13

2026-01-21 23:16:19 +01:00

112 KiB

Raw Blame History

Bakery-IA Pilot Launch Guide

Complete guide for deploying to production for a 10-tenant pilot program

Last Updated: 2026-01-20 Target Environment: clouding.io VPS with MicroK8s Estimated Cost: €41-81/month Time to Deploy: 3-5 hours (first time, including fixes) Status: ⚠️ REQUIRES PRE-DEPLOYMENT FIXES - See Production VPS Deployment Fixes Version: 3.0

Executive Summary
Infrastructure Architecture Overview
⚠️ CRITICAL: Pre-Deployment Fixes
Pre-Launch Checklist
VPS Provisioning
Infrastructure Setup
Domain & DNS Configuration
TLS/SSL Certificates
Email & Communication Setup
Kubernetes Deployment
Configuration & Secrets
Database Migrations
CI/CD Infrastructure Deployment
Mailu Email Server Deployment
Nominatim Geocoding Service
SigNoz Monitoring Deployment
Verification & Testing
Post-Deployment

Executive Summary

What You're Deploying

A complete multi-tenant SaaS platform with:

18 microservices (auth, tenant, ML forecasting, inventory, sales, orders, etc.)
14 PostgreSQL databases with TLS encryption
Redis cache with TLS
RabbitMQ message broker
Monitoring stack (Prometheus, Grafana, AlertManager)
Full security (TLS, RBAC, audit logging)

Total Cost Breakdown

Service	Provider	Monthly Cost
VPS Server (20GB RAM, 8 vCPU, 200GB SSD)	clouding.io	€40-80
Domain	Namecheap/Cloudflare	€1.25 (€15/year)
Email	Zoho Free / Gmail	€0
WhatsApp API	Meta Business	€0 (1k free conversations)
DNS	Cloudflare	€0
SSL	Let's Encrypt	€0
TOTAL		€41-81/month

Timeline

Phase	Duration	Description
Pre-Launch Setup	1-2 hours	Domain, VPS provisioning, accounts setup
Infrastructure Setup	1 hour	MicroK8s installation, firewall config
Deployment	30-60 min	Deploy all services and databases
Verification	30-60 min	Test everything works
Total	2-4 hours	First-time deployment

Infrastructure Architecture Overview

Component Layers

The Bakery-IA platform is organized into distinct infrastructure layers, each with specific deployment dependencies.

┌─────────────────────────────────────────────────────────────────────────────┐
│                         LAYER 6: APPLICATION                                 │
│  Frontend │ Gateway │ 18 Microservices │ CronJobs & Workers                 │
├─────────────────────────────────────────────────────────────────────────────┤
│                         LAYER 5: MONITORING                                  │
│  SigNoz (Unified Observability) │ AlertManager │ OTel Collector             │
├─────────────────────────────────────────────────────────────────────────────┤
│                         LAYER 4: PLATFORM SERVICES (Optional)               │
│  Mailu (Email) │ Nominatim (Geocoding) │ CI/CD (Tekton, Flux, Gitea)       │
├─────────────────────────────────────────────────────────────────────────────┤
│                         LAYER 3: DATA & STORAGE                              │
│  PostgreSQL (18 DBs) │ Redis │ RabbitMQ │ MinIO                             │
├─────────────────────────────────────────────────────────────────────────────┤
│                         LAYER 2: NETWORK & SECURITY                          │
│  Unbound DNS │ CoreDNS │ Ingress Controller │ Cert-Manager │ TLS           │
├─────────────────────────────────────────────────────────────────────────────┤
│                         LAYER 1: FOUNDATION                                  │
│  Namespaces │ Storage Classes │ RBAC │ ConfigMaps │ Secrets                 │
├─────────────────────────────────────────────────────────────────────────────┤
│                         LAYER 0: KUBERNETES CLUSTER                          │
│  MicroK8s (Production) │ Kind (Local Dev) │ EKS (AWS Alternative)           │
└─────────────────────────────────────────────────────────────────────────────┘

Deployment Order & Dependencies

Components must be deployed in a specific order due to dependencies:

1. Namespaces (bakery-ia, tekton-pipelines, flux-system)
   ↓
2. Cert-Manager & ClusterIssuers
   ↓
3. TLS Certificates (internal + ingress)
   ↓
4. Unbound DNS Resolver (required for Mailu DNSSEC)
   ↓
5. CoreDNS Configuration (forward to Unbound)
   ↓
6. Ingress Controller & Resources
   ↓
7. Data Layer: PostgreSQL, Redis, RabbitMQ, MinIO
   ↓
8. Database Migrations
   ↓
9. Application Services (18 microservices)
   ↓
10. Gateway & Frontend
    ↓
11. (Optional) CI/CD: Gitea → Tekton → Flux
    ↓
12. (Optional) Mailu Email Server
    ↓
13. (Optional) Nominatim Geocoding
    ↓
14. (Optional) SigNoz Monitoring

Infrastructure Components Summary

Component	Purpose	Required	Namespace
MicroK8s	Kubernetes cluster	Yes	-
Cert-Manager	TLS certificate management	Yes	cert-manager
Ingress-Nginx	External traffic routing	Yes	ingress
PostgreSQL	18 service databases	Yes	bakery-ia
Redis	Caching & sessions	Yes	bakery-ia
RabbitMQ	Message broker	Yes	bakery-ia
MinIO	Object storage (ML models)	Yes	bakery-ia
Unbound DNS	DNSSEC resolver	For Mailu	bakery-ia
Mailu	Self-hosted email server	Optional	bakery-ia
Nominatim	Geocoding service	Optional	bakery-ia
Gitea	Git server + container registry	Optional	gitea
Tekton	CI/CD pipelines	Optional	tekton-pipelines
Flux CD	GitOps deployment	Optional	flux-system
SigNoz	Unified observability	Recommended	bakery-ia

Quick Reference: What to Deploy

Minimal Production Setup:

Kubernetes cluster + addons
Core infrastructure (databases, cache, broker)
Application services
External email (Zoho/Gmail)

Full Production Setup (Recommended):

Everything above, plus:
Mailu (self-hosted email)
SigNoz (monitoring)
CI/CD (Gitea + Tekton + Flux)
Nominatim (if geocoding needed)

⚠️ CRITICAL: Pre-Deployment Configuration

READ THIS FIRST: Review and complete these configuration steps before deploying to production.

Infrastructure Architecture (Updated)

The infrastructure has been reorganized with the following structure:

infrastructure/
├── environments/              # Environment-specific configs
│   ├── common/configs/        # Shared ConfigMaps and Secrets
│   │   ├── configmap.yaml     # Application configuration
│   │   ├── secrets.yaml       # All secrets (database, JWT, Redis, etc.)
│   │   └── kustomization.yaml
│   ├── dev/k8s-manifests/     # Development Kustomization
│   └── prod/k8s-manifests/    # Production Kustomization & patches
├── platform/                  # Platform-level infrastructure
│   ├── cert-manager/          # TLS certificate issuers (Let's Encrypt)
│   ├── networking/ingress/    # NGINX ingress (base + overlays)
│   ├── storage/               # PostgreSQL, Redis, MinIO
│   ├── gateway/               # API Gateway service
│   └── mail/mailu-helm/       # Email server (Helm chart)
├── services/                  # Application services
│   ├── databases/             # 19 PostgreSQL database instances
│   └── microservices/         # 19 microservices
├── cicd/                      # CI/CD (deployed via Helm, NOT kustomize)
│   ├── gitea/                 # Git server + container registry
│   ├── tekton-helm/           # CI pipelines
│   └── flux/                  # GitOps deployment
└── monitoring/signoz/         # SigNoz observability (via Helm)

🔴 Configuration Status

Item	Status	File Location
Production Secrets	✅ Configured	`infrastructure/environments/common/configs/secrets.yaml`
Cert-Manager Email	✅ Configured	`infrastructure/platform/cert-manager/cluster-issuer-production.yaml`
SigNoz Namespace	✅ Uses bakery-ia	`infrastructure/environments/prod/k8s-manifests/kustomization.yaml`
imagePullSecrets	✅ Auto-patched	Production kustomization adds `gitea-registry-secret` automatically
Image Tags	⚠️ Update for releases	`infrastructure/environments/prod/k8s-manifests/kustomization.yaml`
Stripe Keys	⚠️ Configure before launch	ConfigMap + Secrets
Pilot Coupon	✅ Auto-seeded	`app/jobs/startup_seeder.py`

Required Configuration Changes

1. imagePullSecrets - ✅ AUTOMATICALLY HANDLED

Status: ✅ The production kustomization automatically patches all workloads File: infrastructure/environments/prod/k8s-manifests/kustomization.yaml

The production overlay adds gitea-registry-secret to all Deployments, StatefulSets, Jobs, and CronJobs via Kustomize patches:

patches:
  - target:
      kind: Deployment
    patch: |-
      - op: add
        path: /spec/template/spec/imagePullSecrets
        value:
          - name: gitea-registry-secret

Note: The gitea-registry-secret is created by infrastructure/cicd/gitea/sync-registry-secret.sh after Gitea deployment.

2. Update Image Tags to Semantic Versions (FOR RELEASES)

Why: Using 'latest' causes non-deterministic deployments Impact if skipped: Unpredictable behavior, impossible rollbacks File: infrastructure/environments/prod/k8s-manifests/kustomization.yaml

For production releases, update the images: section from latest to semantic versions.

3. Production Secrets - ✅ ALREADY CONFIGURED

Status: ✅ Strong production secrets have been pre-generated File: infrastructure/environments/common/configs/secrets.yaml

Pre-configured secrets include:

19 database passwords (24-character URL-safe random strings)
JWT secrets (256-bit cryptographically secure)
Redis password (24-character random string)
RabbitMQ credentials
PostgreSQL monitoring user for SigNoz metrics collection

4. Cert-Manager Email - ✅ ALREADY CONFIGURED

Status: ✅ Email set to admin@bakewise.ai File: infrastructure/platform/cert-manager/cluster-issuer-production.yaml

5. Update Stripe Keys (HIGH PRIORITY)

Why: Payment processing requires production Stripe keys Impact if skipped: Payments will use test mode (no real charges)

ConfigMap (infrastructure/environments/common/configs/configmap.yaml):

VITE_STRIPE_PUBLISHABLE_KEY: "pk_live_XXXXXXXXXXXXXXXXXXXX"

Secrets (infrastructure/environments/common/configs/secrets.yaml):

# Add to payment-secrets section (base64 encoded)
STRIPE_SECRET_KEY: <base64-encoded-secret-key>
STRIPE_WEBHOOK_SECRET: <base64-encoded-webhook-secret>

Get your keys from: https://dashboard.stripe.com/apikeys

6. Pilot Coupon Configuration - ✅ AUTO-SEEDED

Status: ✅ Automatically created when tenant-service starts How it works: app/jobs/startup_seeder.py creates the PILOT2025 coupon

Default pilot settings (in configmap, can be customized):

VITE_PILOT_MODE_ENABLED: "true" - Enables pilot UI features
VITE_PILOT_COUPON_CODE: "PILOT2025" - Coupon code for 3 months free
VITE_PILOT_TRIAL_MONTHS: "3" - Trial extension duration

✅ Already Correct (No Changes Needed)

Storage Class - Uses MicroK8s default storage provisioner
Domain Names - bakewise.ai configured in production overlay
Service Types - ClusterIP + Ingress is correct architecture
Network Policies - Defined in infrastructure/platform/security/network-policies/
SigNoz Namespace - ✅ Uses bakery-ia namespace (unified with application)
OTEL Configuration - ✅ Pre-configured for SigNoz in production patches
Replica Counts - ✅ Production replicas defined in kustomization (2-3 per service)

Step-by-Step Configuration Script

Run these commands on your local machine before deployment:

# Navigate to repository root
cd /path/to/bakery-ia

# ========================================
# STEP 1: Verify Infrastructure Structure
# ========================================
echo "Step 1: Verifying new infrastructure structure..."
echo "Checking directories..."
ls -d infrastructure/environments/common/configs/ && echo "✅ Common configs"
ls -d infrastructure/environments/prod/k8s-manifests/ && echo "✅ Prod kustomization"
ls -d infrastructure/platform/cert-manager/ && echo "✅ Cert-manager"
ls -d infrastructure/cicd/gitea/ && echo "✅ Gitea CI/CD"

# ========================================
# STEP 2: Update Image Tags (for releases)
# ========================================
echo -e "\nStep 2: Updating image tags for release..."
export VERSION="1.0.0"  # Change this to your version

# Update application image tags in production kustomization
sed -i.bak "s/newTag: latest/newTag: v${VERSION}/g" \
  infrastructure/environments/prod/k8s-manifests/kustomization.yaml

# Verify (show first 10 image entries)
echo "Current image tags:"
grep -A 1 "name: bakery/" infrastructure/environments/prod/k8s-manifests/kustomization.yaml | head -20

# ========================================
# STEP 3: Verify Production Secrets
# ========================================
echo -e "\nStep 3: Verifying production secrets..."
echo "✅ Production secrets are pre-configured with strong passwords:"
echo "   - 19 database passwords (24-char URL-safe random)"
echo "   - JWT secrets (256-bit cryptographically secure)"
echo "   - Redis password (24-char random)"
echo "   - RabbitMQ credentials"
echo "   - PostgreSQL monitoring user for SigNoz"
echo ""
echo "Location: infrastructure/environments/common/configs/secrets.yaml"

# Quick verification
grep -c "_DB_PASSWORD:" infrastructure/environments/common/configs/secrets.yaml
echo "database password entries found"

# ========================================
# STEP 4: Verify Cert-Manager Email
# ========================================
echo -e "\nStep 4: Verifying cert-manager email..."
grep "email:" infrastructure/platform/cert-manager/cluster-issuer-production.yaml
# Should show: email: admin@bakewise.ai

# ========================================
# STEP 5: Verify imagePullSecrets Patch
# ========================================
echo -e "\nStep 5: Verifying imagePullSecrets configuration..."
grep -A 5 "gitea-registry-secret" infrastructure/environments/prod/k8s-manifests/kustomization.yaml && \
  echo "✅ imagePullSecrets patch configured" || \
  echo "⚠️ WARNING: imagePullSecrets patch missing"

# ========================================
# STEP 6: Configure Stripe Keys (MANUAL)
# ========================================
echo -e "\nStep 6: Stripe Configuration..."
echo "================================================================"
echo "⚠️  MANUAL STEP REQUIRED"
echo ""
echo "1. Edit ConfigMap:"
echo "   File: infrastructure/environments/common/configs/configmap.yaml"
echo "   Add:  VITE_STRIPE_PUBLISHABLE_KEY: \"pk_live_XXXX\""
echo ""
echo "2. Edit Secrets:"
echo "   File: infrastructure/environments/common/configs/secrets.yaml"
echo "   Add to payment-secrets (base64 encoded):"
echo "     STRIPE_SECRET_KEY: <base64-encoded>"
echo "     STRIPE_WEBHOOK_SECRET: <base64-encoded>"
echo ""
echo "Get keys from: https://dashboard.stripe.com/apikeys"
echo "================================================================"
read -p "Press Enter when Stripe keys are configured..."

# ========================================
# STEP 7: Validate Kustomization Build
# ========================================
echo -e "\nStep 7: Validating Kustomization..."
cd infrastructure/environments/prod/k8s-manifests
kustomize build . > /dev/null 2>&1 && \
  echo "✅ Kustomization builds successfully" || \
  echo "⚠️ WARNING: Kustomization build failed"
cd - > /dev/null

# ========================================
# FINAL VALIDATION
# ========================================
echo -e "\n========================================"
echo "Pre-Deployment Configuration Complete!"
echo "========================================"
echo ""
echo "Validation Checklist:"
echo "  ✅ Infrastructure structure verified"
echo "  ✅ Image tags updated to v${VERSION}"
echo "  ✅ Production secrets pre-configured"
echo "  ✅ Cert-manager email: admin@bakewise.ai"
echo "  ✅ imagePullSecrets auto-patched via Kustomize"
echo "  ⚠️  Stripe keys configured (manual verification)"
echo "  ✅ Pilot coupon auto-seeded on startup"
echo ""
echo "Next Steps:"
echo "  1. Deploy CI/CD: Gitea, Tekton, Flux (via Helm)"
echo "  2. Push images to Gitea registry"
echo "  3. Apply Kustomization to cluster"

Manual Verification

After running the script above:

Verify production secrets are configured:

# Check secrets file has strong passwords
head -80 infrastructure/environments/common/configs/secrets.yaml
# Should show base64-encoded passwords for all 19 databases

Check image tags in production overlay:

grep "newTag:" infrastructure/environments/prod/k8s-manifests/kustomization.yaml | head -10
# For releases: should show v1.0.0 (your version)
# For development: 'latest' is acceptable

Verify imagePullSecrets patch:

grep -B 2 -A 6 "imagePullSecrets" infrastructure/environments/prod/k8s-manifests/kustomization.yaml
# Should show patches for Deployment, StatefulSet, Job, CronJob

Verify OTEL/SigNoz configuration:

grep "OTEL_EXPORTER" infrastructure/environments/prod/k8s-manifests/kustomization.yaml
# Should show: http://signoz-otel-collector.bakery-ia.svc.cluster.local:4317

Test Kustomize build:

cd infrastructure/environments/prod/k8s-manifests
kustomize build . | kubectl apply --dry-run=client -f -
# Should complete without errors

Key File Locations Reference

Configuration	File Path
ConfigMap	`infrastructure/environments/common/configs/configmap.yaml`
Secrets	`infrastructure/environments/common/configs/secrets.yaml`
Prod Kustomization	`infrastructure/environments/prod/k8s-manifests/kustomization.yaml`
Cert-Manager Issuer	`infrastructure/platform/cert-manager/cluster-issuer-production.yaml`
Ingress	`infrastructure/platform/networking/ingress/base/ingress.yaml`
Gitea Values	`infrastructure/cicd/gitea/values.yaml`
Mailu Values	`infrastructure/platform/mail/mailu-helm/values.yaml`

Pre-Launch Checklist

Required Accounts & Services

Domain Name
- Register at Namecheap or Cloudflare (€10-15/year)
- Suggested: bakeryforecast.es or bakery-ia.com
VPS Account
- Sign up at clouding.io
- Payment method configured
Email Service - Self-hosted Mailu with Mailgun relay
- Mailu deployed via Helm chart (see Mailu Email Server Deployment)
- Mailgun account for outbound relay (improves deliverability)
- DNS records configured (MX, SPF, DKIM, DMARC)
WhatsApp Business API
- Create Meta Business Account (free)
- Verify business identity
- Phone number ready (non-VoIP)
DNS Access
- Cloudflare account (free, recommended)
- Or domain registrar DNS panel access
Container Registry (Choose ONE)
- Option A: Docker Hub account (recommended)
- Option B: GitHub Container Registry
- Option C: MicroK8s built-in registry

Required Tools on Local Machine

# Verify you have these installed:
kubectl version --client
docker --version
git --version
ssh -V
openssl version

# Install if missing (macOS):
brew install kubectl docker git openssh openssl

Repository Setup

# Clone the repository
git clone https://github.com/yourusername/bakery-ia.git
cd bakery-ia

# Verify structure
ls infrastructure/kubernetes/overlays/prod/

VPS Provisioning

Recommended Configuration

For 10-tenant pilot program:

RAM: 20 GB
CPU: 8 vCPU cores
Storage: 200 GB NVMe SSD (triple replica)
Network: 1 Gbps connection
OS: Ubuntu 22.04 LTS
Monthly Cost: €40-80 (check current pricing)

Why These Specs?

Memory Breakdown:

Application services: 14.1 GB
Databases (18 instances): 4.6 GB
Infrastructure (Redis, RabbitMQ): 0.8 GB
Gateway/Frontend: 1.8 GB
Monitoring: 1.5 GB
System overhead: ~3 GB
Total: ~26 GB capacity needed, 20 GB is sufficient with HPA

Storage Breakdown:

Databases: 36 GB (18 × 2GB)
ML Models: 10 GB
Redis: 1 GB
RabbitMQ: 2 GB
Prometheus metrics: 20 GB
Container images: ~30 GB
Growth buffer: 100 GB
Total: 199 GB

Provisioning Steps

Create VPS at clouding.io:

1. Log in to clouding.io dashboard
2. Click "Create New Server"
3. Select:
   - OS: Ubuntu 22.04 LTS
   - RAM: 20 GB
   - CPU: 8 vCPU
   - Storage: 200 GB NVMe SSD
   - Location: Barcelona (best for Spain)
4. Set hostname: bakery-ia-prod-01
5. Add SSH key (or use password)
6. Create server

Note your server details:

# Save these for later:
VPS_IP="YOUR_VPS_IP_ADDRESS"
VPS_ROOT_PASSWORD="YOUR_ROOT_PASSWORD"  # If not using SSH key

Initial SSH connection:

# Test connection
ssh root@$VPS_IP

# Update system
apt update && apt upgrade -y

Infrastructure Setup

Step 1: Install MicroK8s

Using MicroK8s for production VPS deployment on clouding.io

# SSH into your VPS
ssh root@$VPS_IP

# Update system
apt update && apt upgrade -y

# Install MicroK8s
snap install microk8s --classic --channel=1.28/stable

# Add your user to microk8s group
usermod -a -G microk8s $USER
chown -f -R $USER ~/.kube
newgrp microk8s

# Verify installation
microk8s status --wait-ready

Step 2: Enable Required MicroK8s Addons

All required components are available as MicroK8s addons:

# Enable core addons
microk8s enable dns                  # DNS resolution within cluster
microk8s enable hostpath-storage     # Provides microk8s-hostpath storage class
microk8s enable ingress              # Nginx ingress controller (uses class "public")
microk8s enable cert-manager         # Let's Encrypt SSL certificates
microk8s enable metrics-server       # For HPA autoscaling
microk8s enable rbac                 # Role-based access control

# Setup kubectl alias
echo "alias kubectl='microk8s kubectl'" >> ~/.bashrc
source ~/.bashrc

# Verify all components are running
kubectl get nodes
# Should show: Ready

kubectl get storageclass
# Should show: microk8s-hostpath (default)

kubectl get pods -A
# Should show pods in: kube-system, ingress, cert-manager namespaces

# Verify ingress controller is running
kubectl get pods -n ingress
# Should show: nginx-ingress-microk8s-controller-xxx  Running

# Verify cert-manager is running
kubectl get pods -n cert-manager
# Should show: cert-manager-xxx, cert-manager-webhook-xxx, cert-manager-cainjector-xxx

# Verify metrics-server is working
kubectl top nodes
# Should return CPU/Memory metrics

Important - MicroK8s Ingress Class:

MicroK8s ingress addon uses class name public (NOT nginx)
The ClusterIssuers in this repo are already configured with class: public
If you see cert-manager challenges failing, verify the ingress class matches

Optional but Recommended:

# Enable Prometheus for additional monitoring (optional)
microk8s enable prometheus

# Enable registry if you want local image storage (optional)
microk8s enable registry

Step 3: Enhanced Infrastructure Components

The platform includes additional infrastructure components that enhance security, monitoring, and operations:

# The platform includes Mailu for email services
# Deploy Mailu via Helm (optional but recommended for production):
kubectl create namespace bakery-ia --dry-run=client -o yaml | kubectl apply -f -
helm repo add mailu https://mailu.github.io/helm-charts
helm repo update
helm install mailu mailu/mailu \
  -n bakery-ia \
  -f infrastructure/platform/mail/mailu-helm/values.yaml \
  --timeout 10m \
  --wait

# Verify Mailu deployment
kubectl get pods -n bakery-ia | grep mailu

For development environments, ensure the prepull-base-images script is run:

# On your local machine, run the prepull script to cache base images
cd bakery-ia
chmod +x scripts/prepull-base-images.sh
./scripts/prepull-base-images.sh

For production environments, ensure CI/CD infrastructure is properly configured:

# Tekton Pipelines for CI/CD (optional - can be deployed separately)
kubectl create namespace tekton-pipelines
kubectl apply -f https://storage.googleapis.com/tekton-releases/pipeline/latest/release.yaml
kubectl apply -f https://storage.googleapis.com/tekton-releases/triggers/latest/release.yaml

# Flux CD for GitOps (already enabled in MicroK8s if needed)
# flux install --namespace=flux-system --network-policy=false

Step 4: Configure Firewall

CRITICAL: Ports 80 and 443 must be open for Let's Encrypt HTTP-01 challenges to work.

# Allow necessary ports
ufw allow 22/tcp      # SSH
ufw allow 80/tcp      # HTTP - REQUIRED for Let's Encrypt HTTP-01 challenge
ufw allow 443/tcp     # HTTPS - For your application traffic
ufw allow 16443/tcp   # Kubernetes API (optional, for remote kubectl access)

# Enable firewall
ufw enable

# Check status
ufw status verbose

# Expected output should include:
# 80/tcp    ALLOW   Anywhere
# 443/tcp   ALLOW   Anywhere

Also check clouding.io firewall:

Log in to clouding.io dashboard
Go to your VPS → Firewall settings
Ensure ports 80 and 443 are allowed from anywhere (0.0.0.0/0)

Step 5: Create Namespace

# Create bakery-ia namespace
kubectl create namespace bakery-ia

# Verify
kubectl get namespaces

Domain & DNS Configuration

Step 1: Register Domain at Namecheap

Go to Namecheap
Search for your desired domain (e.g., bakewise.ia)
Complete purchase (~€10-15/year)
Save domain credentials

Step 2: Configure DNS at Namecheap

Access DNS settings:

1. Log in to Namecheap
2. Go to Domain List → Manage → Advanced DNS

Add DNS records pointing to your VPS:

Type    Host    Value               TTL
A       @       YOUR_VPS_IP         Automatic
A       *       YOUR_VPS_IP         Automatic

This points both bakewise.ia and all subdomains (*.bakewise.ia) to your VPS.

Test DNS propagation:

# Wait 5-10 minutes, then test
nslookup bakewise.ia
nslookup api.bakewise.ia
nslookup mail.bakewise.ia

Step 3 (Optional): Configure Cloudflare DNS

Add site to Cloudflare:

1. Log in to Cloudflare
2. Click "Add a Site"
3. Enter your domain name
4. Choose Free plan
5. Cloudflare will scan existing DNS records

Update nameservers at registrar:

Point your domain's nameservers to Cloudflare:
- NS1: assigned.cloudflare.com
- NS2: assigned.cloudflare.com
(Cloudflare will provide the exact values)

Add DNS records:

Type    Name        Content         TTL     Proxy
A       @           YOUR_VPS_IP     Auto    Yes
A       www         YOUR_VPS_IP     Auto    Yes
A       api         YOUR_VPS_IP     Auto    Yes
A       monitoring  YOUR_VPS_IP     Auto    Yes
CNAME   *           yourdomain.com  Auto    No

Configure SSL/TLS mode:

SSL/TLS tab → Overview → Set to "Full (strict)"

Test DNS propagation:

# Wait 5-10 minutes, then test
nslookup yourdomain.com
nslookup api.yourdomain.com

TLS/SSL Certificates

Understanding Certificate Setup

The platform uses two layers of SSL/TLS:

External (Ingress) SSL: Let's Encrypt for public HTTPS
Internal (Database) SSL: Self-signed certificates for database connections

Step 1: Generate Internal Certificates

# On your local machine
cd infrastructure/tls

# Generate certificates
./generate-certificates.sh

# This creates:
# - ca/ (Certificate Authority)
# - postgres/ (PostgreSQL server certs)
# - redis/ (Redis server certs)

Certificate Details:

Root CA: 10-year validity (expires 2035)
Server certs: 3-year validity (expires October 2028)
Algorithm: RSA 4096-bit
Signature: SHA-256

Step 2: Create Kubernetes Secrets

# Create PostgreSQL TLS secret
kubectl create secret generic postgres-tls \
  --from-file=server-cert.pem=infrastructure/tls/postgres/server-cert.pem \
  --from-file=server-key.pem=infrastructure/tls/postgres/server-key.pem \
  --from-file=ca-cert.pem=infrastructure/tls/postgres/ca-cert.pem \
  -n bakery-ia

# Create Redis TLS secret
kubectl create secret generic redis-tls \
  --from-file=redis-cert.pem=infrastructure/tls/redis/redis-cert.pem \
  --from-file=redis-key.pem=infrastructure/tls/redis/redis-key.pem \
  --from-file=ca-cert.pem=infrastructure/tls/redis/ca-cert.pem \
  -n bakery-ia

# Verify secrets created
kubectl get secrets -n bakery-ia | grep tls

Step 3: Configure Let's Encrypt (External SSL)

cert-manager is already enabled via microk8s enable cert-manager. The ClusterIssuer is pre-configured in the repository.

Important: MicroK8s ingress addon uses ingress class public (not nginx). This is already configured in:

infrastructure/platform/cert-manager/cluster-issuer-production.yaml
infrastructure/platform/cert-manager/cluster-issuer-staging.yaml

# On VPS, apply the pre-configured ClusterIssuers
kubectl apply -k infrastructure/platform/cert-manager/

# Verify ClusterIssuers are ready
kubectl get clusterissuer
kubectl describe clusterissuer letsencrypt-production

# Expected output:
# NAME                     READY   AGE
# letsencrypt-production   True    1m
# letsencrypt-staging      True    1m

Configuration details (already set):

Email: admin@bakewise.ai (receives Let's Encrypt expiry notifications)
Ingress class: public (MicroK8s default)
Challenge type: HTTP-01 (requires port 80 open)

If you need to customize the email, edit before applying:

# Edit the production issuer
nano infrastructure/platform/cert-manager/cluster-issuer-production.yaml
# Change: email: admin@bakewise.ai → email: your-email@yourdomain.com

Email & Communication Setup

Self-Hosted Mailu with Mailgun Relay

Architecture:

Mailu - Self-hosted email server (Postfix, Dovecot, Rspamd, Roundcube webmail)
Mailgun - External SMTP relay for improved outbound deliverability
Helm deployment - infrastructure/platform/mail/mailu-helm/

Features:

✅ Full control over email infrastructure
✅ Mailgun relay improves deliverability (avoids VPS IP reputation issues)
✅ Built-in antispam (rspamd) with DNSSEC validation
✅ Webmail interface (Roundcube) at /webmail
✅ Admin panel at /admin
✅ IMAP/SMTP with TLS
✅ Professional addresses: admin@bakewise.ai, noreply@bakewise.ai

Configuration Files:

File	Purpose
`infrastructure/platform/mail/mailu-helm/values.yaml`	Base Mailu configuration
`infrastructure/platform/mail/mailu-helm/prod/values.yaml`	Production overrides
`infrastructure/platform/mail/mailu-helm/configs/mailgun-credentials-secret.yaml`	Mailgun SMTP credentials

Internal SMTP for Application Services:

# Services use Mailu's internal postfix for sending
SMTP_HOST: mailu-postfix.bakery-ia.svc.cluster.local
SMTP_PORT: 587

Prerequisites

Before deploying Mailu, ensure:

Unbound DNS is deployed (for DNSSEC validation)
CoreDNS is configured to forward to Unbound
DNS records are configured for your domain

Step 1: Configure DNS Records

Add these DNS records for your domain (e.g., bakewise.ai):

Type    Name            Value                           TTL
A       mail            YOUR_VPS_IP                     Auto
MX      @               mail.bakewise.ai (priority 10)  Auto
TXT     @               v=spf1 mx a -all                Auto
TXT     _dmarc          v=DMARC1; p=reject; rua=...     Auto

DKIM record will be generated after Mailu is running - you'll add it later.

Step 2: Deploy Unbound DNS Resolver

Unbound provides DNSSEC validation required by Mailu for email authentication.

# On VPS - Deploy Unbound via Helm
helm upgrade --install unbound infrastructure/platform/networking/dns/unbound-helm \
    -n bakery-ia \
    --create-namespace \
    -f infrastructure/platform/networking/dns/unbound-helm/values.yaml \
    -f infrastructure/platform/networking/dns/unbound-helm/prod/values.yaml \
    --timeout 5m \
    --wait

# Verify Unbound is running
kubectl get pods -n bakery-ia | grep unbound
# Should show: unbound-xxx  1/1  Running

# Get Unbound service IP (needed for CoreDNS configuration)
UNBOUND_IP=$(kubectl get svc unbound-dns -n bakery-ia -o jsonpath='{.spec.clusterIP}')
echo "Unbound DNS IP: $UNBOUND_IP"

Step 3: Configure CoreDNS for DNSSEC

Mailu requires DNSSEC validation. Configure CoreDNS to forward external queries to Unbound:

# Get the Unbound service IP
UNBOUND_IP=$(kubectl get svc unbound-dns -n bakery-ia -o jsonpath='{.spec.clusterIP}')

# Patch CoreDNS to forward to Unbound
kubectl patch configmap coredns -n kube-system --type merge -p "{
  \"data\": {
    \"Corefile\": \".:53 {\\n    errors\\n    health {\\n       lameduck 5s\\n    }\\n    ready\\n    kubernetes cluster.local in-addr.arpa ip6.arpa {\\n       pods insecure\\n       fallthrough in-addr.arpa ip6.arpa\\n       ttl 30\\n    }\\n    prometheus :9153\\n    forward . $UNBOUND_IP {\\n       max_concurrent 1000\\n    }\\n    cache 30 {\\n       disable success cluster.local\\n       disable denial cluster.local\\n    }\\n    loop\\n    reload\\n    loadbalance\\n}\\n\"
  }
}"

# Restart CoreDNS to apply changes
kubectl rollout restart deployment coredns -n kube-system
kubectl rollout status deployment coredns -n kube-system --timeout=60s

# Verify DNSSEC is working
kubectl run -it --rm debug --image=alpine --restart=Never -- \
    sh -c "apk add drill && drill -D google.com"
# Should show: ;; flags: ... ad ...  (ad = authenticated data = DNSSEC valid)

Step 4: Create TLS Certificate Secret

Mailu Front pod requires a TLS certificate:

# Generate self-signed certificate for internal use
# (Let's Encrypt handles external TLS via Ingress)
TEMP_DIR=$(mktemp -d)
cd "$TEMP_DIR"

openssl req -x509 -nodes -days 365 -newkey rsa:2048 \
    -keyout tls.key -out tls.crt \
    -subj "/CN=mail.bakewise.ai/O=bakewise"

kubectl create secret tls mailu-certificates \
    --cert=tls.crt \
    --key=tls.key \
    -n bakery-ia

rm -rf "$TEMP_DIR"

# Verify secret created
kubectl get secret mailu-certificates -n bakery-ia

Step 5: Create Admin Credentials Secret

# Generate a secure password (or use your own)
ADMIN_PASSWORD=$(openssl rand -base64 16 | tr -d '/+=' | head -c 16)
echo "Admin password: $ADMIN_PASSWORD"
echo "SAVE THIS PASSWORD SECURELY!"

# Create the admin credentials secret
kubectl create secret generic mailu-admin-credentials \
    --from-literal=password="$ADMIN_PASSWORD" \
    -n bakery-ia

Step 6: Deploy Mailu via Helm

# Add Mailu Helm repository
helm repo add mailu https://mailu.github.io/helm-charts
helm repo update mailu

# Deploy Mailu with production values
# Admin user is created automatically via initialAccount feature
helm upgrade --install mailu mailu/mailu \
    -n bakery-ia \
    --create-namespace \
    -f infrastructure/platform/mail/mailu-helm/values.yaml \
    -f infrastructure/platform/mail/mailu-helm/prod/values.yaml \
    --timeout 10m

# Wait for pods to be ready (may take 5-10 minutes for ClamAV)
kubectl get pods -n bakery-ia -l app.kubernetes.io/instance=mailu -w

# Admin user (admin@bakewise.ai) is created automatically!
# Password is the one you set in Step 5

Step 7: Configure DKIM

After Mailu is running, get the DKIM key and add it to DNS:

# Get DKIM public key
kubectl exec -n bakery-ia deployment/mailu-admin -- \
    cat /dkim/bakewise.ai.dkim.pub

# Add this as a TXT record in your DNS:
# Name: dkim._domainkey
# Value: (the key from above)

Step 8: Verify Email Setup

# Check all Mailu pods are running
kubectl get pods -n bakery-ia | grep mailu
# Expected: All 10 pods in Running state

# Test SMTP connectivity
kubectl run -it --rm smtp-test --image=alpine --restart=Never -- \
    sh -c "apk add swaks && swaks --to test@example.com --from admin@bakewise.ai --server mailu-front.bakery-ia.svc.cluster.local:25"

# Access webmail (via port-forward for testing)
kubectl port-forward -n bakery-ia svc/mailu-front 8080:80
# Open: http://localhost:8080/webmail

Production Email Endpoints

Service	URL/Address
Admin Panel	https://mail.bakewise.ai/admin
Webmail	https://mail.bakewise.ai/webmail
SMTP (STARTTLS)	mail.bakewise.ai:587
SMTP (SSL)	mail.bakewise.ai:465
IMAP (SSL)	mail.bakewise.ai:993

Troubleshooting Mailu

Issue: Admin pod CrashLoopBackOff with "DNSSEC validation" error

# Verify CoreDNS is forwarding to Unbound
kubectl get configmap coredns -n kube-system -o yaml | grep forward
# Should show: forward . <unbound-ip>

# If not, re-run Step 3 above

Issue: Front pod stuck in ContainerCreating

# Check for missing certificate secret
kubectl describe pod -n bakery-ia -l app.kubernetes.io/component=front | grep -A5 Events

# If missing mailu-certificates, re-run Step 4 above

Issue: Admin pod can't connect to Redis

# Verify externalRedis is disabled in values
helm get values mailu -n bakery-ia | grep -A5 externalRedis
# Should show: enabled: false

# If enabled: true, upgrade with correct values
helm upgrade mailu mailu/mailu -n bakery-ia \
    -f infrastructure/platform/mail/mailu-helm/values.yaml \
    -f infrastructure/platform/mail/mailu-helm/prod/values.yaml

WhatsApp Business API Setup

Features:

✅ First 1,000 conversations/month FREE
✅ Perfect for 10 tenants (~500 messages/month)

Setup Steps:

Create Meta Business Account:

1. Go to business.facebook.com
2. Create Business Account
3. Complete business verification

Add WhatsApp Product:

1. Go to developers.facebook.com
2. Create New App → Business
3. Add WhatsApp product
4. Complete setup wizard

Configure Phone Number:

1. Test with your personal number initially
2. Later: Get dedicated business number
3. Verify phone number with SMS code

Create Message Templates:

1. Go to WhatsApp Manager
2. Create templates for:
   - Low inventory alert
   - Expired product alert
   - Forecast summary
   - Order notification
3. Submit for approval (15 min - 24 hours)

Get API Credentials:

Save these values:
- Phone Number ID: (from WhatsApp Manager)
- Access Token: (from App Dashboard)
- Business Account ID: (from WhatsApp Manager)
- Webhook Verify Token: (create your own secure string)

Kubernetes Deployment

Step 1: Prepare Container Images

Option A: Using Docker Hub (Recommended)

# On your local machine
docker login

# Build all images
docker-compose build

# Tag images for Docker Hub
# Replace YOUR_USERNAME with your Docker Hub username
export DOCKER_USERNAME="YOUR_USERNAME"

./scripts/tag-images.sh $DOCKER_USERNAME

# Push to Docker Hub
./scripts/push-images.sh $DOCKER_USERNAME

# Update prod kustomization with your username
# Edit: infrastructure/kubernetes/overlays/prod/kustomization.yaml
# Replace all "bakery/" with "$DOCKER_USERNAME/"

Option B: Using MicroK8s Registry

# On VPS
microk8s enable registry

# Get registry address (usually localhost:32000)
kubectl get service -n container-registry

# On local machine, configure insecure registry
# Edit /etc/docker/daemon.json:
{
  "insecure-registries": ["YOUR_VPS_IP:32000"]
}

# Restart Docker
sudo systemctl restart docker

# Tag and push images
docker tag bakery/auth-service YOUR_VPS_IP:32000/bakery/auth-service
docker push YOUR_VPS_IP:32000/bakery/auth-service
# Repeat for all services...

Step 2: Update Production Configuration

⚠️ CRITICAL: The default configuration uses bakewise.ai domain. You MUST update this before deployment if using a different domain.

Required Configuration Updates

Step 2.1: Remove imagePullSecrets

# On your local machine
cd bakery-ia

# Remove imagePullSecrets from all deployment files
find infrastructure/kubernetes/base -name "*.yaml" -type f -exec sed -i.bak '/imagePullSecrets:/,+1d' {} \;

# Verify removal
grep -r "imagePullSecrets" infrastructure/kubernetes/base/
# Should return NO results

Step 2.2: Update Image Tags (Use Semantic Versions)

# Edit kustomization.yaml to replace 'latest' with actual version
nano infrastructure/kubernetes/overlays/prod/kustomization.yaml

# Find the images section (lines 163-196) and update:
# BEFORE:
#   - name: bakery/auth-service
#     newTag: latest
# AFTER:
#   - name: bakery/auth-service
#     newTag: v1.0.0

# Do this for ALL 22 services, or use this helper:
export VERSION="1.0.0"  # Your version

# Create a script to update all image tags
cat > /tmp/update-tags.sh <<'EOF'
#!/bin/bash
VERSION="${1:-1.0.0}"
sed -i "s/newTag: latest/newTag: v${VERSION}/g" infrastructure/kubernetes/overlays/prod/kustomization.yaml
EOF

chmod +x /tmp/update-tags.sh
/tmp/update-tags.sh ${VERSION}

# Verify no 'latest' tags remain
grep "newTag:" infrastructure/kubernetes/overlays/prod/kustomization.yaml | grep -c "latest"
# Should return: 0

Step 2.3: Fix SigNoz Namespace References

# Update SigNoz patches to use bakery-ia namespace instead of signoz
sed -i 's/namespace: signoz/namespace: bakery-ia/g' infrastructure/kubernetes/overlays/prod/kustomization.yaml

# Verify changes (should show bakery-ia in all 3 patches)
grep -A 3 "name: signoz" infrastructure/kubernetes/overlays/prod/kustomization.yaml

Step 2.4: Update Cert-Manager Email

# Update Let's Encrypt notification email to your production email
sed -i "s/admin@bakery-ia.local/admin@bakewise.ai/g" \
  infrastructure/kubernetes/base/components/cert-manager/cluster-issuer-production.yaml

Step 2.5: Verify Production Secrets (Already Configured) ✅

# Production secrets have been pre-configured with strong cryptographic passwords
# No manual action required - secrets are already set in secrets.yaml

# Verify the secrets are configured (optional)
echo "Verifying production secrets configuration..."
grep "JWT_SECRET_KEY" infrastructure/kubernetes/base/secrets.yaml | head -1
grep "AUTH_DB_PASSWORD" infrastructure/kubernetes/base/secrets.yaml | head -1
grep "REDIS_PASSWORD" infrastructure/kubernetes/base/secrets.yaml | head -1

echo "✅ All production secrets are configured and ready for deployment"

Production URLs:

Main Application: https://bakewise.ai
API Endpoints: https://bakewise.ai/api/v1/...
SigNoz (Monitoring): https://monitoring.bakewise.ai/signoz
AlertManager: https://monitoring.bakewise.ai/alertmanager

Configuration & Secrets

Production Secrets Status ✅

All core secrets have been pre-configured with strong cryptographic passwords:

✅ Database passwords (19 databases) - 24-character random strings
✅ JWT secrets - 256-bit cryptographically secure tokens
✅ Service API key - 64-character hexadecimal string
✅ Redis password - 24-character random string
✅ RabbitMQ password - 24-character random string
✅ RabbitMQ Erlang cookie - 64-character hexadecimal string

Step 1: Configure External Service Credentials (Email & WhatsApp)

You still need to update these external service credentials:

# Edit the secrets file
nano infrastructure/kubernetes/base/secrets.yaml

# Update ONLY these external service credentials:

# SMTP settings (from email setup):
SMTP_USER: <base64-encoded-username>      # your email
SMTP_PASSWORD: <base64-encoded-password>  # app password

# WhatsApp credentials (from WhatsApp setup - optional):
WHATSAPP_API_KEY: <base64-encoded-key>

# Payment processing (from Stripe setup):
STRIPE_SECRET_KEY: <base64-encoded-key>
STRIPE_WEBHOOK_SECRET: <base64-encoded-secret>

To base64 encode:

echo -n "your-value-here" | base64

CRITICAL: Never commit real secrets to git! The secrets.yaml file should be in .gitignore.

Step 2: CI/CD Secrets Configuration

For production CI/CD setup, additional secrets are required:

# Create Docker Hub credentials secret (for image pulls)
kubectl create secret docker-registry dockerhub-creds \
  --docker-server=docker.io \
  --docker-username=YOUR_DOCKERHUB_USERNAME \
  --docker-password=YOUR_DOCKERHUB_TOKEN \
  --docker-email=your-email@example.com \
  -n bakery-ia

# Create Gitea registry credentials (if using Gitea for CI/CD)
kubectl create secret docker-registry gitea-registry-credentials \
  -n tekton-pipelines \
  --docker-server=gitea.bakery-ia.local:5000 \
  --docker-username=your-username \
  --docker-password=your-password

# Create Git credentials for Flux (if using GitOps)
kubectl create secret generic gitea-credentials \
  -n flux-system \
  --from-literal=username=your-username \
  --from-literal=password=your-password

Step 3: Apply Application Secrets

# Copy manifests to VPS (from local machine)
scp -r infrastructure/kubernetes root@YOUR_VPS_IP:~/

# SSH to VPS
ssh root@YOUR_VPS_IP

# Apply application secrets
kubectl apply -f ~/infrastructure/kubernetes/base/secrets.yaml -n bakery-ia

# Verify secrets created
kubectl get secrets -n bakery-ia
# Should show multiple secrets including postgres-tls, redis-tls, app-secrets, etc.

Database Migrations

Step 0: Deploy CI/CD Infrastructure (Optional but Recommended)

For production environments, deploy CI/CD infrastructure components:

# Deploy Tekton Pipelines for CI/CD (optional but recommended for production)
kubectl create namespace tekton-pipelines

# Install Tekton Pipelines
kubectl apply -f https://storage.googleapis.com/tekton-releases/pipeline/latest/release.yaml

# Install Tekton Triggers
kubectl apply -f https://storage.googleapis.com/tekton-releases/triggers/latest/release.yaml

# Apply Tekton configurations
kubectl apply -f ~/infrastructure/cicd/tekton/tasks/
kubectl apply -f ~/infrastructure/cicd/tekton/pipelines/
kubectl apply -f ~/infrastructure/cicd/tekton/triggers/

# Verify Tekton deployment
kubectl get pods -n tekton-pipelines

Step 1: Deploy SigNoz Monitoring (BEFORE Application)

⚠️ CRITICAL: SigNoz must be deployed BEFORE the application into the bakery-ia namespace because the production kustomization patches SigNoz resources.

# On VPS
# 1. Ensure bakery-ia namespace exists
kubectl get namespace bakery-ia || kubectl create namespace bakery-ia

# 2. Add Helm repo
helm repo add signoz https://charts.signoz.io
helm repo update

# 3. Install SigNoz into bakery-ia namespace (NOT separate signoz namespace)
helm install signoz signoz/signoz \
  -n bakery-ia \
  --set frontend.service.type=ClusterIP \
  --set clickhouse.persistence.size=20Gi \
  --set clickhouse.persistence.storageClass=microk8s-hostpath

# 4. Wait for SigNoz to be ready (this may take 10-15 minutes)
kubectl wait --for=condition=ready pod \
  -l app.kubernetes.io/instance=signoz \
  -n bakery-ia \
  --timeout=900s

# 5. Verify SigNoz components running in bakery-ia namespace
kubectl get pods -n bakery-ia -l app.kubernetes.io/instance=signoz
# Should show: signoz-0, signoz-otel-collector, signoz-clickhouse, signoz-zookeeper, signoz-alertmanager

# 6. Verify StatefulSets exist (kustomization will patch these)
kubectl get statefulset -n bakery-ia | grep signoz
# Should show: signoz, signoz-clickhouse

⚠️ Important: Do NOT create a separate signoz namespace. SigNoz must be in bakery-ia namespace for the overlays to work correctly.

Step 2: Deploy Application and Databases

# On VPS
kubectl apply -k ~/infrastructure/kubernetes/overlays/prod

# Wait for databases to be ready (5-10 minutes)
kubectl wait --for=condition=ready pod \
  -l app.kubernetes.io/component=database \
  -n bakery-ia \
  --timeout=600s

# Check status
kubectl get pods -n bakery-ia -l app.kubernetes.io/component=database

Step 2: Run Migrations

Migrations are automatically handled by init containers in each service. Verify they completed:

# Check migration job status
kubectl get jobs -n bakery-ia | grep migration

# All should show "COMPLETIONS = 1/1"

# Check logs if any failed
kubectl logs -n bakery-ia job/auth-migration

Step 3: Verify Database Schemas

# Connect to a database to verify
kubectl exec -n bakery-ia deployment/auth-db -it -- psql -U auth_user -d auth_db

# Inside psql:
\dt          # List tables
\d users     # Describe users table
\q           # Quit

CI/CD Infrastructure Deployment

This section covers deploying the complete CI/CD stack: Gitea (Git server + container registry), Tekton (CI pipelines), and Flux CD (GitOps deployments).

Overview

┌─────────────────────────────────────────────────────────────────────────────┐
│                         CI/CD ARCHITECTURE                                   │
│                                                                              │
│  Developer Push                                                              │
│       │                                                                      │
│       ▼                                                                      │
│  ┌─────────┐     Webhook      ┌─────────────┐     Build/Test    ┌─────────┐ │
│  │  Gitea  │ ───────────────► │   Tekton    │ ─────────────────►│ Images  │ │
│  │  (Git)  │                  │  (Pipelines)│                   │(Registry)│ │
│  └─────────┘                  └─────────────┘                   └─────────┘ │
│       │                              │                               │      │
│       │                              │ Update manifests              │      │
│       │                              ▼                               │      │
│       │                       ┌─────────────┐                        │      │
│       └──────────────────────►│   Flux CD   │◄───────────────────────┘      │
│         Monitor changes       │  (GitOps)   │     Pull images               │
│                               └─────────────┘                                │
│                                      │                                       │
│                                      ▼                                       │
│                               ┌─────────────┐                                │
│                               │ Kubernetes  │                                │
│                               │   Cluster   │                                │
│                               └─────────────┘                                │
└─────────────────────────────────────────────────────────────────────────────┘

Prerequisites

Before deploying CI/CD infrastructure:

Kubernetes cluster is running
Ingress controller is configured
TLS certificates are available
DNS records configured for gitea.bakewise.ai

Step 1: Deploy Gitea (Git Server + Container Registry)

Gitea provides a self-hosted Git server with built-in container registry support. The setup is fully automated - admin user and initial repository are created automatically.

1.1 Create Secrets and Init Job (One Command)

The setup script creates all necessary secrets and applies the initialization job.

For Production Deployment:

# Generate a secure password (minimum 16 characters required for production)
export GITEA_ADMIN_PASSWORD=$(openssl rand -base64 32)
echo "Gitea Admin Password: $GITEA_ADMIN_PASSWORD"
echo "⚠️  Save this password securely - you'll need it for Tekton setup!"

# Run the setup script with --production flag
# This enforces password requirements and uses production registry URL
./infrastructure/cicd/gitea/setup-admin-secret.sh --production

What the --production flag does:

Requires GITEA_ADMIN_PASSWORD environment variable (won't use defaults)
Validates password is at least 16 characters
Uses production registry URL (registry.bakewise.ai)
Hides password in output for security
Shows production-specific next steps

This creates:

gitea-admin-secret in gitea namespace - admin credentials for Gitea
gitea-registry-secret in bakery-ia namespace - for imagePullSecrets
gitea-init-job - Kubernetes Job that creates the bakery-ia repository automatically

For dev environments only: Run without flags to use the default static password:
./infrastructure/cicd/gitea/setup-admin-secret.sh

1.2 Install Gitea via Helm

# Add Gitea Helm repository
helm repo add gitea https://dl.gitea.io/charts
helm repo update gitea

# Install Gitea with PRODUCTION values (includes TLS, proper domains, resources)
helm upgrade --install gitea gitea/gitea \
  -n gitea \
  -f infrastructure/cicd/gitea/values.yaml \
  -f infrastructure/cicd/gitea/values-prod.yaml \
  --timeout 10m \
  --wait

# Wait for Gitea to be ready
kubectl wait --for=condition=ready pod -n gitea -l app.kubernetes.io/name=gitea --timeout=300s

# Verify Gitea is running
kubectl get pods -n gitea
kubectl get svc -n gitea

Production values (values-prod.yaml) include:

Domain: gitea.bakewise.ai and registry.bakewise.ai
TLS via cert-manager with Let's Encrypt production issuer
50Gi storage (vs 10Gi in dev)
Increased resource limits

1.3 Verify Repository Initialization

The init job automatically creates the bakery-ia repository once Gitea is ready:

# Check init job completed successfully
kubectl logs -n gitea job/gitea-init-repo

# Expected output:
# === Gitea Repository Initialization ===
# Gitea is ready!
# Repository 'bakery-ia' created successfully!

If the job needs to be re-run:

kubectl delete job gitea-init-repo -n gitea
kubectl apply -f infrastructure/cicd/gitea/gitea-init-job.yaml

1.4 Configure DNS for Gitea

Add DNS record pointing to your VPS:

Type    Name      Value           TTL
A       gitea     YOUR_VPS_IP     Auto

1.5 Verify Gitea Access

# Check ingress is configured
kubectl get ingress -n gitea

# Test access (after DNS propagation)
curl -I https://gitea.bakewise.ai

# Access web interface
# URL: https://gitea.bakewise.ai
# Username: bakery-admin
# Password: (from step 1.1)

# Verify repository was created via API
curl -u bakery-admin:$GITEA_ADMIN_PASSWORD \
  https://gitea.bakewise.ai/api/v1/repos/bakery-admin/bakery-ia

1.6 Push Code to Repository

The bakery-ia repository is already created with a README. Push your code:

# Add Gitea as remote and push code
cd /path/to/bakery-ia
git remote add gitea https://gitea.bakewise.ai/bakery-admin/bakery-ia.git
git push gitea main

Step 2: Deploy Tekton Pipelines

Tekton provides cloud-native CI/CD pipelines.

2.1 Install Tekton Core Components

# Create Tekton namespace
kubectl create namespace tekton-pipelines

# Install Tekton Pipelines
kubectl apply -f https://storage.googleapis.com/tekton-releases/pipeline/latest/release.yaml

# Wait for Tekton Pipelines to be ready
kubectl wait --for=condition=available --timeout=300s \
  deployment/tekton-pipelines-controller -n tekton-pipelines
kubectl wait --for=condition=available --timeout=300s \
  deployment/tekton-pipelines-webhook -n tekton-pipelines

# Install Tekton Triggers (for webhook-based automation)
kubectl apply -f https://storage.googleapis.com/tekton-releases/triggers/latest/release.yaml
kubectl apply -f https://storage.googleapis.com/tekton-releases/triggers/latest/interceptors.yaml

# Wait for Tekton Triggers to be ready
kubectl wait --for=condition=available --timeout=300s \
  deployment/tekton-triggers-controller -n tekton-pipelines
kubectl wait --for=condition=available --timeout=300s \
  deployment/tekton-triggers-webhook -n tekton-pipelines

# Verify installation
kubectl get pods -n tekton-pipelines

2.2 Deploy Tekton CI/CD Configuration via Helm

For Production Deployment:

# Generate secure webhook token (save this for Gitea webhook configuration)
export TEKTON_WEBHOOK_TOKEN=$(openssl rand -hex 32)
echo "Webhook Token: $TEKTON_WEBHOOK_TOKEN"
echo "⚠️  Save this token - you'll need it for Gitea webhook setup!"

# Ensure GITEA_ADMIN_PASSWORD is still set from Step 1
echo "Using Gitea password from: GITEA_ADMIN_PASSWORD"

# Install Tekton CI/CD with PRODUCTION values
helm upgrade --install tekton-cicd infrastructure/cicd/tekton-helm \
  -n tekton-pipelines \
  -f infrastructure/cicd/tekton-helm/values.yaml \
  -f infrastructure/cicd/tekton-helm/values-prod.yaml \
  --set secrets.webhook.token=$TEKTON_WEBHOOK_TOKEN \
  --set secrets.registry.password=$GITEA_ADMIN_PASSWORD \
  --set secrets.git.password=$GITEA_ADMIN_PASSWORD \
  --timeout 10m \
  --wait

# Verify resources created
kubectl get pipelines -n tekton-pipelines
kubectl get tasks -n tekton-pipelines
kubectl get eventlisteners -n tekton-pipelines
kubectl get triggerbindings -n tekton-pipelines
kubectl get triggertemplates -n tekton-pipelines

What the production values (values-prod.yaml) provide:

Empty default secrets (must be provided via --set flags)
Increased controller/webhook replicas (2 each)
Higher resource limits for production workloads
10Gi workspace storage (vs 5Gi in dev)

⚠️ Security Note: Never commit actual secrets to values files. Always pass them via --set flags or use external secret management.

2.3 Configure Gitea Webhook

Go to Gitea repository settings → Webhooks
Add webhook:
- Target URL: http://el-bakery-ia-listener.tekton-pipelines.svc.cluster.local:8080
- HTTP Method: POST
- Content Type: application/json
- Secret: (same as secrets.webhook.token from Helm)
- Trigger on: Push events
Save webhook

2.4 Test Pipeline Manually

# Create a manual PipelineRun to test the CI pipeline
cat <<EOF | kubectl create -f -
apiVersion: tekton.dev/v1beta1
kind: PipelineRun
metadata:
  generateName: manual-ci-run-
  namespace: tekton-pipelines
spec:
  pipelineRef:
    name: bakery-ia-ci
  workspaces:
    - name: shared-workspace
      volumeClaimTemplate:
        spec:
          accessModes: ["ReadWriteOnce"]
          resources:
            requests:
              storage: 5Gi
    - name: docker-credentials
      secret:
        secretName: gitea-registry-credentials
  params:
    - name: git-url
      value: "http://gitea-http.gitea.svc.cluster.local:3000/bakery-admin/bakery-ia.git"
    - name: git-revision
      value: "main"
EOF

# Watch pipeline progress
kubectl get pipelineruns -n tekton-pipelines -w

# View logs
tkn pipelinerun logs -n tekton-pipelines -f

Step 3: Deploy Flux CD (GitOps)

Flux CD provides GitOps-based continuous deployment.

3.1 Install Flux CLI

# Install Flux CLI (if not already installed)
curl -s https://fluxcd.io/install.sh | sudo bash

# Verify installation
flux --version

3.2 Install Flux Components

# Install Flux CRDs and controllers
flux install --namespace=flux-system --network-policy=false

# Verify Flux is running
kubectl get pods -n flux-system
flux check

3.3 Deploy Flux Configuration via Helm

# Create Flux namespace if not exists
kubectl create namespace flux-system --dry-run=client -o yaml | kubectl apply -f -

# Create Git credentials secret for Flux
kubectl create secret generic gitea-credentials \
  -n flux-system \
  --from-literal=username=bakery-admin \
  --from-literal=password="$GITEA_ADMIN_PASSWORD"

# Install Flux configuration
helm upgrade --install flux-cd infrastructure/cicd/flux \
  -n flux-system \
  --timeout 10m \
  --wait

# Verify Flux resources
kubectl get gitrepositories -n flux-system
kubectl get kustomizations -n flux-system

3.4 Verify GitOps Sync

# Check GitRepository status
flux get sources git -n flux-system

# Check Kustomization status
flux get kustomizations -n flux-system

# Force reconciliation
flux reconcile source git bakery-ia -n flux-system
flux reconcile kustomization bakery-ia-prod -n flux-system

Step 4: Complete CI/CD Workflow Test

Test the entire CI/CD pipeline end-to-end:

# 1. Make a small change in your local repo
echo "# CI/CD Test $(date)" >> README.md
git add README.md
git commit -m "Test CI/CD pipeline"

# 2. Push to Gitea
git push gitea main

# 3. Watch Tekton pipeline triggered by webhook
kubectl get pipelineruns -n tekton-pipelines -w

# 4. After pipeline completes, watch Flux sync
flux get kustomizations -n flux-system -w

# 5. Verify deployment updated
kubectl get deployments -n bakery-ia -o wide

CI/CD Troubleshooting

Tekton Pipeline Fails

# View pipeline run status
kubectl get pipelineruns -n tekton-pipelines

# Get detailed logs
tkn pipelinerun describe <pipelinerun-name> -n tekton-pipelines
tkn pipelinerun logs <pipelinerun-name> -n tekton-pipelines

# Check EventListener logs (for webhook issues)
kubectl logs -n tekton-pipelines -l app.kubernetes.io/component=eventlistener

Flux Not Syncing

# Check GitRepository status
kubectl describe gitrepository bakery-ia -n flux-system

# Check Kustomization status
kubectl describe kustomization bakery-ia-prod -n flux-system

# View Flux controller logs
kubectl logs -n flux-system deployment/source-controller
kubectl logs -n flux-system deployment/kustomize-controller

# Force reconciliation
flux reconcile source git bakery-ia -n flux-system --with-source

Gitea Webhook Not Triggering

# Check webhook delivery in Gitea UI
# Settings → Webhooks → Recent Deliveries

# Verify EventListener is running
kubectl get eventlisteners -n tekton-pipelines
kubectl get svc -n tekton-pipelines | grep listener

# Check EventListener logs
kubectl logs -n tekton-pipelines -l eventlistener=bakery-ia-listener

CI/CD URLs Summary

Service	URL	Purpose
Gitea	https://gitea.bakewise.ai	Git repository & container registry
Gitea Registry	https://gitea.bakewise.ai/v2/	Docker registry API
Tekton Dashboard	(install separately if needed)	Pipeline visualization
Flux	CLI only	GitOps status via `flux` commands

CI/CD Security Considerations

The CI/CD infrastructure has been configured with production security in mind:

Secrets Management

Secret	Purpose	How to Generate
`GITEA_ADMIN_PASSWORD`	Gitea admin & registry auth	`openssl rand -base64 32`
`TEKTON_WEBHOOK_TOKEN`	Webhook signature validation	`openssl rand -hex 32`

Security Features

Production Mode Enforcement
- The --production flag on setup-admin-secret.sh enforces:
  - Mandatory GITEA_ADMIN_PASSWORD environment variable
  - Minimum 16-character password requirement
  - Password hidden from terminal output
Registry Communication
- Git operations (clone, push) use internal cluster DNS: gitea-http.gitea.svc.cluster.local:3000
- Image references use external HTTPS URL: registry.bakewise.ai (containerd requires HTTPS for auth)
- This ensures image pulls work correctly while git operations stay internal
Credential Isolation
- Secrets are passed via --set flags, never committed to git
- Registry credentials are scoped per-namespace
- Webhook tokens are unique per installation

Post-Deployment Security Checklist

# Verify no default passwords in use
kubectl get secret gitea-admin-secret -n gitea -o jsonpath='{.data.password}' | base64 -d | wc -c
# Should be 32+ characters for production

# Verify webhook secret is set
kubectl get secret gitea-webhook-secret -n tekton-pipelines -o jsonpath='{.data.secretToken}' | base64 -d | wc -c
# Should be 64 characters (hex-encoded 32 bytes)

# Verify no hardcoded URLs in tasks
kubectl get task update-gitops -n tekton-pipelines -o yaml | grep -c "bakery-ia.local"
# Should be 0

Mailu Email Server Deployment

Mailu is a full-featured, self-hosted email server with built-in antispam, webmail, and admin panel. Outbound emails are relayed through Mailgun for improved deliverability and to avoid IP reputation issues.

Prerequisites

Before deploying Mailu:

Unbound DNS resolver deployed (for DNSSEC validation)
DNS records configured for mail domain
TLS certificates available
Mailgun account created and domain verified (for outbound email relay)

Step 1: Deploy Unbound DNS Resolver

Mailu requires DNSSEC validation for email authentication (DKIM/SPF/DMARC).

# Deploy Unbound via Helm
helm upgrade --install unbound infrastructure/platform/networking/dns/unbound-helm \
  -n bakery-ia \
  --create-namespace \
  -f infrastructure/platform/networking/dns/unbound-helm/values.yaml \
  -f infrastructure/platform/networking/dns/unbound-helm/prod/values.yaml \
  --timeout 5m \
  --wait

# Verify Unbound is running
kubectl get pods -n bakery-ia | grep unbound

# Get Unbound service IP
UNBOUND_IP=$(kubectl get svc unbound-dns -n bakery-ia -o jsonpath='{.spec.clusterIP}')
echo "Unbound DNS IP: $UNBOUND_IP"

Step 2: Configure CoreDNS for DNSSEC

# Get Unbound IP
UNBOUND_IP=$(kubectl get svc unbound-dns -n bakery-ia -o jsonpath='{.spec.clusterIP}')

# Create updated CoreDNS ConfigMap
cat > /tmp/coredns-config.yaml <<EOF
apiVersion: v1
kind: ConfigMap
metadata:
  name: coredns
  namespace: kube-system
data:
  Corefile: |
    .:53 {
        errors
        health {
           lameduck 5s
        }
        ready
        kubernetes cluster.local in-addr.arpa ip6.arpa {
           pods insecure
           fallthrough in-addr.arpa ip6.arpa
           ttl 30
        }
        prometheus :9153
        forward . $UNBOUND_IP {
           max_concurrent 1000
        }
        cache 30 {
           disable success cluster.local
           disable denial cluster.local
        }
        loop
        reload
        loadbalance
    }
EOF

# Apply configuration
kubectl apply -f /tmp/coredns-config.yaml

# Restart CoreDNS
kubectl rollout restart deployment coredns -n kube-system
kubectl rollout status deployment coredns -n kube-system --timeout=60s

# Verify DNSSEC is working
kubectl run -it --rm dns-test --image=alpine --restart=Never -- \
  sh -c "apk add drill && drill -D google.com"
# Look for "ad" flag (authenticated data) in output

Step 3: Configure Mailgun (External SMTP Relay)

Mailu uses Mailgun as an external SMTP relay for all outbound emails. This improves deliverability and avoids IP reputation issues common with self-hosted mail servers.

3.1: Create Mailgun Account

Go to https://www.mailgun.com and create an account
Add your domain (bakewise.ai) in the Mailgun dashboard
Verify domain ownership by adding the DNS records Mailgun provides

3.2: Get SMTP Credentials

In Mailgun dashboard, go to Domain Settings > SMTP credentials
Note your credentials:
- SMTP hostname: smtp.mailgun.org
- Port: 587 (TLS/STARTTLS)
- Username: typically postmaster@bakewise.ai
- Password: your Mailgun SMTP password (NOT the API key)

3.3: Create Kubernetes Secret for Mailgun

# Edit the secret template with your Mailgun credentials
nano infrastructure/platform/mail/mailu-helm/configs/mailgun-credentials-secret.yaml

# Replace the placeholder values:
#   RELAY_USERNAME: "postmaster@bakewise.ai"
#   RELAY_PASSWORD: "your-mailgun-smtp-password"

# Apply the secret
kubectl apply -f infrastructure/platform/mail/mailu-helm/configs/mailgun-credentials-secret.yaml -n bakery-ia

# Verify secret created
kubectl get secret mailu-mailgun-credentials -n bakery-ia

Step 4: Configure DNS Records for Mail

Add these DNS records for your domain (e.g., bakewise.ai):

Type    Name            Value                                   TTL     Priority
A       mail            YOUR_VPS_IP                             Auto    -
MX      @               mail.bakewise.ai                        Auto    10
TXT     @               v=spf1 include:mailgun.org mx a ~all    Auto    -
TXT     _dmarc          v=DMARC1; p=quarantine; rua=...         Auto    -

Mailgun-specific DNS records (Mailgun will provide exact values):

Type    Name                        Value                       TTL
TXT     (provided by Mailgun)       (DKIM key from Mailgun)     Auto
TXT     (provided by Mailgun)       (DKIM key from Mailgun)     Auto

Note:

The SPF record includes mailgun.org to authorize Mailgun to send on your behalf
Add the DKIM records exactly as Mailgun provides them
Mailu's own DKIM record will be added after deployment (Step 9)

Step 5: Create TLS Certificate Secret

# Generate self-signed certificate for internal Mailu use
# (Ingress handles external TLS termination)
TEMP_DIR=$(mktemp -d)
cd "$TEMP_DIR"

openssl req -x509 -nodes -days 365 -newkey rsa:2048 \
  -keyout tls.key -out tls.crt \
  -subj "/CN=mail.bakewise.ai/O=bakewise"

kubectl create secret tls mailu-certificates \
  --cert=tls.crt \
  --key=tls.key \
  -n bakery-ia

rm -rf "$TEMP_DIR"

# Verify secret created
kubectl get secret mailu-certificates -n bakery-ia

Step 6: Create Admin Credentials Secret

The admin account is created automatically during Helm deployment using the initialAccount feature. Create a secret with the admin password before deploying.

# Generate a secure password (or use your own)
ADMIN_PASSWORD=$(openssl rand -base64 16 | tr -d '/+=' | head -c 16)
echo "Admin password: $ADMIN_PASSWORD"
echo "SAVE THIS PASSWORD SECURELY!"

# Create the admin credentials secret
kubectl create secret generic mailu-admin-credentials \
  --from-literal=password="$ADMIN_PASSWORD" \
  -n bakery-ia

# Verify secret created
kubectl get secret mailu-admin-credentials -n bakery-ia

Alternative: Use the provided template file:

# Edit the secret template with your password (base64 encoded)
nano infrastructure/platform/mail/mailu-helm/configs/mailu-admin-credentials-secret.yaml

# Apply the secret
kubectl apply -f infrastructure/platform/mail/mailu-helm/configs/mailu-admin-credentials-secret.yaml

Step 7: Deploy Mailu via Helm

# Add Mailu Helm repository
helm repo add mailu https://mailu.github.io/helm-charts
helm repo update mailu

# Deploy Mailu with production values
# Note:
#   - externalRelay uses Mailgun via the secret created in Step 3
#   - initialAccount creates admin user automatically using the secret from Step 6
helm upgrade --install mailu mailu/mailu \
  -n bakery-ia \
  --create-namespace \
  -f infrastructure/platform/mail/mailu-helm/values.yaml \
  -f infrastructure/platform/mail/mailu-helm/prod/values.yaml \
  --timeout 10m

# Wait for pods to be ready (ClamAV may take 5-10 minutes)
kubectl get pods -n bakery-ia -l app.kubernetes.io/instance=mailu -w

# The admin user (admin@bakewise.ai) is created automatically!

Step 8: Apply Mailu Ingress

# Apply Mailu-specific ingress configuration
kubectl apply -f infrastructure/platform/mail/mailu-helm/mailu-ingress.yaml

# Verify ingress
kubectl get ingress -n bakery-ia | grep mailu

Admin Credentials (created automatically in Step 7):

Email: admin@bakewise.ai
Password: The password you set in Step 6 (stored in mailu-admin-credentials secret)

To retrieve the password later:

kubectl get secret mailu-admin-credentials -n bakery-ia -o jsonpath='{.data.password}' | base64 -d

Step 9: Configure DKIM

# Get DKIM public key from Mailu
kubectl exec -n bakery-ia deployment/mailu-admin -- \
  cat /dkim/bakewise.ai.dkim.pub

# Add DKIM record to DNS:
# Type: TXT
# Name: dkim._domainkey
# Value: (output from above command)

Step 10: Verify Email Setup

# Check all Mailu pods are running
kubectl get pods -n bakery-ia | grep mailu
# Expected: All pods in Running state

# Verify Mailgun secret is configured
kubectl get secret mailu-mailgun-credentials -n bakery-ia
kubectl get secret mailu-mailgun-credentials -n bakery-ia -o jsonpath='{.data.RELAY_USERNAME}' | base64 -d
# Should show: postmaster@bakewise.ai

# Test internal SMTP connectivity
kubectl run -it --rm smtp-test --image=alpine --restart=Never -- \
  sh -c "apk add swaks && swaks --to test@example.com --from admin@bakewise.ai --server mailu-front.bakery-ia.svc.cluster.local:25"

# Test outbound email via Mailgun relay (send test email)
kubectl exec -it -n bakery-ia deployment/mailu-admin -- \
  flask mailu alias_create test bakewise.ai 'your-personal-email@gmail.com'
# Then send a test email from webmail to your personal email

# Access webmail (via port-forward for testing)
kubectl port-forward -n bakery-ia svc/mailu-front 8080:80
# Open: http://localhost:8080/webmail

Mailu Endpoints

Service	URL/Address
Admin Panel	https://mail.bakewise.ai/admin
Webmail	https://mail.bakewise.ai/webmail
SMTP (STARTTLS)	mail.bakewise.ai:587
SMTP (SSL)	mail.bakewise.ai:465
IMAP (SSL)	mail.bakewise.ai:993

Mailu Troubleshooting

Admin Pod CrashLoopBackOff with DNSSEC Error

# Verify CoreDNS is forwarding to Unbound
kubectl get configmap coredns -n kube-system -o yaml | grep forward
# Should show: forward . <unbound-ip>

# If not configured, re-run Step 2

Front Pod Stuck in ContainerCreating

# Check for missing certificate secret
kubectl describe pod -n bakery-ia -l app.kubernetes.io/component=front | grep -A5 Events

# If missing mailu-certificates, re-run Step 4

Cannot Connect to Redis

# Verify internal Redis is enabled (not external)
helm get values mailu -n bakery-ia | grep -A5 externalRedis
# Should show: enabled: false

# If enabled: true, upgrade with correct values
helm upgrade mailu mailu/mailu -n bakery-ia \
  -f infrastructure/platform/mail/mailu-helm/values.yaml \
  -f infrastructure/platform/mail/mailu-helm/prod/values.yaml

Outbound Emails Not Delivered (Mailgun Relay Issues)

# Check if Mailgun credentials secret exists
kubectl get secret mailu-mailgun-credentials -n bakery-ia
# If missing, create it (see Step 3)

# Verify credentials are set correctly
kubectl get secret mailu-mailgun-credentials -n bakery-ia -o jsonpath='{.data.RELAY_USERNAME}' | base64 -d
# Should show your Mailgun username (e.g., postmaster@bakewise.ai)

# Check Postfix logs for relay errors
kubectl logs -n bakery-ia deployment/mailu-postfix | grep -i "relay\|mailgun\|sasl"
# Look for authentication errors or connection failures

# Verify Mailgun domain is verified
# Go to Mailgun dashboard > Domain Settings > DNS Records
# All records should show "Verified" status

# Test Mailgun SMTP connectivity directly
kubectl run -it --rm mailgun-test --image=alpine --restart=Never -- \
  sh -c "apk add swaks && swaks --to test@example.com --from postmaster@bakewise.ai \
    --server smtp.mailgun.org:587 --tls \
    --auth-user 'postmaster@bakewise.ai' \
    --auth-password 'YOUR_MAILGUN_PASSWORD'"

Emails Going to Spam

Verify SPF record includes Mailgun: v=spf1 include:mailgun.org mx a ~all
Check DKIM records are properly configured in both Mailgun and Mailu
Verify DMARC record is set
Check your domain reputation at mail-tester.com

Nominatim Geocoding Service

Nominatim provides geocoding (address to coordinates) and reverse geocoding for delivery and distribution features.

When to Deploy

Deploy Nominatim if you need:

Address autocomplete in the frontend
Delivery route optimization
Location-based analytics

Step 1: Deploy Nominatim via Helm

# Deploy Nominatim with production values
helm upgrade --install nominatim infrastructure/platform/nominatim/nominatim-helm \
  -n bakery-ia \
  --create-namespace \
  -f infrastructure/platform/nominatim/nominatim-helm/values.yaml \
  -f infrastructure/platform/nominatim/nominatim-helm/prod/values.yaml \
  --timeout 15m \
  --wait

# Verify deployment
kubectl get pods -n bakery-ia | grep nominatim

Note: Initial deployment may take 10-15 minutes as Nominatim downloads and processes geographic data.

Step 2: Verify Nominatim Service

# Check pod status
kubectl get pods -n bakery-ia -l app=nominatim

# Check service
kubectl get svc -n bakery-ia | grep nominatim

# Test geocoding endpoint
kubectl run -it --rm curl-test --image=curlimages/curl --restart=Never -- \
  curl "http://nominatim-service.bakery-ia.svc.cluster.local:8080/search?q=Madrid&format=json"

Step 3: Configure Application to Use Nominatim

Update the application ConfigMap to use the internal Nominatim service:

# Edit configmap
kubectl edit configmap bakery-ia-config -n bakery-ia

# Set:
# NOMINATIM_URL: "http://nominatim-service.bakery-ia.svc.cluster.local:8080"

Nominatim Service Information

Property	Value
Service Name	nominatim-service.bakery-ia.svc.cluster.local
Port	8080
Health Check	http://nominatim-service:8080/status
Search Endpoint	/search?q={query}&format=json
Reverse Endpoint	/reverse?lat={lat}&lon={lon}&format=json

SigNoz Monitoring Deployment

SigNoz provides unified observability (traces, metrics, logs) for the entire platform.

Step 1: Deploy SigNoz via Helm

# Add SigNoz Helm repository
helm repo add signoz https://charts.signoz.io
helm repo update signoz

# Deploy SigNoz into bakery-ia namespace
helm upgrade --install signoz signoz/signoz \
  -n bakery-ia \
  -f infrastructure/monitoring/signoz/signoz-values-prod.yaml \
  --set frontend.service.type=ClusterIP \
  --set clickhouse.persistence.size=20Gi \
  --set clickhouse.persistence.storageClass=microk8s-hostpath \
  --timeout 15m \
  --wait

# Wait for all components (may take 10-15 minutes)
kubectl wait --for=condition=ready pod \
  -l app.kubernetes.io/instance=signoz \
  -n bakery-ia \
  --timeout=900s

# Verify deployment
kubectl get pods -n bakery-ia -l app.kubernetes.io/instance=signoz

Step 2: Configure Ingress for SigNoz

# Apply SigNoz ingress (if not already included in overlays)
cat <<EOF | kubectl apply -f -
apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: signoz-ingress
  namespace: bakery-ia
  annotations:
    cert-manager.io/cluster-issuer: letsencrypt-production
    nginx.ingress.kubernetes.io/ssl-redirect: "true"
spec:
  ingressClassName: public
  tls:
    - hosts:
        - monitoring.bakewise.ai
      secretName: signoz-tls-cert
  rules:
    - host: monitoring.bakewise.ai
      http:
        paths:
          - path: /
            pathType: Prefix
            backend:
              service:
                name: signoz-frontend
                port:
                  number: 3301
EOF

Step 3: Verify SigNoz Access

# Check ingress
kubectl get ingress -n bakery-ia | grep signoz

# Test access
curl -I https://monitoring.bakewise.ai

# Access SigNoz UI
# URL: https://monitoring.bakewise.ai
# Default credentials: admin / admin (change after first login)

SigNoz Endpoints

Service	URL
SigNoz UI	https://monitoring.bakewise.ai
AlertManager	https://monitoring.bakewise.ai/alertmanager
OTel Collector (gRPC)	signoz-otel-collector:4317 (internal)
OTel Collector (HTTP)	signoz-otel-collector:4318 (internal)

Verification & Testing

Step 1: Check All Pods Running

# View all pods
kubectl get pods -n bakery-ia

# Expected: All pods in "Running" state, none in CrashLoopBackOff

# Check for issues
kubectl get pods -n bakery-ia | grep -vE "Running|Completed"

# View logs for any problematic pods
kubectl logs -n bakery-ia POD_NAME

Step 2: Check Services and Ingress

# View services
kubectl get svc -n bakery-ia

# View ingress
kubectl get ingress -n bakery-ia

# View certificates (should auto-issue from Let's Encrypt)
kubectl get certificate -n bakery-ia

# Describe certificate to check status
kubectl describe certificate bakery-ia-prod-tls-cert -n bakery-ia

Step 3: Test Database Connections

# Test PostgreSQL TLS
kubectl exec -n bakery-ia deployment/auth-db -- sh -c \
  'psql -U auth_user -d auth_db -c "SHOW ssl;"'
# Expected output: on

# Test Redis TLS
kubectl exec -n bakery-ia deployment/redis -- redis-cli \
  --tls \
  --cert /tls/redis-cert.pem \
  --key /tls/redis-key.pem \
  --cacert /tls/ca-cert.pem \
  -a $REDIS_PASSWORD \
  ping
# Expected output: PONG

Step 4: Test Frontend Access

# Test frontend (replace with your domain)
curl -I https://bakery.yourdomain.com

# Expected: HTTP/2 200 OK

# Test API health
curl https://api.yourdomain.com/health

# Expected: {"status": "healthy"}

Step 5: Test Authentication

# Create a test user (using your frontend or API)
curl -X POST https://api.yourdomain.com/api/v1/auth/register \
  -H "Content-Type: application/json" \
  -d '{
    "email": "test@yourdomain.com",
    "password": "TestPassword123!",
    "name": "Test User"
  }'

# Login
curl -X POST https://api.yourdomain.com/api/v1/auth/login \
  -H "Content-Type: application/json" \
  -d '{
    "email": "test@yourdomain.com",
    "password": "TestPassword123!"
  }'

# Expected: JWT token in response

Step 6: Test Email Delivery

# Trigger a password reset to test email
curl -X POST https://api.yourdomain.com/api/v1/auth/forgot-password \
  -H "Content-Type: application/json" \
  -d '{"email": "test@yourdomain.com"}'

# Check your email inbox for the reset link
# Check service logs if email not received:
kubectl logs -n bakery-ia deployment/auth-service | grep -i "email\|smtp"

Step 7: Test WhatsApp (Optional)

# Send a test WhatsApp message
# This requires creating a tenant and configuring WhatsApp in the UI
# Or test via API once authenticated

Post-Deployment

Step 1: Access SigNoz Monitoring Stack

Your production deployment includes SigNoz, a unified observability platform that provides complete visibility into your application:

What is SigNoz?

SigNoz is an open-source, all-in-one observability platform that provides:

📊 Distributed Tracing - See end-to-end request flows across all 18 microservices
📈 Metrics Monitoring - Application performance and infrastructure metrics
📝 Log Management - Centralized logs from all services with trace correlation
🔍 Service Performance Monitoring (SPM) - Automatic RED metrics (Rate, Error, Duration)
🗄️ Database Monitoring - All 18 PostgreSQL databases + Redis + RabbitMQ
☸️ Kubernetes Monitoring - Cluster, node, pod, and container metrics

Why SigNoz instead of Prometheus/Grafana?

Single unified UI for traces, metrics, and logs (no context switching)
Automatic service dependency mapping
Built-in APM (Application Performance Monitoring)
Log-trace correlation with one click
Better query performance with ClickHouse backend
Modern UI designed for microservices

Production Monitoring URLs

Access via domain:

https://monitoring.bakewise.ai/signoz        # SigNoz - Main observability UI
https://monitoring.bakewise.ai/alertmanager  # AlertManager - Alert management

Or via port forwarding (if needed):

# SigNoz Frontend (Main UI)
kubectl port-forward -n bakery-ia svc/signoz 8080:8080 &
# Open: http://localhost:8080

# SigNoz AlertManager
kubectl port-forward -n bakery-ia svc/signoz-alertmanager 9093:9093 &
# Open: http://localhost:9093

# OTel Collector (for debugging)
kubectl port-forward -n bakery-ia svc/signoz-otel-collector 4317:4317 &  # gRPC
kubectl port-forward -n bakery-ia svc/signoz-otel-collector 4318:4318 &  # HTTP

Key SigNoz Features to Explore

Once you open SigNoz (https://monitoring.bakewise.ai/signoz), explore these tabs:

1. Services Tab - Application Performance

View all 18 microservices with live metrics
See request rate, error rate, and latency (P50/P90/P99)
Click on any service to drill down into operations
Identify slow endpoints and error-prone operations

2. Traces Tab - Request Flow Visualization

See complete request journeys across services
Identify bottlenecks (slow database queries, API calls)
Debug errors with full stack traces
Correlate with logs for complete context

3. Dashboards Tab - Infrastructure & Database Metrics

PostgreSQL - Monitor all 18 databases (connections, queries, cache hit ratio)
Redis - Cache performance (memory, hit rate, commands/sec)
RabbitMQ - Message queue health (depth, rates, consumers)
Kubernetes - Cluster metrics (nodes, pods, containers)

4. Logs Tab - Centralized Log Management

Search and filter logs from all services
Click on trace ID in logs to see related request trace
Auto-enriched with Kubernetes metadata (pod, namespace, container)
Identify patterns and anomalies

5. Alerts Tab - Proactive Monitoring

Configure alerts on metrics, traces, or logs
Email/Slack/Webhook notifications
View firing alerts and alert history

Quick Health Check

# Verify SigNoz components are running
kubectl get pods -n bakery-ia -l app.kubernetes.io/instance=signoz

# Expected output:
# signoz-0                              READY 1/1
# signoz-otel-collector-xxx             READY 1/1
# signoz-alertmanager-xxx               READY 1/1
# signoz-clickhouse-xxx                 READY 1/1
# signoz-zookeeper-xxx                  READY 1/1

# Check OTel Collector health
kubectl exec -n bakery-ia deployment/signoz-otel-collector -- wget -qO- http://localhost:13133

# View recent telemetry in OTel Collector logs
kubectl logs -n bakery-ia deployment/signoz-otel-collector --tail=50 | grep -i "traces\|metrics\|logs"

Verify Telemetry is Working

Check Services are Reporting:

# Open SigNoz and navigate to Services tab
# You should see all 18 microservices listed

# If services are missing, check if they're sending telemetry:
kubectl logs -n bakery-ia deployment/auth-service | grep -i "telemetry\|otel"

Check Database Metrics:

# Navigate to Dashboards → PostgreSQL in SigNoz
# You should see metrics from all 18 databases

# Verify OTel Collector is scraping databases:
kubectl logs -n bakery-ia deployment/signoz-otel-collector | grep postgresql

Check Traces are Being Collected:

# Make a test API request
curl https://bakewise.ai/api/v1/health

# Navigate to Traces tab in SigNoz
# Search for "gateway" service
# You should see the trace for your request

Check Logs are Being Collected:

# Navigate to Logs tab in SigNoz
# Filter by namespace: bakery-ia
# You should see logs from all pods

# Verify filelog receiver is working:
kubectl logs -n bakery-ia deployment/signoz-otel-collector | grep filelog

Step 2: Configure CI/CD Infrastructure (Optional but Recommended)

If you deployed the CI/CD infrastructure, configure it for your workflow:

Gitea Setup (Git Server + Registry)

# Access Gitea at: http://gitea.bakery-ia.local (for dev) or http://gitea.bakewise.ai (for prod)
# Make sure to add the appropriate hostname to /etc/hosts or configure DNS

# Create your repositories for each service
# Configure webhook to trigger Tekton pipelines

Tekton Pipeline Configuration

# Verify Tekton pipelines are running
kubectl get pods -n tekton-pipelines

# Create a PipelineRun manually to test:
kubectl create -f - <<EOF
apiVersion: tekton.dev/v1beta1
kind: PipelineRun
metadata:
  name: manual-ci-run
  namespace: tekton-pipelines
spec:
  pipelineRef:
    name: bakery-ia-ci
  workspaces:
    - name: shared-workspace
      volumeClaimTemplate:
        spec:
          accessModes: ["ReadWriteOnce"]
          resources:
            requests:
              storage: 5Gi
    - name: docker-credentials
      secret:
        secretName: gitea-registry-credentials
  params:
    - name: git-url
      value: "http://gitea.bakery-ia.local/bakery-admin/bakery-ia.git"
    - name: git-revision
      value: "main"
EOF

Flux CD Configuration (GitOps)

# Verify Flux is running
kubectl get pods -n flux-system

# Set up GitRepository and Kustomization resources for GitOps deployment
# Example:
cat <<EOF | kubectl apply -f -
apiVersion: source.toolkit.fluxcd.io/v1
kind: GitRepository
metadata:
  name: bakery-ia
  namespace: flux-system
spec:
  interval: 1m
  url: https://github.com/your-org/bakery-ia.git
  ref:
    branch: main
---
apiVersion: kustomize.toolkit.fluxcd.io/v1
kind: Kustomization
metadata:
  name: bakery-ia
  namespace: flux-system
spec:
  interval: 5m
  sourceRef:
    kind: GitRepository
    name: bakery-ia
  path: ./infrastructure/environments/prod/k8s-manifests
  prune: true
  validation: client
EOF

Step 2: Configure Alerting

SigNoz includes integrated alerting with AlertManager. Configure it for your team:

Update Email Notification Settings

The alerting configuration is in the SigNoz Helm values. To update:

# For production, edit the values file:
nano infrastructure/helm/signoz-values-prod.yaml

# Update the alertmanager.config section:
# 1. Update SMTP settings:
#    - smtp_from: 'your-alerts@bakewise.ai'
#    - smtp_auth_username: 'your-alerts@bakewise.ai'
#    - smtp_auth_password: (use Kubernetes secret)
#
# 2. Update receivers:
#    - critical-alerts email: critical-alerts@bakewise.ai
#    - warning-alerts email: oncall@bakewise.ai
#
# 3. (Optional) Add Slack webhook for critical alerts

# Apply the updated configuration:
helm upgrade signoz signoz/signoz \
  -n bakery-ia \
  -f infrastructure/helm/signoz-values-prod.yaml

Create Alerts in SigNoz UI

Open SigNoz Alerts Tab:

https://monitoring.bakewise.ai/signoz → Alerts

Create Common Alerts:

Alert 1: High Error Rate
- Name: HighErrorRate
- Query: error_rate > 5 for 5 minutes
- Severity: critical
- Description: "Service {{service_name}} has error rate >5%"
Alert 2: High Latency
- Name: HighLatency
- Query: P99_latency > 3000ms for 5 minutes
- Severity: warning
- Description: "Service {{service_name}} P99 latency >3s"
Alert 3: Service Down
- Name: ServiceDown
- Query: request_rate == 0 for 2 minutes
- Severity: critical
- Description: "Service {{service_name}} not receiving requests"
Alert 4: Database Connection Issues
- Name: DatabaseConnectionsHigh
- Query: pg_active_connections > 80 for 5 minutes
- Severity: warning
- Description: "Database {{database}} connection count >80%"
Alert 5: High Memory Usage
- Name: HighMemoryUsage
- Query: container_memory_percent > 85 for 5 minutes
- Severity: warning
- Description: "Pod {{pod_name}} using >85% memory"

Test Alert Delivery

# Method 1: Create a test alert in SigNoz UI
# Go to Alerts → New Alert → Set a test condition that will fire

# Method 2: Fire a test alert via stress test
kubectl run memory-test --image=polinux/stress --restart=Never \
  --namespace=bakery-ia -- stress --vm 1 --vm-bytes 600M --timeout 300s

# Check alert appears in SigNoz Alerts tab
# https://monitoring.bakewise.ai/signoz → Alerts

# Also check AlertManager
# https://monitoring.bakewise.ai/alertmanager

# Verify email notification received

# Clean up test
kubectl delete pod memory-test -n bakery-ia

Configure Notification Channels

In SigNoz Alerts tab, configure channels:

Email Channel:
- Already configured via AlertManager
- Emails sent to addresses in signoz-values-prod.yaml

Slack Channel (Optional):

# Add Slack webhook URL to signoz-values-prod.yaml
# Under alertmanager.config.receivers.critical-alerts.slack_configs:
#   - api_url: 'https://hooks.slack.com/services/YOUR/WEBHOOK/URL'
#     channel: '#alerts-critical'

Webhook Channel (Optional):
- Configure custom webhook for integration with PagerDuty, OpsGenie, etc.
- Add to alertmanager.config.receivers

Step 3: Configure Backups

# Create backup script on VPS
cat > ~/backup-databases.sh <<'EOF'
#!/bin/bash
BACKUP_DIR="/backups/$(date +%Y-%m-%d)"
mkdir -p $BACKUP_DIR

# Get all database pods
DBS=$(kubectl get pods -n bakery-ia -l app.kubernetes.io/component=database -o name)

for db in $DBS; do
  DB_NAME=$(echo $db | cut -d'/' -f2)
  echo "Backing up $DB_NAME..."

  kubectl exec -n bakery-ia $db -- pg_dump -U postgres > "$BACKUP_DIR/${DB_NAME}.sql"
done

# Compress backups
tar -czf "$BACKUP_DIR.tar.gz" "$BACKUP_DIR"
rm -rf "$BACKUP_DIR"

# Keep only last 7 days
find /backups -name "*.tar.gz" -mtime +7 -delete

echo "Backup completed: $BACKUP_DIR.tar.gz"
EOF

chmod +x ~/backup-databases.sh

# Test backup
./backup-databases.sh

# Setup daily cron job (2 AM)
(crontab -l 2>/dev/null; echo "0 2 * * * ~/backup-databases.sh") | crontab -

Step 3: Setup Alerting

# Update AlertManager configuration with your email
kubectl edit configmap -n monitoring alertmanager-config

# Update recipient emails in the routes section

Step 4: Verify SigNoz Monitoring is Working

Before proceeding, ensure all monitoring components are operational:

# 1. Verify SigNoz pods are running
kubectl get pods -n bakery-ia -l app.kubernetes.io/instance=signoz

# Expected pods (all should be Running/Ready):
# - signoz-0 (or signoz-1, signoz-2 for HA)
# - signoz-otel-collector-xxx
# - signoz-alertmanager-xxx
# - signoz-clickhouse-xxx
# - signoz-zookeeper-xxx

# 2. Check SigNoz UI is accessible
curl -I https://monitoring.bakewise.ai/signoz
# Should return: HTTP/2 200 OK

# 3. Verify OTel Collector is receiving data
kubectl logs -n bakery-ia deployment/signoz-otel-collector --tail=100 | grep -i "received"
# Should show: "Traces received: X" "Metrics received: Y" "Logs received: Z"

# 4. Check ClickHouse database is healthy
kubectl exec -n bakery-ia deployment/signoz-clickhouse -- clickhouse-client --query="SELECT count() FROM system.tables WHERE database LIKE 'signoz_%'"
# Should return a number > 0 (tables exist)

Complete Verification Checklist:

SigNoz UI loads at https://monitoring.bakewise.ai/signoz
Services tab shows all 18 microservices with metrics
Traces tab has sample traces from gateway and other services
Dashboards tab shows PostgreSQL metrics from all 18 databases
Dashboards tab shows Redis metrics (memory, commands, etc.)
Dashboards tab shows RabbitMQ metrics (queues, messages)
Dashboards tab shows Kubernetes metrics (nodes, pods)
Logs tab displays logs from all services in bakery-ia namespace
Alerts tab is accessible and can create new alerts
AlertManager is reachable at https://monitoring.bakewise.ai/alertmanager

If any checks fail, troubleshoot:

# Check OTel Collector configuration
kubectl describe configmap -n bakery-ia signoz-otel-collector

# Check for errors in OTel Collector
kubectl logs -n bakery-ia deployment/signoz-otel-collector | grep -i error

# Check ClickHouse is accepting writes
kubectl logs -n bakery-ia deployment/signoz-clickhouse | grep -i error

# Restart OTel Collector if needed
kubectl rollout restart deployment/signoz-otel-collector -n bakery-ia

Step 5: Document Everything

Create a secure runbook with all credentials and procedures:

Essential Information to Document:

VPS login credentials (stored securely in password manager)
Database passwords (in password manager)
Grafana admin password
Domain registrar access (for bakewise.ai)
Cloudflare access
Email service credentials (SMTP)
WhatsApp API credentials
Docker Hub / Registry credentials
Emergency contact information
Rollback procedures
Monitoring URLs and access procedures

Step 6: Train Your Team

Conduct a training session covering SigNoz and operational procedures:

Login and Overview
- Show how to access https://monitoring.bakewise.ai/signoz
- Navigate through main tabs: Services, Traces, Dashboards, Logs, Alerts
- Explain the unified nature of SigNoz (all-in-one platform)
Services Tab - Application Performance Monitoring
- Show all 18 microservices
- Explain RED metrics (Request rate, Error rate, Duration/latency)
- Demo: Click on a service → Operations → See endpoint breakdown
- Demo: Identify slow endpoints and high error rates
Traces Tab - Request Flow Debugging
- Show how to search for traces by service, operation, or time
- Demo: Click on a trace → See full waterfall (service → database → cache)
- Demo: Find slow database queries in trace spans
- Demo: Click "View Logs" to correlate trace with logs
Dashboards Tab - Infrastructure Monitoring
- Navigate to PostgreSQL dashboard → Show all 18 databases
- Navigate to Redis dashboard → Show cache metrics
- Navigate to Kubernetes dashboard → Show node/pod metrics
- Explain what metrics indicate issues (connection %, memory %, etc.)
Logs Tab - Log Search and Analysis
- Show how to filter by service, severity, time range
- Demo: Search for "error" in last hour
- Demo: Click on trace_id in log → Jump to related trace
- Show Kubernetes metadata (pod, namespace, container)
Alerts Tab - Proactive Monitoring
- Show how to create alerts on metrics
- Review pre-configured alerts
- Show alert history and firing alerts
- Explain how to acknowledge/silence alerts

Part 2: Operational Tasks (30 minutes)

Check application logs (multiple ways)

# Method 1: Via kubectl (for immediate debugging)
kubectl logs -n bakery-ia deployment/orders-service --tail=100 -f

# Method 2: Via SigNoz Logs tab (for analysis and correlation)
# 1. Open https://monitoring.bakewise.ai/signoz → Logs
# 2. Filter by k8s_deployment_name: orders-service
# 3. Click on trace_id to see related request flow

Restart services when needed

# Restart a service (rolling update, no downtime)
kubectl rollout restart deployment/orders-service -n bakery-ia

# Verify restart in SigNoz:
# 1. Check Services tab → orders-service → Should show brief dip then recovery
# 2. Check Logs tab → Filter by orders-service → See restart logs

Investigate performance issues

# Scenario: "Orders API is slow"
# 1. SigNoz → Services → orders-service → Check P99 latency
# 2. SigNoz → Traces → Filter service:orders-service, duration:>1s
# 3. Click on slow trace → Identify bottleneck (DB query? External API?)
# 4. SigNoz → Dashboards → PostgreSQL → Check orders_db connections/queries
# 5. Fix identified issue (add index, optimize query, scale service)

Respond to alerts
- Show how to access alerts in SigNoz → Alerts tab
- Show AlertManager UI at https://monitoring.bakewise.ai/alertmanager
- Review common alerts and their resolution steps
- Reference the Production Operations Guide

Part 3: Documentation and Resources (10 minutes)

Share documentation
- PILOT_LAUNCH_GUIDE.md - This guide (deployment)
- PRODUCTION_OPERATIONS_GUIDE.md - Daily operations with SigNoz
- security-checklist.md - Security procedures
Bookmark key URLs
- SigNoz: https://monitoring.bakewise.ai/signoz
- AlertManager: https://monitoring.bakewise.ai/alertmanager
- Production app: https://bakewise.ai
Setup on-call rotation (if applicable)
- Configure rotation schedule in AlertManager
- Document escalation procedures
- Test alert delivery to on-call phone/email

Part 4: Hands-On Exercise (15 minutes)

Exercise: Investigate a Simulated Issue

Create a load test to generate traffic
Use SigNoz to find the slowest endpoint
Identify the root cause using traces
Correlate with logs to confirm
Check infrastructure metrics (DB, memory, CPU)
Propose a fix based on findings

This trains the team to use SigNoz effectively for real incidents.

Troubleshooting

Issue: Pods Not Starting

# Check pod status
kubectl describe pod POD_NAME -n bakery-ia

# Common causes:
# 1. Image pull errors
kubectl get events -n bakery-ia | grep -i "pull"

# 2. Resource limits
kubectl describe node

# 3. Volume mount issues
kubectl get pvc -n bakery-ia

Issue: Certificate Not Issuing

# Check certificate status
kubectl describe certificate bakery-ia-prod-tls-cert -n bakery-ia

# Check cert-manager logs
kubectl logs -n cert-manager deployment/cert-manager

# Check challenges
kubectl get challenges -n bakery-ia

# Verify DNS is correct
nslookup bakery.yourdomain.com

Issue: Database Connection Errors

# Check database pod
kubectl get pods -n bakery-ia -l app.kubernetes.io/component=database

# Check database logs
kubectl logs -n bakery-ia deployment/auth-db

# Test connection from service pod
kubectl exec -n bakery-ia deployment/auth-service -- nc -zv auth-db 5432

Issue: Services Can't Connect to Databases

# Check if SSL is enabled
kubectl exec -n bakery-ia deployment/auth-db -- sh -c \
  'psql -U auth_user -d auth_db -c "SHOW ssl;"'

# Check service logs for SSL errors
kubectl logs -n bakery-ia deployment/auth-service | grep -i "ssl\|tls"

# Restart service to pick up new SSL config
kubectl rollout restart deployment/auth-service -n bakery-ia

Issue: Out of Resources

# Check node resources
kubectl top nodes

# Check pod resource usage
kubectl top pods -n bakery-ia

# Identify resource hogs
kubectl top pods -n bakery-ia --sort-by=memory

# Scale down non-critical services temporarily
kubectl scale deployment monitoring -n bakery-ia --replicas=0

Next Steps After Successful Launch

Monitor for 48 Hours
- Check dashboards daily
- Review error logs
- Monitor resource usage
- Test all functionality
Optimize Based on Metrics
- Adjust resource limits if needed
- Fine-tune autoscaling thresholds
- Optimize database queries if slow
Onboard First Tenant
- Create test tenant
- Upload sample data
- Test all features
- Gather feedback
Scale Gradually
- Add 1-2 tenants at a time
- Monitor resource usage
- Upgrade VPS if needed (see scaling guide)
Plan for Growth
- Review PRODUCTION_OPERATIONS_GUIDE.md
- Implement additional monitoring
- Plan capacity upgrades
- Consider managed services for scale

Cost Scaling Path

Tenants	RAM	CPU	Storage	Monthly Cost
10	20 GB	8 cores	200 GB	€40-80
25	32 GB	12 cores	300 GB	€80-120
50	48 GB	16 cores	500 GB	€150-200
100+	Consider multi-node cluster or managed K8s	€300+

Support Resources

Documentation:

Operations Guide: PRODUCTION_OPERATIONS_GUIDE.md - Daily operations, monitoring, incident response
Security Guide: security-checklist.md - Security procedures and compliance
Database Security: database-security.md - Database operations and TLS configuration
TLS Configuration: tls-configuration.md - Certificate management
RBAC Implementation: rbac-implementation.md - Access control

Monitoring Access:

SigNoz (Primary): https://monitoring.bakewise.ai/signoz - All-in-one observability
- Services: Application performance monitoring (APM)
- Traces: Distributed tracing across all services
- Dashboards: PostgreSQL, Redis, RabbitMQ, Kubernetes metrics
- Logs: Centralized log management with trace correlation
- Alerts: Alert configuration and management
AlertManager: https://monitoring.bakewise.ai/alertmanager - Alert routing and notifications

External Resources:

MicroK8s Docs: https://microk8s.io/docs
Kubernetes Docs: https://kubernetes.io/docs
Let's Encrypt: https://letsencrypt.org/docs
Cloudflare DNS: https://developers.cloudflare.com/dns
SigNoz Documentation: https://signoz.io/docs/
OpenTelemetry Documentation: https://opentelemetry.io/docs/

Monitoring Architecture:

OpenTelemetry: Industry-standard instrumentation framework
- Auto-instruments FastAPI, HTTPX, SQLAlchemy, Redis
- Collects traces, metrics, and logs from all services
- Exports to SigNoz via OTLP protocol (gRPC port 4317, HTTP port 4318)
SigNoz Components:
- Frontend: Web UI for visualization and analysis
- OTel Collector: Receives and processes telemetry data
- ClickHouse: Time-series database for fast queries
- AlertManager: Alert routing and notification delivery
- Zookeeper: Coordination service for ClickHouse cluster

Summary Checklist

Pre-Deployment Configuration (LOCAL MACHINE)

Production secrets configured - ✅ JWT, database passwords, API keys (ALREADY DONE)
External service credentials - Update SMTP, WhatsApp, Stripe in secrets.yaml
imagePullSecrets removed - Delete from all 67 manifests
Image tags updated - Change all 'latest' to v1.0.0 (semantic version)
SigNoz namespace fixed - ✅ Already done (bakery-ia namespace)
Cert-manager email updated - ✅ Already set to admin@bakewise.ai
Stripe publishable key updated - Replace pk_test_... with production key in configmap.yaml
Pilot mode verified - ✅ VITE_PILOT_MODE_ENABLED=true (default is correct)
Manifests validated - No 'latest' tags, no imagePullSecrets remaining

Infrastructure Setup

VPS provisioned and accessible
k3s (or Kubernetes) installed and configured
nginx-ingress-controller installed
metrics-server installed and working
cert-manager installed
local-path-provisioner installed
Domain registered and DNS configured
Cloudflare protection enabled (optional but recommended)

Secrets and Configuration

TLS certificates generated (postgres, redis)
Email service configured and tested
WhatsApp API setup (optional for launch)
Container images built and pushed with version tags
Production configs verified (domains, CORS, storage class)
Strong passwords generated for all services
Docker registry secret created (dockerhub-creds)
Application secrets applied

Monitoring

SigNoz deployed via Helm
SigNoz pods running and healthy
SigNoz in bakery-ia namespace

CI/CD Infrastructure (Optional)

Gitea deployed and accessible
Gitea admin user created
Repository created and code pushed
Tekton Pipelines installed
Tekton Triggers configured
Tekton Helm chart deployed
Webhook configured in Gitea
Flux CD installed
GitRepository and Kustomization configured
End-to-end pipeline test successful

Email Infrastructure (Optional - Mailu)

Unbound DNS resolver deployed
CoreDNS configured for DNSSEC
Mailu TLS certificate created
Mailu deployed via Helm
Admin user created
DKIM record added to DNS
Email sending/receiving tested

Geocoding (Optional - Nominatim)

Nominatim deployed
Health check passing
Application configured to use Nominatim

Application Deployment

All pods running successfully
Databases accepting TLS connections
Let's Encrypt certificates issued
Frontend accessible via HTTPS
API health check passing
Test user can login
Email delivery working
SigNoz monitoring accessible
Metrics flowing to SigNoz
Pilot coupon verified - Check tenant-service logs for "Pilot coupon created successfully"

Post-Deployment

Backups configured and tested
Team trained on operations
Documentation complete
Emergency procedures documented
Monitoring alerts configured

🎉 Congratulations! Your Bakery-IA platform is now live in production!

Estimated total time: 2-4 hours for first deployment Subsequent updates: 15-30 minutes

Document Version: 3.0 Last Updated: 2026-01-21 Maintained By: DevOps Team

Changes in v3.0:

NEW: Infrastructure Architecture Overview - Added component layers diagram and deployment dependencies
NEW: CI/CD Infrastructure Deployment - Complete guide for Gitea, Tekton, and Flux CD
- Step-by-step Gitea installation with container registry
- Tekton Pipelines and Triggers setup via Helm
- Flux CD GitOps configuration
- Webhook integration and end-to-end testing
- Troubleshooting guide for CI/CD issues
NEW: Mailu Email Server Deployment - Comprehensive self-hosted email setup
- Unbound DNS resolver deployment for DNSSEC
- CoreDNS configuration for mail authentication
- Mailu Helm deployment with all components
- DKIM/SPF/DMARC configuration
- Troubleshooting common Mailu issues
NEW: Nominatim Geocoding Service - Address lookup service deployment
NEW: SigNoz Monitoring Deployment - Dedicated section (previously embedded)
UPDATED: Table of Contents - Reorganized with new sections (18 sections total)
UPDATED: Summary Checklist - Added CI/CD, Email, and Geocoding verification items
UPDATED: Infrastructure Components Summary - Added all optional components with namespaces

Changes in v2.1:

Updated DNS configuration for Namecheap (primary) with Cloudflare as optional
Clarified MicroK8s ingress class is public (not nginx)
Updated Let's Encrypt ClusterIssuer documentation to reference pre-configured files
Added firewall requirements for clouding.io VPS
Emphasized port 80/443 requirements for HTTP-01 challenges

Changes in v2.0:

Added critical pre-deployment fixes section
Updated infrastructure setup for MicroK8s
Added required component installation (nginx-ingress, metrics-server, etc.)
Updated configuration steps with domain replacement
Added Docker registry secret creation
Added SigNoz Helm deployment before application
Updated storage class configuration
Added image tag version requirements
Expanded verification checklist

112 KiB Raw Blame History Unescape Escape

Bakery-IA Pilot Launch Guide

Table of Contents

Executive Summary

What You're Deploying

Total Cost Breakdown

Timeline

Infrastructure Architecture Overview

Component Layers

Deployment Order & Dependencies

Infrastructure Components Summary

Quick Reference: What to Deploy

⚠️ CRITICAL: Pre-Deployment Configuration

Infrastructure Architecture (Updated)

🔴 Configuration Status

Required Configuration Changes

1. imagePullSecrets - ✅ AUTOMATICALLY HANDLED

2. Update Image Tags to Semantic Versions (FOR RELEASES)

3. Production Secrets - ✅ ALREADY CONFIGURED

4. Cert-Manager Email - ✅ ALREADY CONFIGURED

5. Update Stripe Keys (HIGH PRIORITY)

6. Pilot Coupon Configuration - ✅ AUTO-SEEDED

✅ Already Correct (No Changes Needed)

Step-by-Step Configuration Script

Manual Verification

Key File Locations Reference

Pre-Launch Checklist

Required Accounts & Services

Required Tools on Local Machine

Repository Setup

VPS Provisioning

Recommended Configuration

Why These Specs?

Provisioning Steps

Infrastructure Setup

Step 1: Install MicroK8s

Step 2: Enable Required MicroK8s Addons

Step 3: Enhanced Infrastructure Components

Step 4: Configure Firewall

Step 5: Create Namespace

Domain & DNS Configuration

Step 1: Register Domain at Namecheap

Step 2: Configure DNS at Namecheap

Step 3 (Optional): Configure Cloudflare DNS

TLS/SSL Certificates

Understanding Certificate Setup

Step 1: Generate Internal Certificates

Step 2: Create Kubernetes Secrets

Step 3: Configure Let's Encrypt (External SSL)

Email & Communication Setup

Self-Hosted Mailu with Mailgun Relay

Prerequisites

Step 1: Configure DNS Records

Step 2: Deploy Unbound DNS Resolver

Step 3: Configure CoreDNS for DNSSEC

Step 4: Create TLS Certificate Secret

Step 5: Create Admin Credentials Secret

Step 6: Deploy Mailu via Helm

Step 7: Configure DKIM

Step 8: Verify Email Setup

Production Email Endpoints

Troubleshooting Mailu

WhatsApp Business API Setup

Kubernetes Deployment

Step 1: Prepare Container Images

Option A: Using Docker Hub (Recommended)

Option B: Using MicroK8s Registry

Step 2: Update Production Configuration

Required Configuration Updates

Configuration & Secrets

Production Secrets Status ✅

Step 1: Configure External Service Credentials (Email & WhatsApp)

Step 2: CI/CD Secrets Configuration

Step 3: Apply Application Secrets

Database Migrations

Step 0: Deploy CI/CD Infrastructure (Optional but Recommended)

Step 1: Deploy SigNoz Monitoring (BEFORE Application)

Step 2: Deploy Application and Databases

Step 2: Run Migrations

Step 3: Verify Database Schemas

112 KiB

Raw Blame History