ခဏလေး ဒီလိုပဲ စဉ်းစားကြည့်
Docker container တစ်ခုကို run တာ လွယ်ပါတယ် — ဒါပေမယ့် production မှာ container ၅၀၊ ၁၀၀ ကို တစ်ပြိုင်နက် run ရင်ကော? တစ်ခု crash သွားရင် ဘယ်သူပြန် start ပေးမလဲ? User များလာရင် ဘယ်သူ container အသစ်ထပ်ဖန်တီးမလဲ? Server တစ်လုံး down သွားရင် ဘယ်သူ container တွေကို server ကျန်ဆီ ရွှေ့ပေးမလဲ? Kubernetes က အဲဒီအလုပ်တွေကို အလိုအလျောက် လုပ်ပေးတဲ့ 'container ထိန်းချုပ်ရေးစနစ်' ပါ — Google က internal system (Borg) ကနေ idea ယူပြီး open-source လုပ်ခဲ့တာပါ။
လက်တွေ့ scenario နဲ့ ချိတ်ကြည့်မယ်
Restaurant chain တစ်ခုကို တွေးကြည့်ပါ — branch ၅၀ ရှိပြီး၊ branch တစ်ခု fire ဖြစ်သွားရင် customer ကို ချက်ချင်း အနီးဆုံး branch ကို redirect လုပ်ပေးရမယ်၊ Friday night crowd များလာရင် branch အသစ်တွေ ချက်ချင်းဖွင့်ပေးရမယ်။ လူတစ်ယောက်တည်းနဲ့ manual စီမံလို့ မဖြစ်ပါဘူး — Kubernetes ကတော့ အဲဒီ 'branch manager' အလိုအလျောက် system ပါပဲ။ Netflix, Spotify, Shopify အားလုံး Kubernetes အပေါ်မှာ run ပါတယ်။
အတူတူ ကြည့်မယ်
Without Kubernetes:
You manually SSH into servers, restart crashed containers,
add new servers when traffic spikes, remove them when it drops.
With Kubernetes:
You describe the desired state ("I want 3 copies running"),
Kubernetes keeps making it true — forever, automatically.Kubernetes ကို 'container ထိန်းချုပ်ရေး autopilot' အဖြစ် ရှင်းပြနိုင်မည်။၅ မိနစ် စမ်းကြည့်
Website တစ်ခု traffic ၁၀ ဆ ချက်ချင်းများလာရင် (viral post ဖြစ်လို့) ဘာတွေ ဖြစ်နိုင်လဲ (manual စီမံရင် vs Kubernetes သုံးရင်) စဉ်းစားပြီး ၃ ကြောင်း ရေးကြည့်ပါ။
သတိလေးတစ်ချက်
Kubernetes က powerful ပေမယ့် complexity လည်း မြင့်ပါတယ် — project အသေးလေးတစ်ခုကို Kubernetes နဲ့ start လုပ်ရင် 'ခြင်လုံးကို လက်ဗွေဖြင့် သတ်' လုပ်သလိုပါပဲ။