Ticketmaster

A ticket sales platform built with a microservices architecture in Go, featuring event-driven communication via CDC (Change Data Capture), distributed locking for ticket reservations, and full-text search powered by Elasticsearch.

Architecture Overview

The system is composed of 3 independent Go microservices that share a single PostgreSQL database and communicate asynchronously via Kafka (CDC):

┌─────────┐     ┌──────────────────────┐     ┌─────────────┐
│  Client │────▶│  Event Management    │────▶│  PostgreSQL │
│         │     │  API (:8080)         │◀────│             │
└─────────┘     └──────────┬───────────┘     └──────┬──────┘
     │                     │                        │
     │              ┌──────▼──────┐          ┌──────▼──────┐
     │              │    Redis    │          │  Debezium   │
     │              │  (Cache +   │          │  Connect    │
     │              │   Locks)    │          └──────┬──────┘
     │              └──────▲──────┘                 │
     │                     │                 ┌──────▼──────┐
     │          ┌──────────┴───────────┐     │    Kafka    │
     ├─────────▶│  Booking API (:8081) │     │             │
     │          └──────────────────────┘     └──────┬──────┘
     │                                              │
     │          ┌──────────────────────┐     ┌──────▼───────────┐
     └─────────▶│  Search API (:8082)  │────▶│  Elasticsearch   │
                └──────────────────────┘     └──────────────────┘

Key architectural patterns:

Pattern	Description
Cache-Aside	Redis caches event data (1h TTL) and search results (5min TTL) to reduce DB/ES load
Distributed Lock	Redis-based locks with 15-minute TTL for temporary ticket reservations
Two-Phase Booking	Reserve (temporary lock) → Confirm (persist to DB)
CDC (Change Data Capture)	Debezium captures PostgreSQL WAL changes → publishes to Kafka → Search API consumes and indexes into Elasticsearch
Clean Architecture	Each service follows: entities → repositories → services → usecases → handlers

Tech Stack

Component	Technology	Version
Language	Go	1.25
HTTP Framework	Echo	v5
ORM	GORM (with generics)	v1.31
Database	PostgreSQL	17
Cache / Locks	Redis	8
Message Broker	Apache Kafka	4.0
CDC Connector	Debezium	3.5
Search Engine	Elasticsearch	9.3
Migrations	Goose	—
Hot Reload	Air	—
Containers	Docker Compose	—

Entity Relationship Diagram

┌───────────────────────────┐       ┌───────────────────────────────┐
│          venues           │       │          performers           │
├───────────────────────────┤       ├───────────────────────────────┤
│ PK  id         BIGINT     │       │ PK  id          BIGINT        │
│     uuid       TEXT UNIQUE│       │     uuid        TEXT UNIQUE   │
│     created_at TIMESTAMPTZ│       │     created_at  TIMESTAMPTZ   │
│     updated_at TIMESTAMPTZ│       │     updated_at  TIMESTAMPTZ   │
│     location   TEXT       │       │     name        TEXT          │
│     seat_map   JSON       │       │     age         BIGINT        │
└─────────────┬─────────────┘       │     description TEXT NULL     │
              │ 1                    └──────────────┬────────────────┘
              │                                     │
              │                                     │ M
              │                     ┌───────────────┴────────────────┐
              │                     │       event_performers         │
              │                     ├────────────────────────────────┤
              │                     │ PK,FK  event_id     BIGINT     │
              │                     │ PK,FK  performer_id BIGINT     │
              │                     └───────────────┬────────────────┘
              │                                     │ M
              │                                     │
              │ N    ┌──────────────────────────────┤
              └──────┤           events             │
                     ├──────────────────────────────┤
                     │ PK  id          BIGINT       │
                     │     uuid        TEXT UNIQUE  │
                     │     created_at  TIMESTAMPTZ  │
                     │     updated_at  TIMESTAMPTZ  │
                     │ FK  venue_id    BIGINT       │
                     │     date        TIMESTAMPTZ  │
                     │     name        TEXT         │
                     │     description TEXT NULL       │
                     └──────────────┬────────────────┘
                                    │ 1
                                    │
                                    │ N
                     ┌──────────────┴────────────────┐
                     │           tickets              │
                     ├───────────────────────────────┤
                     │ PK  id        BIGINT           │
                     │     uuid      TEXT UNIQUE       │
                     │     created_at TIMESTAMPTZ      │
                     │     updated_at TIMESTAMPTZ      │
                     │ FK  event_id  BIGINT           │
                     │     price     BIGINT           │
                     │     seat      TEXT             │
                     │     status    TEXT             │
                     │              (available|booked) │
                     └────────────────────────────────┘

Relationships

Relationship	Type	Cascade
venues → events	One-to-Many	ON UPDATE CASCADE, ON DELETE CASCADE
events → tickets	One-to-Many	ON UPDATE CASCADE, ON DELETE CASCADE
events ↔ performers	Many-to-Many (via `event_performers`)	ON UPDATE CASCADE, ON DELETE CASCADE

Indexes

venues: location
performers: name
events: venue_id, date, name
tickets: event_id, seat, status
event_performers: performer_id

Services

Event Management API

Port: 8080 | Base path: /api/v1/management

Responsible for querying event details with all related data (venue, performers, tickets). Uses Redis cache (1h TTL) for event data and checks ticket reservation status in real-time via Redis locks.

Endpoint:

Method	Path	Description
`GET`	`/api/v1/management/events/:uuid`	Find event by UUID with full details

Response example:

{
  "uuid": "d0e1f2a3-b4c5-4d5e-8f9a-0b1c2d3e4f5a",
  "name": "The Eras Tour - New York",
  "description": "Taylor Swift brings her iconic Eras Tour...",
  "date": "2026-07-15T20:00:00Z",
  "venue": {
    "uuid": "a1b2c3d4-e5f6-4a5b-8c9d-0e1f2a3b4c5d",
    "location": "Madison Square Garden, New York",
    "seat_map": { "A1": true, "A2": true, "B1": true }
  },
  "performers": [
    {
      "uuid": "e5f6a7b8-c9d0-4e5f-8a9b-0c1d2e3f4a5b",
      "name": "Taylor Swift",
      "age": 34,
      "description": "Multi-award winning pop and country music artist..."
    }
  ],
  "tickets": [
    { "uuid": "c5d6e7f8-...", "price": 35000, "seat": "A1", "status": "available" },
    { "uuid": "d6e7f8a9-...", "price": 35000, "seat": "A2", "status": "reserved" },
    { "uuid": "f8a9b0c1-...", "price": 30000, "seat": "B1", "status": "booked" }
  ]
}

Note: The status field on tickets reflects real-time state. A ticket can be available, reserved (temporarily locked in Redis), or booked (confirmed in DB).

Booking API

Port: 8081 | Base path: /api/v1/booking

Handles the two-phase ticket purchasing flow. Uses Redis distributed locks with a 15-minute TTL for temporary reservations.

Endpoints:

Method	Path	Description
`POST`	`/api/v1/booking/tickets/reserve`	Reserve a ticket (temporary lock)
`POST`	`/api/v1/booking/tickets/confirm`	Confirm purchase (persist to DB)

Reserve request:

{
  "ticket_uuid": "c5d6e7f8-a9b0-4c5d-8e9f-0a1b2c3d4e5f",
  "user_uuid": "550e8400-e29b-41d4-a716-446655440000"
}

Confirm request:

{
  "ticket_uuid": "c5d6e7f8-a9b0-4c5d-8e9f-0a1b2c3d4e5f",
  "user_uuid": "550e8400-e29b-41d4-a716-446655440000",
  "payment_method_uuid": "660e8400-e29b-41d4-a716-446655440000"
}

Response codes:

Status	Meaning
`200`	Success (no body)
`400`	Invalid request payload
`409`	Ticket not available (reserve) or not reserved for user (confirm)
`500`	Internal server error

Redis key structure:

Key Pattern	Purpose	TTL
`ticket_lock:{ticket_uuid}`	Global lock — prevents other users from reserving	15 min
`ticket_lock_user:{ticket_uuid}:{user_uuid}`	User-specific lock — validates ownership on confirm	15 min

Search API

Port: 8082 | Base path: /api/v1/search

Provides full-text search over events using Elasticsearch. Data is synced from PostgreSQL via CDC (Debezium → Kafka). Results are cached in Redis for 5 minutes.

Endpoint:

Method	Path	Description
`GET`	`/api/v1/search/events?q=&page=&size=`	Search events by text query

Query parameters:

Param	Type	Default	Description
`q`	string	—	Search query (matches name, description, location)
`page`	int	1	Page number
`size`	int	10	Results per page (max 100)

Response example:

{
  "events": [
    {
      "uuid": "d0e1f2a3-b4c5-4d5e-8f9a-0b1c2d3e4f5a",
      "name": "The Eras Tour - New York",
      "description": "Taylor Swift brings her iconic Eras Tour...",
      "date": "2026-07-15T20:00:00Z",
      "location": "Madison Square Garden, New York"
    }
  ],
  "total": 1,
  "page": 1,
  "size": 10
}

Search behavior:

Fuzzy matching (fuzziness: AUTO) across name (4x boost), location (2x boost), and description
Only returns future events (date >= now)
Sorted by relevance score (desc), then date (asc)

CDC Consumer — Kafka topics:

Topic	Trigger	Action
`pg.public.events`	Event created/updated	Upsert document in Elasticsearch `events` index
`pg.public.events`	Event deleted	Delete document from Elasticsearch index
`pg.public.venues`	Venue updated	Update `location` field on all related event documents

Flowcharts

Find Event by UUID Flow

                    GET /events/:uuid
                          │
                          ▼
                ┌─────────────────┐
                │  Check Redis    │
                │  cache for UUID │
                └────────┬────────┘
                         │
                   ┌─────┴─────┐
                   │ Cache hit? │
                   └─────┬─────┘
                    Yes  │  No
              ┌──────────┴──────────┐
              ▼                     ▼
     ┌────────────┐      ┌──────────────────┐
     │ Use cached │      │ Query PostgreSQL  │
     │   event    │      │ (with preloads:   │
     └─────┬──────┘      │ venue, performers,│
           │             │ tickets)          │
           │             └────────┬──────────┘
           │                      │
           │                      ▼
           │             ┌────────────────┐
           │             │ Store in Redis │
           │             │  cache (1h)    │
           │             └────────┬───────┘
           │                      │
           └──────────┬───────────┘
                      ▼
           ┌──────────────────────┐
           │ Check Redis for each │
           │ ticket lock key      │
           │ (ticket_lock:{uuid}) │
           └──────────┬───────────┘
                      ▼
           ┌──────────────────────┐
           │ Mark locked tickets  │
           │ as "reserved"        │
           └──────────┬───────────┘
                      ▼
               Return EventDTO

Reserve Ticket Flow

          POST /tickets/reserve
          { ticket_uuid, user_uuid }
                    │
                    ▼
          ┌───────────────────┐
          │ Check ticket       │
          │ status in DB       │
          │ (status=available?)│
          └─────────┬─────────┘
                    │
              ┌─────┴─────┐
              │ Available? │
              └─────┬─────┘
           No  │         │  Yes
               ▼         ▼
          ┌─────────┐  ┌──────────────────────┐
          │ 409     │  │ Redis Pipeline:       │
          │ Conflict│  │ SETNX ticket_lock:    │
          └─────────┘  │   {uuid} (TTL 15min)  │
                       │ SETNX ticket_lock_    │
                       │   user:{uuid}:{user}  │
                       │   (TTL 15min)          │
                       └──────────┬─────────────┘
                                  │
                            ┌─────┴─────┐
                            │ Lock set? │
                            └─────┬─────┘
                         No  │         │  Yes
                             ▼         ▼
                        ┌─────────┐  ┌─────────┐
                        │ 409     │  │ 200 OK  │
                        │ Conflict│  │         │
                        └─────────┘  └─────────┘

Confirm Ticket Flow

        POST /tickets/confirm
        { ticket_uuid, user_uuid,
          payment_method_uuid }
                    │
                    ▼
          ┌──────────────────────┐
          │ Check Redis for      │
          │ ticket_lock_user:    │
          │ {ticket}:{user}      │
          └──────────┬───────────┘
                     │
               ┌─────┴─────┐
               │ Reserved   │
               │ by user?   │
               └─────┬─────┘
            No  │         │  Yes
                ▼         ▼
           ┌─────────┐  ┌──────────────────┐
           │ 409     │  │ TODO: Process    │
           │ Conflict│  │ payment via      │
           └─────────┘  │ Stripe           │
                        └────────┬─────────┘
                                 ▼
                        ┌──────────────────┐
                        │ Update ticket    │
                        │ status → "booked"│
                        │ in PostgreSQL    │
                        └────────┬─────────┘
                                 ▼
                            ┌─────────┐
                            │ 200 OK  │
                            └─────────┘

Search Events Flow

      GET /search/events?q=taylor&page=1&size=10
                    │
                    ▼
          ┌───────────────────┐
          │ Build cache key:  │
          │ search:taylor:1:10│
          └─────────┬─────────┘
                    ▼
          ┌───────────────────┐
          │ Check Redis cache │
          └─────────┬─────────┘
                    │
              ┌─────┴─────┐
              │ Cache hit? │
              └─────┬─────┘
               Yes  │  No
          ┌─────────┴─────────┐
          ▼                   ▼
   ┌──────────┐     ┌────────────────────┐
   │  Return  │     │ Query Elasticsearch│
   │  cached  │     │ - multi_match on   │
   └──────────┘     │   name^4, location │
                    │   ^2, description  │
                    │ - filter: date≥now │
                    │ - fuzziness: AUTO  │
                    └────────┬───────────┘
                             ▼
                    ┌────────────────────┐
                    │ Cache result in    │
                    │ Redis (5min TTL)   │
                    └────────┬───────────┘
                             ▼
                      Return results

CDC Pipeline Flow

  ┌──────────────┐    WAL     ┌──────────────┐   Kafka    ┌────────────┐
  │  PostgreSQL  │───────────▶│   Debezium   │──────────▶│   Kafka    │
  │              │  (logical  │   Connect    │  topics:   │            │
  │ tables:      │ replicat.) │              │ pg.public. │ pg.public. │
  │ - events     │            │ Monitors:    │ events     │ events     │
  │ - venues     │            │ - events     │ pg.public. │ pg.public. │
  │              │            │ - venues     │ venues     │ venues     │
  └──────────────┘            └──────────────┘            └──────┬─────┘
                                                                 │
                                                                 ▼
                                                        ┌────────────────┐
                                                        │  Search API    │
                                                        │  CDC Consumer  │
                                                        └────────┬───────┘
                                                                 │
                              ┌───────────────────────────┬──────┴────────┐
                              ▼                           ▼               ▼
                    ┌──────────────────┐       ┌───────────────┐  ┌──────────────┐
                    │ op=c or op=u     │       │ op=d          │  │ venue op=u   │
                    │ (create/update)  │       │ (delete)      │  │              │
                    │                  │       │               │  │              │
                    │ Upsert event doc │       │ Delete event  │  │ Update all   │
                    │ in Elasticsearch │       │ from ES index │  │ events with  │
                    │ (fetch venue for │       │               │  │ venue_id →   │
                    │  location)       │       │               │  │ new location │
                    └──────────────────┘       └───────────────┘  └──────────────┘

Running the Project

Prerequisites

Docker and Docker Compose
Go 1.25+ (for local development)
Goose (database migrations)
Air (hot reload, optional for local dev)

Full Stack (Docker Compose)

This is the easiest way to run the entire platform:

# 1. Start all containers (infra + APIs)
make up

# 2. Wait for services, then run migrations + create Debezium connector + seed data
make setup

Services will be available at:

Service	URL
Event Management API	http://localhost:8080
Booking API	http://localhost:8081
Search API	http://localhost:8082
Elasticsearch	http://localhost:9200
Kibana	http://localhost:5601
Kafka Connect	http://localhost:8083
PostgreSQL	localhost:5432
Redis	localhost:6379

Useful commands:

# Check migration status
make migrate-status

# Check Debezium connector status
make connector-status

# Stop all containers
make down

# Stop and remove all data (volumes)
make reset

# Full teardown (delete connector + rollback migrations)
make teardown

Local Development

For developing a single service with hot reload:

# 1. Start only infrastructure containers
make up-components

# 2. Run migrations and setup CDC
make setup

# 3. Navigate to the desired service and start with hot reload
cd event-management && make run-dev   # starts on :8080
cd booking && make run-dev            # starts on :8080 (use different port or docker)
cd search && make run-dev             # starts on :8080

Note: When running services locally, you need to set environment variables or create a .env file matching the variables from docker-compose.yml (e.g., POSTGRES_HOST=localhost).

Environment variables required by each service:

Event Management & Booking

POSTGRES_HOST=localhost
POSTGRES_PORT=5432
POSTGRES_USER=main
POSTGRES_PASSWORD=password
POSTGRES_DBNAME=event-management
POSTGRES_SSLMODE=disable
POSTGRES_TIMEZONE=UTC
POSTGRES_MAX_OPEN_CONNS=25
POSTGRES_CONN_MAX_LIFETIME=300
POSTGRES_MAX_IDLE_CONNS=10
POSTGRES_CONN_MAX_IDLE_TIME=60
SERVER_MODE=development
SERVER_ADDR=:8080
SERVER_READ_TIMEOUT=10
SERVER_WRITE_TIMEOUT=10
SERVER_MAX_HEADER_BYTES=1048576
SERVER_CTX_TIMEOUT=30
REDIS_ADDR=localhost:6379
REDIS_PASSWORD=
REDIS_DB=0

Search (additional variables)

# All PostgreSQL, Server, and Redis vars from above, plus:
KAFKA_LISTENERS=localhost:9092
ELASTICSEARCH_ADDR=http://localhost:9200

API Reference

Event Management

GET /api/v1/management/events/:uuid

Returns full event details including venue, performers, and tickets with real-time reservation status.

Status	Response
200	`EventDTO`
404	`{ "error": "...", "message": "Event not found" }`
500	`{ "error": "...", "message": "Internal server error" }`

Booking

POST /api/v1/booking/tickets/reserve
Content-Type: application/json

{ "ticket_uuid": "string", "user_uuid": "string" }

Status	Response
200	No content
400	`{ "error": "...", "message": "Invalid request payload" }`
409	`{ "error": "...", "message": "Ticket is not available for reservation" }`
500	`{ "error": "...", "message": "Internal server error" }`

POST /api/v1/booking/tickets/confirm
Content-Type: application/json

{ "ticket_uuid": "string", "user_uuid": "string", "payment_method_uuid": "string" }

Status	Response
200	No content
400	`{ "error": "...", "message": "Invalid request payload" }`
409	`{ "error": "...", "message": "Ticket is not reserved for this user" }`
500	`{ "error": "...", "message": "Internal server error" }`

Search

GET /api/v1/search/events?q={query}&page={page}&size={size}

Param	Type	Default	Constraint
q	string	—	Search text
page	int	1	≥ 1
size	int	10	1–100

Status	Response
200	`SearchEventsResponseDTO`
500	`{ "error": "...", "message": "Internal server error" }`

Project Structure

ticketmaster/
├── docker-compose.yml          # Full infrastructure + services
├── connector-config.json       # Debezium PostgreSQL CDC connector config
├── Makefile                    # Orchestration: migrations, CDC setup, Docker
├── @docs/                      # Architecture diagrams
│
├── event-management/           # Event query service
│   ├── cmd/server/main.go
│   ├── internal/
│   │   ├── entities/           # Domain models (Event, Venue, Performer, Ticket)
│   │   ├── repositories/       # Data access layer (GORM)
│   │   ├── services/           # Event cache (Redis), Ticket lock check (Redis)
│   │   ├── usecases/           # FindEventByUUID
│   │   ├── handlers/           # HTTP handlers + DTOs
│   │   └── server/             # Config, DB, Redis, Echo setup
│   └── migrations/             # Goose SQL migrations + seeds
│
├── booking/                    # Ticket reservation & confirmation service
│   ├── cmd/server/main.go
│   ├── internal/
│   │   ├── entities/           # Domain models
│   │   ├── repositories/       # Ticket repository (availability check, status update)
│   │   ├── services/           # Ticket lock (Redis distributed lock)
│   │   ├── usecases/           # ReserveTicket, ConfirmTicket
│   │   ├── handlers/           # HTTP handlers + DTOs
│   │   └── server/             # Config, DB, Redis, Echo setup
│
└── search/                     # Full-text search service
    ├── cmd/server/main.go
    ├── internal/
    │   ├── brokers/cdc/        # Kafka CDC consumer (Debezium events + venues)
    │   ├── entities/           # Domain models
    │   ├── repositories/       # Venue repository (for location enrichment)
    │   ├── services/           # Elasticsearch engine, Search cache (Redis)
    │   ├── usecases/           # SearchEvents
    │   ├── handlers/           # HTTP handlers + DTOs
    │   └── server/             # Config, DB, Redis, ES, Echo setup

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

Ticketmaster

Table of Contents

Architecture Overview

Tech Stack

Entity Relationship Diagram

Relationships

Indexes

Services

Event Management API

Booking API

Search API

Flowcharts

Find Event by UUID Flow

Reserve Ticket Flow

Confirm Ticket Flow

Search Events Flow

CDC Pipeline Flow

Running the Project

Prerequisites

Full Stack (Docker Compose)

Local Development

API Reference

Event Management

Booking

Search

Project Structure

About

Uh oh!

Releases

Packages

Uh oh!

Contributors

Uh oh!

Languages

Name		Name	Last commit message	Last commit date
Latest commit History 107 Commits
@docs		@docs
booking		booking
event-management		event-management
search		search
Makefile		Makefile
README.md		README.md
connector-config.json		connector-config.json
docker-compose.yml		docker-compose.yml

Folders and files

Latest commit

History

Repository files navigation

Ticketmaster

Table of Contents

Architecture Overview

Tech Stack

Entity Relationship Diagram

Relationships

Indexes

Services

Event Management API

Booking API

Search API

Flowcharts

Find Event by UUID Flow

Reserve Ticket Flow

Confirm Ticket Flow

Search Events Flow

CDC Pipeline Flow

Running the Project

Prerequisites

Full Stack (Docker Compose)

Local Development

API Reference

Event Management

Booking

Search

Project Structure

About

Topics

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages