Supabase Realtime Container Orchestration —

Supabase Realtime Orchestration

Supabase Realtime PostgreSQL CDC WebSocket Container Orchestration Kubernetes Docker Scaling Elixir Phoenix Presence Broadcast Channel

Feature	Supabase	Firebase	Appwrite	Nhost
Database	PostgreSQL	Firestore	MariaDB	PostgreSQL
Realtime	CDC + WebSocket	Built-in	WebSocket	Hasura
Open Source	ใช่	ไม่	ใช่	ใช่
Self-host	Docker/K8s	ไม่ได้	Docker	Docker

Supabase Realtime Setup

=== Supabase Realtime ===

JavaScript Client

import { createClient } from '@supabase/supabase-js'

const supabase = createClient(SUPABASE_URL, SUPABASE_ANON_KEY)

Subscribe to table changes

const channel = supabase

.channel('messages')

.on('postgres_changes',

{ event: '*', schema: 'public', table: 'messages' },

(payload) => {

console.log('Change:', payload)

if (payload.eventType === 'INSERT') {

addMessage(payload.new)

}

}

)

.subscribe()

Presence — Who's online

const presenceChannel = supabase.channel('online-users')

presenceChannel

.on('presence', { event: 'sync' }, () => {

const state = presenceChannel.presenceState()

console.log('Online:', Object.keys(state).length)

})

.subscribe(async (status) => {

if (status === 'SUBSCRIBED') {

await presenceChannel.track({

user_id: currentUser.id,

username: currentUser.name,

online_at: new Date().toISOString(),

})

}

})

Broadcast — Send messages to channel

const broadcastChannel = supabase.channel('chat-room')

broadcastChannel

.on('broadcast', { event: 'message' }, (payload) => {

console.log('Message:', payload)

})

.subscribe()

broadcastChannel.send({

type: 'broadcast',

event: 'message',

payload: { text: 'Hello!', user: 'Alice' },

})

from dataclasses import dataclass

from typing import List

@dataclass

class RealtimeChannel:

name: str

type: str

subscribers: int

events_per_sec: float

latency_ms: int

channels = [

RealtimeChannel("messages", "postgres_changes", 250, 45, 12),

RealtimeChannel("online-users", "presence", 500, 5, 8),

RealtimeChannel("chat-room-1", "broadcast", 80, 120, 5),

RealtimeChannel("notifications", "postgres_changes", 1000, 20, 15),

RealtimeChannel("live-dashboard", "postgres_changes", 50, 200, 10),

]

print("=== Realtime Channels ===")

for c in channels:

print(f" [{c.type}] {c.name}")

print(f" Subscribers: {c.subscribers} | Events/s: {c.events_per_sec} | Latency: {c.latency_ms}ms")

Docker และ Kubernetes

=== Container Orchestration ===

Docker Compose — Self-hosted Supabase

services:

db:

image: supabase/postgres:15.1.0

volumes: [./volumes/db:/var/lib/postgresql/data]

environment:

POSTGRES_PASSWORD:

realtime:

image: supabase/realtime:v2.25.0

depends_on: [db]

environment:

DB_HOST: db

DB_PORT: 5432

DB_USER: supabase_admin

DB_PASSWORD:

DB_NAME: postgres

PORT: 4000

SECRET_KEY_BASE:

kong:

image: kong:2.8.1

environment:

KONG_DECLARATIVE_CONFIG: /var/lib/kong/kong.yml

ports: ["8000:8000", "8443:8443"]

studio:

image: supabase/studio:latest

environment:

SUPABASE_URL: http://kong:8000

STUDIO_PG_META_URL: http://meta:8080

Kubernetes Deployment

apiVersion: apps/v1

kind: Deployment

metadata:

name: supabase-realtime

spec:

replicas: 3

selector:

matchLabels: { app: supabase-realtime }

template:

spec:

containers:

• name: realtime

image: supabase/realtime:v2.25.0

ports: [{ containerPort: 4000 }]

resources:

requests: { cpu: "500m", memory: "512Mi" }

limits: { cpu: "2000m", memory: "2Gi" }

env:

• name: DB_HOST

valueFrom:

secretKeyRef: { name: supabase-db, key: host }

@dataclass

class K8sResource:

name: str

kind: str

replicas: int

cpu: str

memory: str

status: str

resources = [

K8sResource("supabase-db", "StatefulSet", 1, "2000m", "4Gi", "Running"),

K8sResource("supabase-realtime", "Deployment", 3, "500m", "512Mi", "Running"),

K8sResource("supabase-kong", "Deployment", 2, "250m", "256Mi", "Running"),

K8sResource("supabase-auth", "Deployment", 2, "250m", "256Mi", "Running"),

K8sResource("supabase-storage", "Deployment", 2, "250m", "512Mi", "Running"),

K8sResource("supabase-studio", "Deployment", 1, "250m", "256Mi", "Running"),

]

print("\n=== Kubernetes Resources ===")

for r in resources:

print(f" [{r.status}] {r.name} ({r.kind})")

print(f" Replicas: {r.replicas} | CPU: {r.cpu} | Memory: {r.memory}")

Scaling และ Monitoring

# === Production Scaling ===

# HPA — Horizontal Pod Autoscaler
# apiVersion: autoscaling/v2
# kind: HorizontalPodAutoscaler
# metadata:
#   name: supabase-realtime-hpa
# spec:
#   scaleTargetRef:
#     apiVersion: apps/v1
#     kind: Deployment
#     name: supabase-realtime
#   minReplicas: 2
#   maxReplicas: 10
#   metrics:
#     - type: Resource
#       resource:
#         name: cpu
#         target: { type: Utilization, averageUtilization: 70 }
#     - type: Pods
#       pods:
#         metric: { name: websocket_connections }
#         target: { type: AverageValue, averageValue: "500" }

# Prometheus Metrics
# supabase_realtime_connected_clients
# supabase_realtime_channels_count
# supabase_realtime_messages_per_second
# supabase_realtime_latency_milliseconds

monitoring = {
    "Connected Clients": "2,500",
    "Active Channels": "150",
    "Messages/sec": "450",
    "Avg Latency": "12ms",
    "P99 Latency": "45ms",
    "Realtime Pods": "5/10 (HPA)",
    "DB Connections": "85/200",
    "WebSocket Memory": "1.2GB / 2GB",
    "CDC Lag": "< 100ms",
}

print("Supabase Realtime Dashboard:")
for k, v in monitoring.items():
    print(f"  {k}: {v}")

# Cost Comparison
costs = {
    "Supabase Cloud (Pro)": "$25/mo + usage",
    "Self-hosted (K8s 3-node)": "$150/mo (cloud VMs)",
    "Self-hosted (Home Lab)": "$0/mo (electricity only)",
    "Firebase (Blaze)": "Pay-as-you-go (~$50-200/mo)",
}

print(f"\n\nCost Comparison:")
for setup, cost in costs.items():
    print(f"  [{setup}]: {cost}")

เคล็ดลับ

• RLS: เปิด Row Level Security ทุกตาราง ป้องกัน Data Leak
• Channel: Subscribe เฉพาะตาราง/คอลัมน์ที่ต้องการ ลด Load
• HPA: ตั้ง HPA สำหรับ Realtime Server Scale ตาม Connection
• CDC: Monitor CDC Lag ถ้าสูงเกิน 1s ต้องตรวจสอบ
• Backup: Backup PostgreSQL เป็นประจำ ใช้ pg_dump หรือ Litestream

การดูแลระบบในสภาพแวดล้อม Production

การบริหารจัดการระบบ Production ที่ดีต้องมี Monitoring ครอบคลุม ใช้เครื่องมืออย่าง Prometheus + Grafana สำหรับ Metrics Collection และ Dashboard หรือ ELK Stack สำหรับ Log Management ตั้ง Alert ให้แจ้งเตือนเมื่อ CPU เกิน 80% RAM ใกล้เต็ม หรือ Disk Usage สูง

Backup Strategy ต้องวางแผนให้ดี ใช้หลัก 3-2-1 คือ มี Backup อย่างน้อย 3 ชุด เก็บใน Storage 2 ประเภทต่างกัน และ 1 ชุดต้องอยู่ Off-site ทดสอบ Restore Backup เป็นประจำ อย่างน้อยเดือนละครั้ง เพราะ Backup ที่ Restore ไม่ได้ก็เหมือนไม่มี Backup

เรื่อง Security Hardening ต้องทำตั้งแต่เริ่มต้น ปิด Port ที่ไม่จำเป็น ใช้ SSH Key แทน Password ตั้ง Fail2ban ป้องกัน Brute Force อัพเดท Security Patch สม่ำเสมอ และทำ Vulnerability Scanning อย่างน้อยเดือนละครั้ง ใช้หลัก Principle of Least Privilege ให้สิทธิ์น้อยที่สุดที่จำเป็น

สรุป

Supabase Realtime PostgreSQL CDC WebSocket Container Kubernetes Docker Presence Broadcast Channel HPA Scaling RLS Monitoring Prometheus Self-hosted Production