Concert Booking

flowchart LR
    Client[Client Browser]
    subgraph App[Spring Boot Cluster]
      App1[App Instance 1]
      App2[App Instance 2]
    end
    Redis[Redis]
    PG[(PostgreSQL)]
    Kafka[Kafka KRaft]
    Client -->|REST API + SSE| App1
    Client -->|REST API + SSE| App2
    App1 -->|Redisson Lock| Redis
    App2 -->|Redisson Lock| Redis
    App1 --> PG
    App2 --> PG
    App1 --> Kafka
    App2 --> Kafka
    Redis -->|Queue / Hold / Token| App1
    Redis -->|Queue / Hold / Token| App2

erDiagram
    users ||--o{ reservations : "예매"
    concerts ||--o{ concert_schedules : "스케줄"
    concert_schedules ||--o{ seats : "좌석"
    concert_schedules ||--o{ reservations : "예매"
    reservations ||--o{ reservation_seats : "좌석 배정"
    seats ||--o{ reservation_seats : "예매됨"
    reservations ||--o| payments : "결제"

    users {
      bigserial id PK
      varchar email UK
      varchar password
      varchar nickname
      timestamp created_at
    }
    concerts {
      bigserial id PK
      varchar title
      text description
      varchar venue
      varchar artist
      timestamp created_at
    }
    concert_schedules {
      bigserial id PK
      bigint concert_id FK
      date schedule_date
      time start_time
      int total_seats
      int available_seats
      bigint version
      timestamp created_at
    }
    seats {
      bigserial id PK
      bigint schedule_id FK
      varchar section
      int row_number
      int seat_number
      int price
      varchar status
      bigint version
      timestamp created_at
    }
    reservations {
      bigserial id PK
      uuid reservation_key UK
      bigint user_id FK
      bigint schedule_id FK
      varchar status
      int total_amount
      timestamp expires_at
      timestamp created_at
    }
    reservation_seats {
      bigserial id PK
      bigint reservation_id FK
      bigint seat_id FK
    }
    payments {
      bigserial id PK
      uuid payment_key UK
      bigint reservation_id FK
      int amount
      varchar status
      timestamp created_at
    }

sequenceDiagram
    participant C as Client
    participant API as QueueController
    participant QS as QueueService
    participant R as Redis
    C->>API: POST /api/queue/enter
    API->>QS: enter(scheduleId, userId)
    QS->>R: ZADD queue:schedule NX
    QS-->>C: 대기열 등록
    C->>API: GET /api/queue/events
    API->>QS: subscribe()
    loop 순번 갱신
      QS->>R: ZRANK / ZCARD
      QS-->>C: SSE position update
    end
    alt 100등 이내
      QS->>R: 입장 토큰 저장(TTL 5분)
      QS-->>C: SSE ready + token
    end

sequenceDiagram
    participant C as Client
    participant API as ReservationController
    participant QI as QueueTokenInterceptor
    participant RS as ReservationService
    participant R as Redis
    participant DB as PostgreSQL
    participant PS as PaymentService
    participant K as Kafka
    C->>API: POST /api/reservations
    API->>QI: queue token validate
    QI-->>API: 통과
    API->>RS: reserve(seatIds)
    RS->>R: stock DECR
    RS->>R: Redisson MultiLock
    RS->>DB: SELECT schedule / seats FOR UPDATE
    RS->>DB: reservation 저장
    RS->>R: unlock
    RS-->>C: reservationId
    C->>PS: POST /api/payments
    PS->>DB: payment 저장
    PS->>K: reservation.completed
    PS-->>C: 결제 성공

sequenceDiagram
    participant S as ExpirationScheduler
    participant DB as PostgreSQL
    participant K as Kafka
    participant C as SeatReleaseConsumer
    participant R as Redis
    S->>DB: 만료된 PENDING 예약 조회
    S->>K: reservation.cancelled 발행
    K->>C: consume
    C->>DB: reservation 상태 EXPIRED
    C->>DB: seat 상태 AVAILABLE
    C->>R: 재고 복구 + hold key 정리