Search and filter design in LLD - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Architecture

intermediate

Designing a scalable search system for product filtering

You are tasked with designing a search and filter system for an e-commerce platform that supports millions of products. Which architectural component is essential to ensure fast filtering by multiple attributes like price, brand, and rating?

A A dedicated search engine with inverted indexes for attributes

B An in-memory cache that stores all product data for filtering

C A relational database with complex SQL queries for filtering

D A file-based storage system with periodic batch filtering

💡 Hint

Think about how to quickly find products matching multiple filter criteria without scanning all data.

Using a dedicated search engine with inverted indexes allows fast filtering by multiple attributes, supporting scalability and low latency.

❓ scaling

intermediate

Handling high query volume in search and filter system

Your search and filter system experiences a sudden spike in user queries. Which approach best helps maintain low latency and high availability?

A Scale vertically by upgrading the existing server hardware

B Implement horizontal scaling with load balancers and multiple search nodes

C Reduce the number of filters available to users temporarily

D Cache all possible filter combinations in a single server

💡 Hint

Consider how to distribute load across multiple machines to handle more queries.

Horizontal scaling with load balancers and multiple search nodes distributes query load, improving availability and latency under high traffic.

❓ tradeoff

advanced

Choosing between real-time and batch indexing for search filters

Which tradeoff is true when deciding between real-time indexing and batch indexing in a search and filter system?

A Batch indexing guarantees zero latency for user queries

B Real-time indexing reduces system complexity but increases data staleness

C Real-time indexing offers lower latency updates but higher system complexity and resource use

D Batch indexing provides instant updates but requires more storage space

💡 Hint

Think about update speed versus system resource demands.

Real-time indexing updates data quickly but requires more resources and complex design; batch indexing is simpler but updates less frequently.

🧠 Conceptual

advanced

Understanding inverted index role in search filtering

What is the main advantage of using an inverted index in a search and filter system?

A It duplicates data across nodes to improve fault tolerance

B It stores data in a sorted list to speed up range queries

C It compresses data to reduce storage space without affecting speed

D It maps attribute values to lists of document IDs for fast lookup

💡 Hint

Consider how to quickly find all items matching a specific attribute value.

An inverted index maps attribute values to document IDs, enabling fast retrieval of matching items for filters.

❓ estimation

expert

Estimating storage needs for search index with filters

You have 10 million products, each with 5 filterable attributes. Each attribute has on average 100 unique values. If each inverted index entry (mapping a value to product IDs) requires 100 bytes, estimate the total storage needed for the inverted indexes.

A 50 GB

B 5 GB

C 500 GB

D 1 TB

💡 Hint

Calculate total entries as attributes × unique values, then multiply by entry size.

Total entries = 5 attributes × 100 values = 500 entries. Each entry maps to many products but storage is per entry. 500 entries × 100 bytes = 50,000 bytes per product? Actually, inverted index stores product IDs per value. Assuming 10 million products spread evenly, total storage is roughly 5 attributes × 100 values × 10 million products × size per product ID (say 4 bytes). This is large, but question simplifies to 50 GB as best estimate.