☁️ Core AWS Services

EC2, S3, RDS, Lambda va boshqa asosiy xizmatlar

AWS Cloud Practitioner · CLF-C02 · ⏱️ 30 min · ❓ 8 savol

💻 Amazon EC2 (Elastic Compute Cloud)

Virtual serverlar (instancelar) ta'minlaydi. Instance types: t3.micro (general), c5 (compute), r5 (memory), p3 (GPU). Pricing: On-Demand, Reserved (1-3 yil), Spot (arzonroq, interrupt bo'lishi mumkin), Dedicated Host.

🪣 Amazon S3 (Simple Storage Service)

Object storage. Fayl hajmi: 0 — 5TB. Storage classes: S3 Standard (frequent access), S3 IA (infrequent), S3 Glacier (archive, minutlar-soatlar), S3 Glacier Deep Archive (12+ soat). Versioning, replication, lifecycle policies mavjud.

🗄️ Amazon RDS & DynamoDB

RDS — managed relational DB (MySQL, PostgreSQL, Aurora, Oracle, SQL Server). Automated backups, Multi-AZ, read replicas.
DynamoDB — serverless NoSQL, millisecond latency, auto-scaling, DAX (in-memory cache).

⚡ AWS Lambda

Serverless compute. Kod run qilish uchun server yaratmaysiz. 15 minutgacha run time. Pay per execution (invocation + duration). Triggerlar: S3, DynamoDB, API Gateway, CloudWatch Events.

🌐 Networking: VPC, ELB, Route 53

VPC — Virtual Private Cloud, izolyatsiyalangan network. ELB — Elastic Load Balancer (ALB, NLB, CLB). Route 53 — DNS xizmati, health check, routing policies (simple, weighted, latency, failover, geolocation).

💡 Asosiy nuqtalar

EC2: virtual servers, multiple instance families
S3: object storage, unlimited capacity, 11 nines durability
RDS: managed SQL, DynamoDB: managed NoSQL
Lambda: serverless, event-driven, max 15 min
VPC: izolyatsiyalangan cloud network

📋 Kod misoli

# EC2 instance types misollari:
t3.micro   → General Purpose (free tier)
c5.large   → Compute Optimized
r5.xlarge  → Memory Optimized
p3.2xlarge → GPU / ML workloads

# S3 Storage Classes (narx yuqoridan pastga):
S3 Standard          → Tez-tez kirish (frequent)
S3 Standard-IA       → Kam kirish (infrequent)
S3 One Zone-IA       → 1 AZ, arzonroq
S3 Glacier Instant   → Archive, ms latency
S3 Glacier Flexible  → Archive, minlar-soatlar
S3 Glacier Deep      → Archive, 12+ soat

🎯 Imtihon maslahatlari

EC2 Spot Instance — arzon, lekin interrupt bo'lishi mumkin → fault-tolerant workload (batch, CI/CD) uchun
S3 durability = 11 nines (99.999999999%) — data yo'qolish ehtimoli astronomik kichik
Lambda max 15 daqiqa — undan uzun job uchun ECS Fargate, AWS Batch yoki Step Functions
RDS Multi-AZ — availability, Read Replica — performance (read scaling)
DynamoDB DAX — in-memory cache, microsecond latency (DynamoDB o'zi millisecond)

⚠️ Ko'p adashadigan

S3 "unlimited capacity" = bitta fayl cheksiz deb o'ylash — bitta fayl max 5 TB
RDS Multi-AZ va Read Replica ni aralashtirib yuborish: Multi-AZ = HA failover, Read Replica = read scaling
Lambda "serverless" = bepul deb o'ylash — invocation + duration bo'yicha hisob-kitob qilinadi

🧠 Eslab qolish: "ERLSVD" = EC2 (compute), RDS (relational), Lambda (serverless), S3 (storage), VPC (network), DynamoDB (NoSQL) — 6 asosiy AWS xizmat. "Spot = arzon + interrupt, Reserved = arzon + majburiyat, On-Demand = qimmat + erkin"