sriphat-dataplatform/document-obsidiant/2026-SRI-PJ-001 Sriphat AI Transformation/08-Operations-Runbook.md

tags, created, status

tags	created	status
project/sriphat operations runbook deployment troubleshooting	2026-05-07	active

Operations Runbook

Quick Reference — Service Status

# ดู containers ทั้งหมดที่รันอยู่
docker ps

# ดู resource usage
docker stats

# ดู logs แบบ realtime (ผ่าน Dozzle)
# https://ai.sriphat.com/dozzle

---

First-Time Deployment

Prerequisites

• Docker + Docker Compose installed
• RAM ≥ 8 GB
• Disk ≥ 50 GB
• Port 80, 443 accessible

Step 1: Setup Network

# สร้าง Docker network ร่วม
docker network create shared_data_network

Step 2: Configure Environment

# Copy และแก้ไขค่า
cp .env.example .env.global
nano .env.global

# Supabase env
cd 02-supabase
cp .env.example .env
nano .env

# API Service env
cd ../03-apiservice
cp .env.example .env
nano .env

# MinIO env
cd ../07-minio
cp .env.example .env
nano .env

Step 3: Start Services (ตามลำดับ)

# 1. Infrastructure (Nginx + Keycloak + PostgreSQL + Redis)
cd 01-infra
docker compose --env-file ../.env.global up -d

# รอ PostgreSQL พร้อม (~30 วินาที)
sleep 30

# 2. Supabase
cd ../02-supabase
docker compose up -d

# รอ Supabase DB พร้อม (~60 วินาที)
sleep 60

# 3. API Service
cd ../03-apiservice
docker compose --env-file ../.env.global up --build -d

# 4. Airflow (ถ้าใช้งาน)
cd ../05-airflow
docker compose up airflow-init   # รอให้ init เสร็จ
docker compose up -d

# 5. Analytics (Superset)
cd ../06-analytics
docker compose --env-file ../.env.global build
docker compose --env-file ../.env.global up -d

# 6. MinIO
cd ../07-minio
docker compose up -d

Step 4: Verify Services

docker ps --format "table {{.Names}}\t{{.Status}}\t{{.Ports}}"

Expected containers:

• nginx-proxy-manager ✅
• keycloak ✅
• postgres ✅
• redis ✅
• dozzle ✅
• sdp-supabase-db ✅
• sdp-supabase-studio ✅
• sdp-supabase-kong ✅
• sdp-supabase-auth ✅
• sdp-supabase-rest ✅
• sdp-realtime-dev ✅
• sdp-supabase-storage ✅
• sdp-supabase-pooler ✅
• sdp-supabase-analytics ✅
• sdp-supabase-meta ✅
• apiservice ✅
• superset ✅
• minio ✅

---

Post-Installation Setup

Setup Keycloak

1. เข้า https://ai.sriphat.com/keycloak/admin
2. Login ด้วย KEYCLOAK_ADMIN credentials
3. สร้าง Realm: sriphat
4. สร้าง Clients:
   • apiservice (API Service)
   • superset-client (Apache Superset)
   • minio-client (MinIO)
   • airflow-client (Apache Airflow)
5. เชื่อมต่อ LDAP/AD (optional)

Initialize API Service

# เข้า Admin UI
https://ai.sriphat.com/apiservice/admin/

# สร้าง API Client และ Generate Key
curl -X POST "https://ai.sriphat.com/apiservice/admin/api-keys/generate" \
  -H "Cookie: session=<admin-session>" \
  -d "client_id=1&permissions=feed.checkpoint:write&name=his-key"

Setup Superset Data Sources

1. เข้า https://ai.sriphat.com/superset
2. Settings → Database Connections → Add
3. เพิ่ม PostgreSQL:

   postgresql://postgres:<password>@postgres:5432/postgres
   

---

Daily Operations

Start All Services

# Infrastructure
cd 01-infra && docker compose --env-file ../.env.global up -d && cd ..

# Supabase
cd 02-supabase && docker compose up -d && cd ..

# API Service
cd 03-apiservice && docker compose --env-file ../.env.global up -d && cd ..

# Airflow
cd 05-airflow && docker compose up -d && cd ..

# Analytics
cd 06-analytics && docker compose --env-file ../.env.global up -d && cd ..

# MinIO
cd 07-minio && docker compose up -d && cd ..

Stop All Services

cd 01-infra && docker compose down && cd ..
cd 02-supabase && docker compose down && cd ..
cd 03-apiservice && docker compose down && cd ..
cd 05-airflow && docker compose down && cd ..
cd 06-analytics && docker compose down && cd ..
cd 07-minio && docker compose down && cd ..

---

Backup & Restore

Backup PostgreSQL (Infra)

# Backup ทั้ง database
docker exec postgres pg_dumpall -U postgres > backup_all_$(date +%Y%m%d_%H%M).sql

# Backup เฉพาะ database
docker exec postgres pg_dump -U postgres postgres > backup_postgres_$(date +%Y%m%d).sql
docker exec postgres pg_dump -U postgres superset > backup_superset_$(date +%Y%m%d).sql
docker exec postgres pg_dump -U postgres airflow > backup_airflow_$(date +%Y%m%d).sql

Backup Supabase PostgreSQL

docker exec sdp-supabase-db pg_dump -U postgres postgres > backup_supabase_$(date +%Y%m%d).sql

Backup MinIO

# ใช้ mc mirror
mc mirror sriphat/ ./minio-backup-$(date +%Y%m%d)/

# หรือ tar data directory
tar -czf minio-backup-$(date +%Y%m%d).tar.gz 07-minio/data/

Restore PostgreSQL

# Restore
docker exec -i postgres psql -U postgres postgres < backup_postgres_20260501.sql

---

Update Services

# Pull latest images
cd 01-infra && docker compose --env-file ../.env.global pull && cd ..
cd 02-supabase && docker compose pull && cd ..
cd 06-analytics && docker compose --env-file ../.env.global pull && cd ..
cd 07-minio && docker compose pull && cd ..

# Rebuild API Service (มี code changes)
cd 03-apiservice
docker compose --env-file ../.env.global build
docker compose --env-file ../.env.global up -d

---

Troubleshooting

PostgreSQL ไม่ start / connection refused

# ตรวจสอบ health
docker exec postgres pg_isready -U postgres

# ดู logs
docker logs postgres --tail 50

# Check schemas
docker exec postgres psql -U postgres -c "\dn"

Keycloak ไม่ start

# ดู logs (มักเกิดจาก PostgreSQL ยังไม่พร้อม)
docker logs keycloak --tail 50

# Restart หลัง PostgreSQL พร้อม
docker restart keycloak

API Service ไม่ connect database

# ตรวจสอบ network
docker network inspect shared_data_network

# ตรวจสอบ env vars
docker exec apiservice env | grep DB_

# Test connection จาก container
docker exec apiservice python -c "import psycopg2; psycopg2.connect(host='postgres', user='postgres', password='<pass>')"

Supabase services unhealthy

# ตรวจสอบทุก container
docker ps --filter "name=sdp-"

# Restart ตามลำดับ dependency
docker restart sdp-supabase-db
sleep 10
docker restart sdp-supabase-analytics
sleep 10
docker restart sdp-supabase-kong
docker restart sdp-supabase-auth
docker restart sdp-supabase-rest

Airflow worker ไม่ pick tasks

# ตรวจสอบ Redis connectivity
docker exec airflow-worker redis-cli -h redis ping

# ตรวจสอบ worker
docker logs airflow-worker --tail 50

# Restart worker
docker restart airflow-worker

Nginx 502 Bad Gateway

# ตรวจสอบว่า backend container ทำงานอยู่
docker ps | grep <service-name>

# ตรวจสอบ logs
docker logs nginx-proxy-manager --tail 50
docker logs <service-name> --tail 50

# ตรวจสอบ network
docker network inspect shared_data_network | grep -A5 <service-name>

MinIO ไม่ accessible

# Health check
curl -f http://localhost:9000/minio/health/live

# Logs
docker logs minio --tail 50

# Disk space
df -h

---

Monitoring

Dozzle (Docker Logs)

URL: https://ai.sriphat.com/dozzle

Dozzle monitor ทั้ง:

• Main server (local)
• Remote agent: 192.168.100.9:7007

Container Health

# ดู health status
docker ps --format "table {{.Names}}\t{{.Status}}\t{{.Health}}"

# ดู resource usage
docker stats --format "table {{.Name}}\t{{.CPUPerc}}\t{{.MemUsage}}"

---

Access Points Summary

Service	URL	Credentials
Nginx Proxy	http://192.168.100.9:8020	—
Keycloak Admin	https://ai.sriphat.com/keycloak/admin	KEYCLOAK_ADMIN
Supabase Studio	https://ai.sriphat.com/supabase	DB credentials
Supabase API	https://ai.sriphat.com/supabase-api	ANON_KEY / SERVICE_ROLE_KEY
API Service	https://ai.sriphat.com/apiservice	Admin credentials
Airflow	https://ai.sriphat.com/airflow	Airflow admin
Superset	https://ai.sriphat.com/superset	SUPERSET_ADMIN_*
MinIO Console	https://ai.sriphat.com/minio-console	MINIO_ROOT_*
MinIO API	https://ai.sriphat.com/minio	S3 credentials
Dozzle	https://ai.sriphat.com/dozzle	— (no auth default)

---

Related

• 00-Project-Overview
• 01-Infrastructure
• 07-Security-Strategy