Write sharding in DynamoDB - Mini Project: Build & Apply

Write Sharding in DynamoDB

📖 Scenario: You are building a high-traffic web application that stores user activity logs in DynamoDB. To handle many simultaneous writes without slowing down, you want to distribute the writes evenly across multiple partitions using write sharding.

🎯 Goal: Create a DynamoDB table schema and implement write sharding by assigning shard IDs to distribute writes evenly. You will set up the data structure, configure the number of shards, write the logic to assign shard IDs, and finalize the table key design.

📋 What You'll Learn

Create a DynamoDB table schema with a partition key and sort key

Define a variable for the number of shards

Write logic to assign a shard ID to each write based on user ID

Complete the table key design to include the shard ID for even write distribution

💡 Why This Matters

🌍 Real World

Write sharding helps large-scale applications handle many simultaneous writes by spreading data across multiple partitions, preventing bottlenecks.

💼 Career

Understanding write sharding is important for backend developers and database engineers working with NoSQL databases like DynamoDB to optimize performance and scalability.

1

Create DynamoDB table schema

Create a DynamoDB table schema with a partition key called ShardId (string) and a sort key called UserId (string). Define the table name as UserActivity.

DynamoDB

# Define DynamoDB table schema with partition key ShardId and sort key UserId
# Your code here

Need a hint?

Use ShardId as the partition key and UserId as the sort key in your table schema.

2

Define number of shards

Define a variable called num_shards and set it to 10 to represent the total number of shards for write distribution.

DynamoDB

TableName: UserActivity
KeySchema:
  - AttributeName: ShardId
    KeyType: HASH
  - AttributeName: UserId
    KeyType: RANGE
AttributeDefinitions:
  - AttributeName: ShardId
    AttributeType: S
  - AttributeName: UserId
    AttributeType: S

# Define number of shards
# Your code here

Need a hint?

Set num_shards to 10 to split writes into 10 partitions.

3

Assign shard ID for each write

Write a function called get_shard_id that takes a user_id string and returns a shard ID string by computing hash(user_id) % num_shards. The shard ID should be a string of the number.

DynamoDB

TableName: UserActivity
KeySchema:
  - AttributeName: ShardId
    KeyType: HASH
  - AttributeName: UserId
    KeyType: RANGE
AttributeDefinitions:
  - AttributeName: ShardId
    AttributeType: S
  - AttributeName: UserId
    AttributeType: S

num_shards = 10

# Define get_shard_id function
# Your code here

Need a hint?

Use the built-in hash() function and modulo operator to assign shard IDs.

4

Complete table key design with shard ID

Complete the write operation by using the ShardId from get_shard_id(user_id) as the partition key and UserId as the sort key in the item to write to DynamoDB.

DynamoDB

TableName: UserActivity
KeySchema:
  - AttributeName: ShardId
    KeyType: HASH
  - AttributeName: UserId
    KeyType: RANGE
AttributeDefinitions:
  - AttributeName: ShardId
    AttributeType: S
  - AttributeName: UserId
    AttributeType: S

num_shards = 10

def get_shard_id(user_id):
    return str(hash(user_id) % num_shards)

# Use get_shard_id to assign ShardId in the item to write
# Your code here

Need a hint?

Use the get_shard_id function to set the ShardId key in the item dictionary.