Event Seat Booking on Website

Implementing Event Seat Booking on a Website

Event seat booking differs from other types: seats are fixed on a venue map, sales come in waves (early bird → general sale → last chance), and load can peak at sales start.

Data Model

CREATE TABLE events (
    id              SERIAL PRIMARY KEY,
    title           VARCHAR(255) NOT NULL,
    description     TEXT,
    venue_name      VARCHAR(255),
    venue_address   TEXT,
    starts_at       TIMESTAMP NOT NULL,
    ends_at         TIMESTAMP,
    sales_open_at   TIMESTAMP,
    sales_close_at  TIMESTAMP,
    status          VARCHAR(20) DEFAULT 'draft',  -- draft|on_sale|sold_out|finished
    schema_id       INTEGER,     -- venue map
    cover_image     VARCHAR(500)
);

CREATE TABLE seat_categories (
    id          SERIAL PRIMARY KEY,
    event_id    INTEGER REFERENCES events(id),
    name        VARCHAR(100),   -- 'Orchestra', 'Balcony', 'VIP'
    color       VARCHAR(7),     -- color on venue map
    price       NUMERIC(10,2),
    total_seats INTEGER,
    available_seats INTEGER
);

CREATE TABLE seats (
    id          BIGSERIAL PRIMARY KEY,
    event_id    INTEGER REFERENCES events(id),
    category_id INTEGER REFERENCES seat_categories(id),
    row_label   VARCHAR(10),
    seat_number VARCHAR(10),
    x_pos       NUMERIC(6,2),   -- coordinate on map
    y_pos       NUMERIC(6,2),
    status      VARCHAR(20) DEFAULT 'available',
    -- available | held | sold | blocked | unavailable
    booking_id  BIGINT,
    held_until  TIMESTAMP,
    held_by     VARCHAR(100)   -- session_id
);

CREATE TABLE ticket_bookings (
    id              BIGSERIAL PRIMARY KEY,
    event_id        INTEGER REFERENCES events(id),
    customer_name   VARCHAR(255),
    customer_email  VARCHAR(255),
    customer_phone  VARCHAR(50),
    total_amount    NUMERIC(10,2),
    status          VARCHAR(20) DEFAULT 'pending',
    payment_id      VARCHAR(100),
    created_at      TIMESTAMP DEFAULT NOW(),
    paid_at         TIMESTAMP,
    cancelled_at    TIMESTAMP
);

CREATE TABLE ticket_booking_seats (
    booking_id  BIGINT REFERENCES ticket_bookings(id),
    seat_id     BIGINT REFERENCES seats(id),
    price       NUMERIC(10,2),
    PRIMARY KEY (booking_id, seat_id)
);

Seat Hold with TTL

When the client selects seats, they are temporarily blocked until payment completes or timer expires:

HOLD_TTL_SECONDS = 600  # 10 minutes

def hold_seats(seat_ids: list[int], session_id: str) -> bool:
    with db.transaction():
        # Atomically check and block
        result = db.execute("""
            UPDATE seats
            SET status = 'held',
                held_by = %(session)s,
                held_until = NOW() + INTERVAL '10 minutes'
            WHERE id = ANY(%(ids)s)
              AND status = 'available'
            RETURNING id
        """, {'ids': seat_ids, 'session': session_id})

        held_count = len(result)
        if held_count < len(seat_ids):
            # Not all seats available — rollback transaction
            raise db.Rollback("Some seats are no longer available")

    return True

Expired holds are released by a background process running every minute:

UPDATE seats
SET status = 'available', held_by = NULL, held_until = NULL
WHERE status = 'held' AND held_until < NOW();

Venue Map (Seat Map)

Venue layout is rendered on SVG or Canvas. Seat data is requested from backend:

{
  "sections": [
    {
      "id": 1,
      "name": "Orchestra",
      "rows": [
        {
          "label": "A",
          "seats": [
            { "id": 1001, "number": "1", "x": 100, "y": 200, "status": "available", "price": 2500 },
            { "id": 1002, "number": "2", "x": 130, "y": 200, "status": "sold", "price": 2500 }
          ]
        }
      ]
    }
  ]
}

User clicks a seat, it highlights, gets added to cart. On attempt to add already taken seat — error shows without page reload (WebSocket or polling every 5 sec).

Real-Time Updates

During high demand, seats are snapped up fast. Real-time schema updates via WebSocket:

const ws = new WebSocket(`wss://api.example.com/events/${eventId}/seats`);

ws.onmessage = (event) => {
  const { seat_id, status } = JSON.parse(event.data);
  updateSeatStatus(seat_id, status);  // update color on map
};

Backend broadcasts updates to all connected clients on each seat status change.

Pricing Categories and Sales Waves

CREATE TABLE pricing_tiers (
    id              SERIAL PRIMARY KEY,
    event_id        INTEGER REFERENCES events(id),
    name            VARCHAR(100),   -- 'Early Bird', 'Standard'
    category_id     INTEGER REFERENCES seat_categories(id),
    price           NUMERIC(10,2),
    available_from  TIMESTAMP,
    available_until TIMESTAMP,
    quota           INTEGER         -- seat limit at this price
);

System automatically switches between tiers on schedule.

E-Tickets

After payment, each ticket is generated as PDF with QR code:

import qrcode
from reportlab.pdfgen import canvas

def generate_ticket_pdf(ticket: dict) -> bytes:
    qr = qrcode.make(ticket['verification_code'])
    # ... overlay QR on PDF template
    return pdf_bytes

QR contains a signed token with ticket_id and booking_id. At entry — scanned and verified via API.

Implementation Timeline

Without venue map, simple seat numbering, online payment — 10–12 business days. Full seat map (SVG), real-time updates, e-tickets with QR, pricing tiers, entry verification — 16–22 business days.