Betteruptime Testing Strategy QA

BetterUptime QA

BetterUptime Testing Strategy QA Uptime Monitoring Incident Management Status Page On-call Alert SLA SLO SLI Chaos Testing Game Day HTTP TCP DNS Heartbeat

Monitor Setup

# === BetterUptime Configuration ===

# Monitor Types:
# 1. HTTP Monitor — ตรวจ Status Code + Response Time
# 2. TCP Monitor — ตรวจ Port เปิด
# 3. DNS Monitor — ตรวจ DNS Resolution
# 4. Heartbeat — ตรวจ Cron Job ทำงาน
# 5. Keyword — ตรวจ Keyword ใน Page

# API — Create Monitor
# curl -X POST https://betteruptime.com/api/v2/monitors \
#   -H "Authorization: Bearer TOKEN" \
#   -H "Content-Type: application/json" \
#   -d '{
#     "monitor_type": "status",
#     "url": "https://api.example.com/health",
#     "pronounceable_name": "API Health Check",
#     "check_frequency": 30,
#     "request_timeout": 15,
#     "confirmation_period": 3,
#     "regions": ["us", "eu", "ap"],
#     "expected_status_codes": [200],
#     "domain_expiration": 30,
#     "ssl_expiration": 14,
#     "follow_redirects": true,
#     "policy_id": "escalation-policy-id"
#   }'

# Terraform — Infrastructure as Code
# resource "betteruptime_monitor" "api_health" {
#   monitor_type     = "status"
#   url              = "https://api.example.com/health"
#   check_frequency  = 30
#   request_timeout  = 15
#   regions          = ["us", "eu", "ap"]
#   policy_id        = betteruptime_policy.default.id
# }

from dataclasses import dataclass
from typing import List

@dataclass
class Monitor:
    name: str
    type: str
    url: str
    frequency_sec: int
    uptime_30d: float
    avg_response_ms: int
    status: str

monitors = [
    Monitor("API Health", "HTTP", "https://api.example.com/health", 30, 99.95, 120, "Up"),
    Monitor("Web Frontend", "HTTP", "https://www.example.com", 60, 99.99, 250, "Up"),
    Monitor("Database", "TCP", "db.example.com:5432", 30, 99.90, 15, "Up"),
    Monitor("DNS Resolution", "DNS", "example.com", 60, 100.0, 25, "Up"),
    Monitor("Cron Backup", "Heartbeat", "Expected every 6h", 21600, 99.8, 0, "Up"),
    Monitor("SSL Certificate", "HTTP", "https://example.com", 86400, 100.0, 0, "Valid 45d"),
]

print("=== Monitors ===")
for m in monitors:
    print(f"  [{m.status}] {m.name} ({m.type})")
    print(f"    URL: {m.url}")
    print(f"    Frequency: {m.frequency_sec}s | Uptime: {m.uptime_30d}% | Response: {m.avg_response_ms}ms")

Testing Strategy

# === QA Testing for Monitoring ===

@dataclass
class TestCase:
    id: str
    category: str
    description: str
    method: str
    expected: str
    status: str

test_cases = [
    TestCase("TC-01", "Detection", "Monitor ตรวจจับ HTTP 500", "หยุด Service แล้วดู Alert", "Alert ภายใน 3 นาที", "Pass"),
    TestCase("TC-02", "Detection", "Monitor ตรวจจับ Timeout", "ตั้ง Delay > Timeout", "Alert ภายใน 3 นาที", "Pass"),
    TestCase("TC-03", "Alert", "Alert ส่งถึง Slack", "Trigger Downtime", "Slack message received", "Pass"),
    TestCase("TC-04", "Alert", "Alert ส่งถึง Email", "Trigger Downtime", "Email received < 1 min", "Pass"),
    TestCase("TC-05", "Alert", "SMS Alert ทำงาน", "Trigger Downtime", "SMS received < 2 min", "Pass"),
    TestCase("TC-06", "Escalation", "Escalate ถ้าไม่ Acknowledge", "ไม่ตอบ Alert 10 นาที", "Escalate to Level 2", "Pass"),
    TestCase("TC-07", "Status Page", "Status Page อัพเดทอัตโนมัติ", "Trigger Downtime", "Status = Degraded", "Pass"),
    TestCase("TC-08", "Recovery", "ตรวจจับ Recovery", "Start Service กลับ", "Alert resolved < 2 min", "Pass"),
    TestCase("TC-09", "SSL", "SSL Expiry Alert", "Certificate < 14 days", "Alert sent", "Pass"),
    TestCase("TC-10", "Heartbeat", "Cron Job Missing", "ไม่ส่ง Heartbeat 6h", "Alert sent", "Pass"),
]

print("\n=== QA Test Cases ===")
passed = sum(1 for t in test_cases if t.status == "Pass")
print(f"  Results: {passed}/{len(test_cases)} Passed\n")
for t in test_cases:
    print(f"  [{t.status}] {t.id} — {t.category}: {t.description}")
    print(f"    Method: {t.method}")
    print(f"    Expected: {t.expected}")

# Chaos Testing Schedule
chaos_tests = [
    "Weekly: ปิด Non-critical Service ดู Alert + Recovery",
    "Monthly: Game Day ซ้อม Full Incident Response",
    "Quarterly: Failover Test ย้าย Region ดู Monitoring",
    "On Change: ทดสอบ Monitor ใหม่ทุกครั้งที่เพิ่ม/แก้ไข",
]

print(f"\n\nChaos Testing Schedule:")
for i, c in enumerate(chaos_tests, 1):
    print(f"  {i}. {c}")

SLA Management

# === SLA/SLO/SLI ===

# SLI (Service Level Indicator) — ตัววัด
# - Availability: % time service is up
# - Latency: p50, p95, p99 response time
# - Error Rate: % of 5xx responses
# - Throughput: requests per second

# SLO (Service Level Objective) — เป้าหมาย
# - Availability: 99.9%
# - Latency p99: < 500ms
# - Error Rate: < 0.1%

# SLA (Service Level Agreement) — สัญญา
# - 99.9% uptime = max 8h 45m downtime/year
# - Credit if SLA breached

@dataclass
class SLATarget:
    service: str
    sla_pct: float
    max_downtime_month: str
    current_uptime: float
    error_budget_remaining: str
    status: str

sla_targets = [
    SLATarget("API Gateway", 99.95, "21 min", 99.97, "65%", "Healthy"),
    SLATarget("Web App", 99.9, "43 min", 99.95, "80%", "Healthy"),
    SLATarget("Database", 99.99, "4 min", 99.995, "90%", "Healthy"),
    SLATarget("CDN", 99.9, "43 min", 99.85, "15%", "At Risk"),
    SLATarget("Auth Service", 99.95, "21 min", 99.93, "35%", "Warning"),
]

print("SLA Dashboard:")
for s in sla_targets:
    emoji = "OK" if s.status == "Healthy" else s.status.upper()
    print(f"  [{emoji}] {s.service}")
    print(f"    SLA: {s.sla_pct}% | Current: {s.current_uptime}%")
    print(f"    Max Down: {s.max_downtime_month}/mo | Budget: {s.error_budget_remaining}")

# Uptime Calculation
print(f"\n\nUptime Reference:")
uptimes = {
    "99%": "7h 18m/month, 3.65 days/year",
    "99.9%": "43m 50s/month, 8h 46m/year",
    "99.95%": "21m 55s/month, 4h 23m/year",
    "99.99%": "4m 23s/month, 52m 36s/year",
    "99.999%": "26s/month, 5m 16s/year",
}
for pct, downtime in uptimes.items():
    print(f"  {pct}: {downtime}")

เคล็ดลับ

• Multi-region: Monitor จากหลาย Region ป้องกัน False Positive
• Confirmation: ตั้ง Confirmation Period ก่อน Alert
• Game Day: ซ้อม Incident Response ทุกเดือน
• Error Budget: ติดตาม Error Budget ไม่ให้หมด
• Status Page: สร้าง Status Page สำหรับลูกค้า

การนำความรู้ไปประยุกต์ใช้งานจริง

แหล่งเรียนรู้ที่แนะนำ ได้แก่ Official Documentation ที่อัพเดทล่าสุดเสมอ Online Course จาก Coursera Udemy edX ช่อง YouTube คุณภาพทั้งไทยและอังกฤษ และ Community อย่าง Discord Reddit Stack Overflow ที่ช่วยแลกเปลี่ยนประสบการณ์กับนักพัฒนาทั่วโลก

เปรียบเทียบข้อดีและข้อเสีย

จากตารางเปรียบเทียบจะเห็นว่าข้อดีมีมากกว่าข้อเสียอย่างชัดเจน โดยเฉพาะในแง่ของประสิทธิภาพและความสามารถในการ Scale สำหรับข้อเสียส่วนใหญ่สามารถแก้ไขได้ด้วยการเรียนรู้อย่างเป็นระบบและวางแผนทรัพยากรให้เหมาะสม

สรุป

BetterUptime Testing Strategy QA Uptime Monitoring Incident Status Page On-call SLA SLO SLI Error Budget Chaos Testing Game Day Alert Escalation Multi-region