Shard key selection in HLD - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is a shard key in distributed databases?

A shard key is a specific field or set of fields used to divide data into smaller parts called shards. It helps distribute data across multiple servers to improve performance and scalability.

beginner

Why is choosing the right shard key important?

The right shard key ensures even data distribution, reduces hotspots, and improves query performance. A poor choice can cause uneven load and slow down the system.

intermediate

What are common criteria for selecting a shard key?

Common criteria include high cardinality (many unique values), frequent use in queries, even data distribution, and low update frequency to avoid moving data between shards.

intermediate

Explain the impact of a low-cardinality shard key.

A low-cardinality shard key has few unique values, causing data to cluster in few shards. This leads to uneven load, hotspots, and poor scalability.

intermediate

What is a hotspot in sharding, and how does shard key selection affect it?

A hotspot is a shard that receives a disproportionate amount of traffic or data. Choosing a shard key that evenly distributes data and queries helps prevent hotspots.

Which property is most important for a shard key to ensure even data distribution?

A High cardinality

B Low update frequency

C Used in rare queries

D Small data size

High cardinality means many unique values, which helps distribute data evenly across shards.

What happens if a shard key causes data to cluster on a few shards?

A Improved query speed

B Hotspots and uneven load

C Automatic data balancing

D Reduced storage needs

Clustering data on few shards creates hotspots, causing uneven load and performance issues.

Which shard key characteristic helps reduce data movement between shards?

A Used only in writes

B High update frequency

C Low update frequency

D Random values

Low update frequency means shard key values rarely change, reducing data movement between shards.

Why should a shard key be frequently used in queries?

A To avoid data partitioning

B To increase data duplication

C To reduce shard count

D To improve query routing and performance

Using the shard key in queries helps route queries directly to relevant shards, improving performance.

Which of these is NOT a good shard key choice?

A Country code with 3-4 unique values

B User ID with millions of unique users

C Order ID with unique values

D Session token with high uniqueness

Country code has low cardinality, causing uneven data distribution and hotspots.

Explain how to select a shard key for a large user database.

Think about what uniquely identifies users and how queries are made.

Describe the consequences of a poor shard key choice in a distributed system.

Consider what happens when data is not balanced across servers.