ai
CircleCI Orbs High Availability HA Setup
CircleCI Orbs High Availability HA Setup คืออะไร
CircleCI เป็น CI/CD platform ยอดนิยมที่รองรับ automated build, test และ deploy CircleCI Orbs คือ reusable configuration packages ที่ช่วยลด boilerplate config และแชร์ best practices ข้าม projects High Availability (HA) Setup คือการออกแบบระบบให้มี uptime สูงสุด (99.9%+) โดยไม่มี single point of failure การรวมแนวคิดเหล่านี้ช่วยให้ CI/CD pipeline มีความเสถียร deploy applications ที่มี HA architecture และใช้ Orbs จัดการ infrastructure อย่างมีประสิทธิภาพ
CircleCI Orbs พื้นฐาน
# orbs_basics.py — CircleCI Orbs fundamentals
import json
class CircleCIOrbs:
CONCEPTS = {
"orb": {
"name": "Orb",
"description": "Reusable package ของ CircleCI config (commands, jobs, executors)",
"registry": "circleci.com/developer/orbs (public registry)",
},
"command": {
"name": "Command",
"description": "Reusable step sequence ที่ใช้ใน jobs",
"example": "aws-cli/setup — ติดตั้งและ configure AWS CLI",
},
"job": {
"name": "Job",
"description": "Pre-defined job ที่พร้อมใช้ใน workflow",
"example": "docker/publish — build + push Docker image",
},
"executor": {
"name": "Executor",
"description": "Pre-configured execution environment",
"example": "node/default — Node.js environment พร้อมใช้",
},
}
POPULAR_ORBS = {
"aws-cli": {"name": "aws-cli", "use": "AWS CLI setup and commands", "installs": "2B+"},
"docker": {"name": "docker", "use": "Build, tag, push Docker images", "installs": "500M+"},
"kubernetes": {"name": "kubernetes", "use": "Deploy to Kubernetes clusters", "installs": "100M+"},
"slack": {"name": "slack", "use": "Send notifications to Slack", "installs": "300M+"},
"terraform": {"name": "terraform", "use": "Terraform plan and apply", "installs": "200M+"},
"node": {"name": "node", "use": "Node.js environment and caching", "installs": "400M+"},
}
BASIC_CONFIG = """
# .circleci/config.yml — Basic Orbs usage
version: 2.1
orbs:
aws-cli: circleci/aws-cli@4.1
docker: circleci/docker@2.6
slack: circleci/slack@4.13
jobs:
build-and-test:
docker:
- image: cimg/python:3.11
steps:
- checkout
- run: pip install -r requirements.txt
- run: pytest tests/ -v
- slack/notify:
event: fail
template: basic_fail_1
deploy:
docker:
- image: cimg/python:3.11
steps:
- checkout
- aws-cli/setup
- docker/build:
image: myapp
tag:
- docker/push:
image: myapp
tag:
workflows:
build-deploy:
jobs:
- build-and-test
- deploy:
requires: [build-and-test]
filters:
branches:
only: main
"""
def show_concepts(self):
print("=== Orbs Concepts ===\n")
for key, concept in self.CONCEPTS.items():
print(f"[{concept['name']}] {concept['description']}")
def show_popular(self):
print(f"\n=== Popular Orbs ===")
for key, orb in self.POPULAR_ORBS.items():
print(f" [{orb['name']}] {orb['use']}")
def show_config(self):
print(f"\n=== Basic Config ===")
print(self.BASIC_CONFIG[:500])
orbs = CircleCIOrbs()
orbs.show_concepts()
orbs.show_popular()
orbs.show_config()HA Architecture Patterns
# ha_patterns.py — High Availability architecture patterns
import json
class HAPatterns:
PATTERNS = {
"active_active": {
"name": "Active-Active",
"description": "ทุก instance รับ traffic พร้อมกัน load balanced",
"use": "Web servers, API gateways, stateless services",
"uptime": "99.99%+",
"complexity": "Medium-High",
},
"active_passive": {
"name": "Active-Passive (Failover)",
"description": "Primary instance ทำงาน, secondary standby พร้อม takeover",
"use": "Databases, stateful services",
"uptime": "99.9%+",
"complexity": "Medium",
},
"multi_az": {
"name": "Multi-AZ (Availability Zone)",
"description": "Deploy ข้าม AZ ใน region เดียว",
"use": "Cloud workloads (AWS, GCP, Azure)",
"uptime": "99.99%",
"complexity": "Medium",
},
"multi_region": {
"name": "Multi-Region",
"description": "Deploy ข้าม regions (geographic redundancy)",
"use": "Global applications, disaster recovery",
"uptime": "99.999%",
"complexity": "High",
},
}
HA_COMPONENTS = {
"load_balancer": "ALB/NLB, Nginx, HAProxy — กระจาย traffic",
"auto_scaling": "EC2 ASG, K8s HPA — scale ตาม load",
"database_ha": "RDS Multi-AZ, Aurora, PostgreSQL streaming replication",
"cache_ha": "Redis Sentinel/Cluster, ElastiCache Multi-AZ",
"dns_failover": "Route53 health checks, Cloudflare load balancing",
"monitoring": "CloudWatch, Prometheus, PagerDuty — detect failures",
}
def show_patterns(self):
print("=== HA Patterns ===\n")
for key, pattern in self.PATTERNS.items():
print(f"[{pattern['name']}] {pattern['description']}")
print(f" Use: {pattern['use']} | Uptime: {pattern['uptime']}")
print()
def show_components(self):
print("=== HA Components ===")
for comp, desc in self.HA_COMPONENTS.items():
print(f" [{comp}] {desc}")
ha = HAPatterns()
ha.show_patterns()
ha.show_components()CircleCI HA Deployment Pipeline
# ha_pipeline.py — HA deployment with CircleCI
import json
class HAPipeline:
CONFIG = """
# .circleci/config.yml — HA Deployment Pipeline
version: 2.1
orbs:
aws-cli: circleci/aws-cli@4.1
aws-ecs: circleci/aws-ecs@4.0
kubernetes: circleci/kubernetes@1.3
slack: circleci/slack@4.13
rollbar: circleci/rollbar@1.0
executors:
deploy-executor:
docker:
- image: cimg/python:3.11
environment:
AWS_REGION: ap-southeast-1
jobs:
test:
docker:
- image: cimg/python:3.11
- image: cimg/postgres:15.0
steps:
- checkout
- restore_cache:
keys: [deps-{{ checksum "requirements.txt" }}]
- run: pip install -r requirements.txt
- save_cache:
key: deps-{{ checksum "requirements.txt" }}
paths: [~/.cache/pip]
- run: pytest tests/ -v --junitxml=results.xml
- store_test_results:
path: results.xml
build-image:
executor: deploy-executor
steps:
- checkout
- setup_remote_docker:
version: 20.10.24
- run:
name: Build and push to ECR
command: |
aws ecr get-login-password | docker login --username AWS --password-stdin $ECR_URL
docker build -t $ECR_URL/myapp: .
docker push $ECR_URL/myapp:
deploy-blue-green:
executor: deploy-executor
steps:
- aws-cli/setup
- run:
name: Blue-Green Deploy to ECS
command: |
# Update task definition with new image
aws ecs update-service \\
--cluster production \\
--service myapp-green \\
--task-definition myapp: \\
--desired-count 3
# Wait for green deployment stable
aws ecs wait services-stable \\
--cluster production \\
--services myapp-green
# Switch ALB target group
aws elbv2 modify-listener \\
--listener-arn $LISTENER_ARN \\
--default-actions Type=forward, TargetGroupArn=$GREEN_TG_ARN
- slack/notify:
event: pass
template: success_tagged_deploy_1
health-check:
executor: deploy-executor
steps:
- run:
name: Post-deploy health check
command: |
for i in $(seq 1 10); do
STATUS=$(curl -s -o /dev/null -w "%{http_code}" https://api.example.com/health)
if [ "$STATUS" = "200" ]; then
echo "Health check passed"
exit 0
fi
sleep 10
done
echo "Health check failed"
exit 1
workflows:
ha-deploy:
jobs:
- test
- build-image:
requires: [test]
filters:
branches:
only: main
- deploy-blue-green:
requires: [build-image]
- health-check:
requires: [deploy-blue-green]
"""
def show_config(self):
print("=== HA Pipeline Config ===")
print(self.CONFIG[:800])
def deployment_strategies(self):
print(f"\n=== Deployment Strategies for HA ===")
strategies = [
{"name": "Blue-Green", "downtime": "Zero", "rollback": "Instant (switch back)", "complexity": "Medium"},
{"name": "Canary", "downtime": "Zero", "rollback": "Fast (route back)", "complexity": "High"},
{"name": "Rolling Update", "downtime": "Near-zero", "rollback": "Slow (redeploy)", "complexity": "Low"},
{"name": "A/B Testing", "downtime": "Zero", "rollback": "Instant", "complexity": "High"},
]
for s in strategies:
print(f" [{s['name']}] Downtime: {s['downtime']} | Rollback: {s['rollback']}")
pipeline = HAPipeline()
pipeline.show_config()
pipeline.deployment_strategies()Infrastructure as Code for HA
# iac_ha.py — Terraform HA infrastructure
import json
class IACHA:
TERRAFORM = """
# main.tf — HA infrastructure with Terraform
provider "aws" {
region = "ap-southeast-1"
}
# VPC with Multi-AZ
module "vpc" {
source = "terraform-aws-modules/vpc/aws"
version = "5.0"
name = "ha-vpc"
cidr = "10.0.0.0/16"
azs = ["ap-southeast-1a", "ap-southeast-1b", "ap-southeast-1c"]
private_subnets = ["10.0.1.0/24", "10.0.2.0/24", "10.0.3.0/24"]
public_subnets = ["10.0.101.0/24", "10.0.102.0/24", "10.0.103.0/24"]
enable_nat_gateway = true
single_nat_gateway = false # NAT per AZ for HA
}
# ALB (Application Load Balancer)
resource "aws_lb" "main" {
name = "ha-alb"
internal = false
load_balancer_type = "application"
subnets = module.vpc.public_subnets
enable_deletion_protection = true
enable_cross_zone_load_balancing = true
}
# ECS Cluster with Fargate
resource "aws_ecs_cluster" "main" {
name = "ha-cluster"
setting {
name = "containerInsights"
value = "enabled"
}
}
# ECS Service (Multi-AZ)
resource "aws_ecs_service" "app" {
name = "myapp"
cluster = aws_ecs_cluster.main.id
task_definition = aws_ecs_task_definition.app.arn
desired_count = 3 # Spread across 3 AZs
launch_type = "FARGATE"
network_configuration {
subnets = module.vpc.private_subnets
security_groups = [aws_security_group.app.id]
}
load_balancer {
target_group_arn = aws_lb_target_group.app.arn
container_name = "myapp"
container_port = 8080
}
}
# RDS Multi-AZ
resource "aws_db_instance" "main" {
identifier = "ha-postgres"
engine = "postgres"
engine_version = "15"
instance_class = "db.r6g.large"
multi_az = true
storage_encrypted = true
backup_retention_period = 7
db_subnet_group_name = aws_db_subnet_group.main.name
}
"""
def show_terraform(self):
print("=== Terraform HA Config ===")
print(self.TERRAFORM[:700])
def ha_checklist(self):
print(f"\n=== HA Checklist ===")
checks = [
("Multi-AZ deployment (3 AZs)", True),
("Load balancer health checks", True),
("Auto-scaling configured", True),
("Database Multi-AZ failover", True),
("Redis/Cache HA (Sentinel/Cluster)", True),
("DNS failover configured", True),
("Monitoring + alerting active", True),
("Disaster recovery plan documented", True),
("Runbook for failover scenarios", True),
]
for name, status in checks:
icon = "PASS" if status else "TODO"
print(f" [{icon:>4}] {name}")
iac = IACHA()
iac.show_terraform()
iac.ha_checklist()Monitoring & Failover
# monitoring.py — HA monitoring and failover
import json
import random
class HAMonitoring:
def cluster_status(self):
print("=== HA Cluster Status ===\n")
services = [
{"name": "API Gateway (ALB)", "status": "healthy", "instances": "3/3", "az": "3 AZs"},
{"name": "App Service (ECS)", "status": "healthy", "instances": "6/6", "az": "3 AZs"},
{"name": "Database (RDS)", "status": "healthy", "instances": "Primary + Standby", "az": "Multi-AZ"},
{"name": "Cache (Redis)", "status": "healthy", "instances": "3 nodes", "az": "3 AZs"},
{"name": "CDN (CloudFront)", "status": "healthy", "instances": "Global", "az": "Edge locations"},
]
for s in services:
print(f" [{s['status']:>7}] {s['name']:<25} Instances: {s['instances']}")
def uptime_metrics(self):
print(f"\n=== Uptime Metrics ===")
metrics = {
"Current uptime": f"{random.uniform(99.95, 99.99):.3f}%",
"Last 30 days": f"{random.uniform(99.90, 99.99):.3f}%",
"MTTR": f"{random.randint(1, 10)} minutes",
"MTBF": f"{random.randint(30, 180)} days",
"Incidents (30d)": f"{random.randint(0, 3)}",
"RTO target": "< 5 minutes",
"RPO target": "< 1 minute",
}
for m, v in metrics.items():
print(f" {m}: {v}")
def failover_test(self):
print(f"\n=== Failover Test Results ===")
tests = [
{"scenario": "Single instance failure", "recovery": f"{random.randint(5, 30)}s", "result": "PASS"},
{"scenario": "AZ failure", "recovery": f"{random.randint(30, 120)}s", "result": "PASS"},
{"scenario": "Database failover", "recovery": f"{random.randint(30, 90)}s", "result": "PASS"},
{"scenario": "Region failover", "recovery": f"{random.randint(120, 600)}s", "result": "PASS"},
]
for t in tests:
print(f" [{t['result']:>4}] {t['scenario']:<30} Recovery: {t['recovery']}")
mon = HAMonitoring()
mon.cluster_status()
mon.uptime_metrics()
mon.failover_test()FAQ - คำถามที่พบบ่อย
Q: CircleCI Orbs กับ GitHub Actions reusable workflows อันไหนดี?
A: CircleCI Orbs: mature ecosystem, versioned packages, central registry, reusable across orgs GitHub Actions: larger marketplace, tighter GitHub integration, easier to start ใช้ CircleCI: ถ้าใช้ CircleCI อยู่แล้ว, ต้องการ versioned reusable configs ใช้ GitHub Actions: ถ้าใช้ GitHub, ต้องการ simplicity
เนื้อหาเกี่ยวข้อง — Network Security สำหรับ Home Office
Q: HA ต้องใช้ minimum กี่ AZs?
แนะนำเพิ่มเติม — ระบบเทรดของ iCafeForex
A: 2 AZs: minimum สำหรับ HA (ถ้า 1 AZ ล่ม ยังมี 1 AZ) 3 AZs: แนะนำ (quorum-based systems เช่น etcd, Consul ต้อง 3+) Cost: Multi-AZ เพิ่มค่าใช้จ่าย ~50-100% แต่คุ้มสำหรับ production
เนื้อหาเกี่ยวข้อง — ทำความเข้าใจ BigQuery Scheduled Query Progressive Delivery
Q: Blue-Green กับ Canary อันไหนดี?
A: Blue-Green: ง่ายกว่า, switch 100% traffic ทีเดียว, rollback instant Canary: ค่อยๆ route traffic (1% → 10% → 50% → 100%), ลด blast radius ใช้ Blue-Green: simple apps, small team, ต้องการ fast rollback ใช้ Canary: large scale, ต้อง validate กับ real traffic ก่อน full deploy
แนะนำเพิ่มเติม — SiamCafeBook
เนื้อหาเกี่ยวข้อง — บทความที่เกี่ยวข้อง: ups แบบ true online คือ
Q: HA มีค่าใช้จ่ายเท่าไหร่?
A: Multi-AZ: เพิ่ม ~50-100% (double instances + cross-AZ traffic) Multi-Region: เพิ่ม ~200-400% (full stack per region) ประหยัด: ใช้ spot instances สำหรับ non-critical, reserved instances สำหรับ baseline ROI: downtime 1 ชั่วโมง อาจเสียหาย $10K-$1M+ → HA คุ้มค่าสำหรับ business-critical
เนื้อหาเกี่ยวข้อง — บทความที่เกี่ยวข้อง: Ollama Local LLM Clean Architecture
