Game state management in LLD - Scalability & System Analysis

Scalability Analysis - Game state management

Growth Table: Game State Management

Users	Game State Size	Server Load	Latency	Storage Needs
100 users	Small (few MBs)	Single server handles all	Low (real-time)	Minimal, local storage
10,000 users	Medium (GBs)	Multiple servers, load balancer	Low to medium	Distributed cache + DB
1,000,000 users	Large (TBs)	Cluster of servers, sharded DB	Medium (optimized)	Sharded DB + caching layers
100,000,000 users	Very large (PBs)	Massive clusters, global distribution	Medium to high (edge caching)	Multi-region DB, archival storage

First Bottleneck

The first bottleneck is the database that stores and retrieves game states. As user count grows, the number of read/write operations increases rapidly. A single database instance can handle only a limited number of queries per second (QPS), typically up to 5,000-10,000 QPS for a relational DB. Beyond this, latency increases and requests queue up, causing delays in game state updates and retrievals.

Scaling Solutions

• Read Replicas: Use read replicas to distribute read queries and reduce load on the primary database.
• Caching: Implement in-memory caches (e.g., Redis) for frequently accessed game states to reduce DB hits.
• Sharding: Partition the database by user ID or game session to spread load across multiple DB instances.
• Horizontal Scaling: Add more application servers behind a load balancer to handle more concurrent connections.
• Eventual Consistency: Use asynchronous updates where strict real-time consistency is not critical to reduce DB write pressure.
• Edge Caching: For global users, cache game state snapshots closer to users to reduce latency.

Back-of-Envelope Cost Analysis

Assuming 1 million concurrent users, each sending 1 state update per second:

• Requests per second: ~1,000,000 QPS (too high for single DB)
• Storage: If each game state is 10 KB, total active data ~10 GB in memory/cache; historical data grows daily.
• Bandwidth: 1,000,000 updates * 10 KB = ~10 GB/s (requires high network capacity)

This shows the need for sharding, caching, and horizontal scaling to handle load and bandwidth.

Interview Tip

Start by explaining the components involved: game clients, servers, database, and cache. Discuss how game state updates flow and where bottlenecks appear as users grow. Then, propose scaling strategies step-by-step, focusing on database scaling first, followed by application and network layers. Use real numbers to justify your choices and show understanding of trade-offs.

Self Check Question

Your database handles 1,000 QPS. Traffic grows 10x to 10,000 QPS. What do you do first?

Answer: Add read replicas and implement caching to reduce direct database load before considering sharding or adding more servers.

Key Result

Game state management first breaks at the database layer due to high read/write load; scaling requires caching, read replicas, and sharding to maintain low latency and handle millions of users.

Practice

1. What is the main purpose of game state management in a video game?

easy

A. To handle the sound effects and music

B. To keep track of what is happening in the game and control transitions between different screens or modes

C. To improve the graphics quality of the game

D. To manage the player's score only

Solution

1. Step 1: Understand the role of game state management

   Game state management is about tracking the current status of the game, such as menus, playing, or paused states.

2. Step 2: Identify the correct purpose

   It controls how the game moves between these states and keeps the game organized and less buggy.

3. Final Answer:

   To keep track of what is happening in the game and control transitions between different screens or modes -> Option B

4. Quick Check:

   Game state management = Track and control game modes [OK]

Hint: Game state manages screens and modes, not graphics or sound [OK]

Common Mistakes:

• Confusing game state with graphics or sound management
• Thinking it only manages scores
• Assuming it handles player input directly

2. Which of the following is the correct way to represent a simple game state using an enum in a low-level design?

easy

A. enum GameState { MENU, PLAYING, PAUSED, GAME_OVER }

B. class GameState { int MENU = 1; int PLAYING = 2; int PAUSED = 3; int GAME_OVER = 4; }

C. var GameState = ['MENU', 'PLAYING', 'PAUSED', 'GAME_OVER']

D. GameState = { MENU: 1, PLAYING: 2, PAUSED: 3, GAME_OVER: 4 }

Solution

1. Step 1: Identify enum syntax for game states

   Enums are used to define a fixed set of named constants, perfect for game states.

2. Step 2: Check which option uses enum correctly

   enum GameState { MENU, PLAYING, PAUSED, GAME_OVER } uses enum syntax correctly to define game states clearly and safely.

3. Final Answer:

   enum GameState { MENU, PLAYING, PAUSED, GAME_OVER } -> Option A

4. Quick Check:

   Enum syntax for states = enum GameState { MENU, PLAYING, PAUSED, GAME_OVER } [OK]

Hint: Enums clearly name fixed states, use enum keyword [OK]

Common Mistakes:

• Using arrays or objects instead of enums for fixed states
• Defining states as class variables without enum
• Mixing syntax from different languages

3. Given this pseudocode for a game state manager, what will be the output after calling changeState('PAUSED') twice?

class GameStateManager:
  def __init__(self):
    self.state = 'MENU'
  def changeState(self, new_state):
    if self.state != new_state:
      self.state = new_state
      print(f'State changed to {self.state}')
    else:
      print(f'State already {self.state}')

manager = GameStateManager()
manager.changeState('PAUSED')
manager.changeState('PAUSED')

medium

A. State already PAUSED
State already PAUSED

B. State changed to PAUSED
State changed to PAUSED

C. State changed to PAUSED
State already PAUSED

D. State changed to MENU
State changed to PAUSED

Solution

1. Step 1: Analyze first changeState call

   Initial state is 'MENU'. Changing to 'PAUSED' triggers state change and prints 'State changed to PAUSED'.

2. Step 2: Analyze second changeState call

   State is already 'PAUSED', so it prints 'State already PAUSED' without changing.

3. Final Answer:

   State changed to PAUSED State already PAUSED -> Option C

4. Quick Check:

   Second call same state = no change message [OK]

Hint: Second same state call prints 'already' message [OK]

Common Mistakes:

• Assuming state changes again on same value
• Ignoring else branch output
• Confusing initial state with changed state

4. In the following code snippet, what is the main bug that can cause incorrect game state transitions?

class GameStateManager:
  def __init__(self):
    self.state = 'MENU'
  def changeState(self, new_state):
    if self.state == new_state:
      self.state = new_state
      print(f'State changed to {self.state}')
    else:
      print(f'State already {self.state}')

medium

A. The method does not accept new_state parameter

B. The print statements are swapped

C. The initial state is not set properly

D. The condition is reversed; it changes state only if states are equal

Solution

1. Step 1: Review the if condition logic

   The code changes state only if current state equals new_state, which is wrong because state should change when states differ.

2. Step 2: Identify correct condition

   The condition should be if current state != new_state to update state and print change message.

3. Final Answer:

   The condition is reversed; it changes state only if states are equal -> Option D

4. Quick Check:

   State change condition reversed = bug [OK]

Hint: Check if condition matches when states differ, not equal [OK]

Common Mistakes:

• Not noticing reversed if condition
• Assuming print statements cause bug
• Ignoring initial state setup

5. You are designing a multiplayer game with complex states like LOBBY, MATCHMAKING, IN_GAME, PAUSED, and GAME_OVER. Which approach best supports scalability and easy state transitions for many players?

hard

A. Use a centralized state manager with a state machine pattern and event-driven updates per player

B. Store each player's state in a simple variable and update it directly without structure

C. Use global variables for all states and check them in every game loop iteration

D. Hardcode state transitions inside each player's input handler

Solution

1. Step 1: Understand scalability needs

   Many players and complex states require organized, scalable management to avoid bugs and support concurrency.

2. Step 2: Evaluate approaches

   A centralized state manager using a state machine and event-driven updates cleanly handles transitions and scales well.

3. Final Answer:

   Use a centralized state manager with a state machine pattern and event-driven updates per player -> Option A

4. Quick Check:

   Centralized state machine + events = scalable design [OK]

Hint: Centralized state machine with events scales best [OK]

Common Mistakes:

• Using global variables causing race conditions
• Hardcoding transitions making maintenance hard
• No structure causing bugs with many players