Cloud Cost Anomaly Sentinel: Autonomous FinOps Waste Hunting
Automatically identifying and remediating orphan EBS volumes, unattached Elastic IPs, and idle development RDS instances using AWS Config and EventBridge.
Quick Navigation
The 3 Pillars of Telemetry: Metrics, Logs, Traces & Automated Self-Healing
Monitoring tells you when a system is failing; Observability allows you to understand why a complex, distributed failure mode is occurring:
Interactive Architecture Diagram (Use controls to zoom & pan)
Amazon CloudWatch, OpenSearch, AWS X-Ray Distributed Tracing နှင့် Self-Healing EventBridge
ရှုပ်ထွေးသော Cloud စနစ်များတွင် ပြဿနာဖြစ်ပွားပါက မည်သည့်နေရာတွင် မည်သည့်အတွက် ချို့ယွင်းသွားသည်ကို သိရှိနိုင်သော Observability မဏ္ဍိုင် ၃ ရပ် (Metrics, Logs, Traces) နှင့် စနစ်ချို့ယွင်းချက်ကို အလိုအလျောက် ပြင်ဆင်ပေးသော Self-Healing Architecture။
Metrics, Logs နှင့် Distributed Traces (The 3 Pillars of Observability) ကို စနစ်တကျ ချိတ်ဆက်နိုင်ခြင်း။
AWS X-Ray သုံး၍ Microservices များအကြား Request ပျောက်ဆုံးမှုနှင့် Bottleneck Latency များကို Pinpoint ရှာဖွေနိုင်ခြင်း။
CloudWatch Metric Alarms နှင့် Composite Alarms များ ဖန်တီး၍ Alarm Fatigue (အချက်ပေးသံများလွန်း၍ မေ့လျော့ခြင်း) ကို လျှော့ချနိုင်ခြင်း။
Amazon EventBridge + Lambda ဖြင့် EC2/RDS ပျက်စီးမှုများကို လူမပါဘဲ အလိုအလျောက် ကုစားပေးသော Self-Healing စနစ် တည်ဆောက်နိုင်ခြင်း။
စနစ်ကြီးမားလာသည်နှင့်အမျှ "ကျွနု်ပ်တို့ Server ပျက်နေသလား" ဟု ကြည့်ရုံဖြင့် မလုံလောက်တော့ဘဲ "User များထဲမှ 0.1% သည် ဘာကြောင့် 500 Error ကြုံနေရသလဲ" ကို ချက်ချင်း ဖြေရှင်းနိုင်ရပါမည်။
Observability သည် စနစ်၏ အခြေအနေကို အချိန်နှင့်တပြေးညီ မြင်သာစေပြီး စီးပွားရေးဆုံးရှုံးမှုကို အနည်းဆုံးဖြစ်အောင် ကာကွယ်ပေးပါသည်။
User Request တစ်ခုသည် API Gateway -> Lambda -> DynamoDB -> 3rd-party API သို့ သွားသော ခရီးစဉ်တစ်ခုလုံးကို ID တစ်ခုတည်းဖြင့် ခြေရာခံခြင်း။
Service Level Indicator (လက်ရှိ အခြေအနေ), Service Level Objective (အတွင်းပိုင်း ပစ်မှတ်), Service Level Agreement (သုံးစွဲသူနှင့် ချုပ်ဆိုထားသော ကတိကဝတ်)။
Disk Space ပြည့်သွားပါက သို့မဟုတ် Memory Leak ဖြစ်ပါက Lambda က အလိုအလျောက် သန့်ရှင်းရေးလုပ်ပေးသော စနစ်။
လက်တွေ့ လေ့လာနိုင်သော Case Studies များ
Automatically identifying and remediating orphan EBS volumes, unattached Elastic IPs, and idle development RDS instances using AWS Config and EventBridge.
Detecting production degradations with CloudWatch Composite Alarms, X-Ray distributed tracing, and executing automated Lambda self-healing.