Skip to content

system-architecture-diagram.md

Latest commit

 

History

History
250 lines (206 loc) · 7.1 KB

system-architecture-diagram.md

File metadata and controls

250 lines (206 loc) · 7.1 KB
Raw

리눅스로 한 학기 살기 - 전체 시스템 아키텍처

🏗️ 전체 시스템 연결 구조

graph TB
    subgraph "Development Environment"
        DEV[개발자 💻]
        VSCODE[VS Code]
        GIT[Git Repository]
    end
    
    subgraph "CI/CD Pipeline"
        GITHUB[GitHub Repository 📁]
        JENKINS[Jenkins 🔧]
        DOCKERHUB[Docker Hub 📦]
    end
    
    subgraph "Kubernetes Cluster (Minikube)"
        subgraph "Default Namespace"
            DISCORDBOT[Discord Bot Pod 🤖]
        end
        
        subgraph "Monitoring Namespace"
            PROMETHEUS[Prometheus 📊]
            GRAFANA[Grafana 📈]
            ALERTMANAGER[AlertManager 🚨]
            SLACKBOT[Slack Bot Pod 💬]
        end
        
        subgraph "Kubernetes Services"
            DISCORD_SVC[Discord Bot Service]
            PROM_SVC[Prometheus Service]
            GRAFANA_SVC[Grafana Service]
            ALERT_SVC[AlertManager Service]
            SLACK_SVC[Slack Bot Service]
        end
    end
    
    subgraph "External Services"
        DISCORD_API[Discord API 💬]
        SLACK_API[Slack API 📢]
        DOCKER_REGISTRY[Docker Registry 🐳]
    end
    
    subgraph "Storage & Config"
        CONFIGMAPS[ConfigMaps ⚙️]
        SECRETS[Secrets 🔐]
        VOLUMES[Persistent Volumes 💾]
    end
    
    %% Development Flow
    DEV --> VSCODE
    VSCODE --> GIT
    GIT --> GITHUB
    
    %% CI/CD Flow
    GITHUB -->|Webhook| JENKINS
    JENKINS -->|Build & Test| JENKINS
    JENKINS -->|Push Image| DOCKERHUB
    JENKINS -->|Deploy| DISCORDBOT
    JENKINS -->|Deploy| SLACKBOT
    
    %% Docker Hub Connections
    DOCKERHUB -->|Pull Image| DISCORDBOT
    DOCKERHUB -->|Pull Image| SLACKBOT
    DOCKERHUB --> DOCKER_REGISTRY
    
    %% Service Connections
    DISCORDBOT --> DISCORD_SVC
    PROMETHEUS --> PROM_SVC
    GRAFANA --> GRAFANA_SVC
    ALERTMANAGER --> ALERT_SVC
    SLACKBOT --> SLACK_SVC
    
    %% Monitoring Flow
    DISCORDBOT -->|Metrics :8000/metrics| PROMETHEUS
    SLACKBOT -->|Metrics :8000/metrics| PROMETHEUS
    PROMETHEUS -->|Data Source| GRAFANA
    PROMETHEUS -->|Alerts| ALERTMANAGER
    ALERTMANAGER -->|Notifications| SLACK_API
    
    %% External API Connections
    DISCORDBOT -->|Bot Commands| DISCORD_API
    SLACKBOT -->|Messages| SLACK_API
    
    %% Configuration
    CONFIGMAPS -->|Config| PROMETHEUS
    CONFIGMAPS -->|Config| GRAFANA
    CONFIGMAPS -->|Config| ALERTMANAGER
    SECRETS -->|Tokens| DISCORDBOT
    SECRETS -->|Webhook URL| ALERTMANAGER
    VOLUMES -->|Storage| PROMETHEUS
    VOLUMES -->|Storage| GRAFANA
    
    %% Styling
    classDef development fill:#e1f5fe
    classDef cicd fill:#f3e5f5
    classDef k8s fill:#e8f5e8
    classDef external fill:#fff3e0
    classDef storage fill:#fce4ec
    
    class DEV,VSCODE,GIT development
    class GITHUB,JENKINS,DOCKERHUB cicd
    class DISCORDBOT,PROMETHEUS,GRAFANA,ALERTMANAGER,SLACKBOT,DISCORD_SVC,PROM_SVC,GRAFANA_SVC,ALERT_SVC,SLACK_SVC k8s
    class DISCORD_API,SLACK_API,DOCKER_REGISTRY external
    class CONFIGMAPS,SECRETS,VOLUMES storage
Loading

📊 포트 및 네트워크 구성

graph LR
    subgraph "Host Machine (Ubuntu)"
        HOST_30090[":30090 NodePort"]
        HOST_30300[":30300 NodePort"]
        HOST_30500[":30500 NodePort"]
        HOST_9093[":9093 Port-Forward"]
    end
    
    subgraph "Kubernetes Cluster"
        subgraph "Discord Bot"
            DISCORD_POD[Discord Bot Pod]
            DISCORD_8000[":8000 Metrics"]
        end
        
        subgraph "Prometheus"
            PROM_POD[Prometheus Pod]
            PROM_9090[":9090 Web UI"]
        end
        
        subgraph "Grafana"
            GRAFANA_POD[Grafana Pod]
            GRAFANA_3000[":3000 Web UI"]
        end
        
        subgraph "AlertManager"
            ALERT_POD[AlertManager Pod]
            ALERT_9093_INT[":9093 Web UI"]
        end
        
        subgraph "Slack Bot"
            SLACK_POD[Slack Bot Pod]
            SLACK_8000[":8000 Metrics"]
            SLACK_5000[":5000 Test Endpoint"]
        end
    end
    
    %% Port Mappings
    HOST_30090 --> PROM_9090
    HOST_30300 --> GRAFANA_3000
    HOST_30500 --> SLACK_5000
    HOST_9093 --> ALERT_9093_INT
    
    %% Internal Connections
    PROM_POD -->|Scrape| DISCORD_8000
    PROM_POD -->|Scrape| SLACK_8000
    GRAFANA_POD -->|Query| PROM_9090
    PROM_POD -->|Send Alerts| ALERT_9093_INT
    
    classDef host fill:#ffeb3b
    classDef internal fill:#4caf50
    class HOST_30090,HOST_30300,HOST_30500,HOST_9093 host
    class DISCORD_8000,PROM_9090,GRAFANA_3000,ALERT_9093_INT,SLACK_8000,SLACK_5000 internal
Loading

🔄 데이터 플로우

sequenceDiagram
    participant User as 사용자
    participant Discord as Discord Bot
    participant Prom as Prometheus
    participant Grafana as Grafana
    participant Alert as AlertManager
    participant Slack as Slack
    
    Note over User,Slack: 정상 운영 플로우
    User->>Discord: ?ping 명령어
    Discord->>User: 응답 + 메트릭 업데이트
    Discord->>Prom: 메트릭 노출 (:8000/metrics)
    
    Note over Prom: 30초마다 스크래핑
    Prom->>Discord: GET /metrics
    Prom->>Grafana: 데이터 제공
    Grafana->>User: 대시보드 표시 (:30300)
    
    Note over User,Slack: 알람 플로우
    Discord->>Discord: 에러 발생 (5회 이상)
    Prom->>Prom: 알람 규칙 평가
    Prom->>Alert: 알람 전송
    Alert->>Slack: Slack 알림 전송
    Slack->>User: 알림 메시지
Loading

🛠️ 배포 플로우

sequenceDiagram
    participant Dev as 개발자
    participant Git as GitHub
    participant Jenkins as Jenkins
    participant Hub as Docker Hub
    participant K8s as Kubernetes
    participant Bot as Discord Bot
    
    Dev->>Git: git push
    Git->>Jenkins: Webhook 트리거
    
    Note over Jenkins: Jenkins Pipeline 실행
    Jenkins->>Jenkins: 1. Checkout 코드
    Jenkins->>Jenkins: 2. Python 테스트
    Jenkins->>Jenkins: 3. Docker 이미지 빌드
    Jenkins->>Hub: 4. 이미지 푸시
    Jenkins->>K8s: 5. Deployment 업데이트
    
    K8s->>Hub: 새 이미지 Pull
    K8s->>Bot: Pod 재배포
    Bot->>Bot: 새 버전 실행
Loading

📋 주요 연결점 요약

1. 개발 → 배포

  • GitHub → Jenkins (Webhook)
  • Jenkins → Docker Hub (Image Push)
  • Jenkins → Kubernetes (Deployment)

2. 모니터링 체인

  • Discord Bot → Prometheus (Metrics :8000)
  • Slack Bot → Prometheus (Metrics :8000)
  • Prometheus → Grafana (Data Source)
  • Prometheus → AlertManager (Alerts)

3. 알림 체인

  • Prometheus → AlertManager (Alert Rules)
  • AlertManager → Slack (Webhook Notifications)

4. 네트워크 접근

  • Prometheus: localhost:30090
  • Grafana: localhost:30300
  • Slack Bot Test: localhost:30500
  • AlertManager: kubectl port-forward 필요

5. 설정 관리

  • ConfigMaps: Prometheus, Grafana, AlertManager 설정
  • Secrets: Discord Token, Slack Webhook URL
  • Persistent Volumes: Prometheus, Grafana 데이터 저장

이 구조를 통해 코드 변경부터 모니터링, 알림까지 완전 자동화된 DevOps 파이프라인이 구현되어 있습니다! 🚀