Question 1

What is consistent hashing?

Accepted Answer

**Target Audience:** Beginners to system design who know what a hash function and a cache are; no distributed-systems background assumed.
**Difficulty:** Medium
**Verified with:** Node 22.22, Python 3.12, Go 1.24 and Java 21 (all four print identical output). History is cited from Karger et al., STOC 1997; system usage from the Dynamo paper (SOSP 2007) and the Apache Cassandra documentation.

> **How to read this doc:** Every idea is explained in plain language first, then tagged with `📌 Interv

Question 2

When do you use a read replica vs a cache?

Accepted Answer

**Target Audience:** Beginners to system design — you know what a database and a query are; replication and caching are explained here.
**Difficulty:** Medium
**Verified with:** Node 22.22 and Python 3.12 (both print identical output).

> **How to read this doc:** Every idea is explained in plain language first, then tagged with `📌 Interview term:` so you also learn the exact words an interviewer expects. Behaviour claims (stale reads, counts) come from running the example; the **Verified** sec

Question 3

How would you design a 'nearby drivers' / proximity search?

Accepted Answer

**Target Audience:** Developers preparing for system design interviews. You know what a database index and a cache are; no maps or geometry background needed.
**Difficulty:** Hard
**Verified with:** Node 22.22, Python 3.12, Go 1.24 and Java 21 (all four print identical output). Redis behaviour and memory were checked on Redis 7.0.15.

> **How to read this doc:** Every idea is explained in plain language first, then tagged with `📌 Interview term:` so you also learn the exact words an interviewer

Question 4

How would you design a distributed key-value store / query cache for search?

Accepted Answer

## Step 1: Use Cases, Constraints, Assumptions, Back-of-Envelope

### Use cases
We'll scope to handle only the core flows that expose the bottlenecks; everything else is deferred.

- **User** searches — GET `/search?q=...`, cache lookup by normalized query hash, hit returns in <20ms, miss fetches search tier.
- **Service** caches query results — write-through on miss with TTL + jitter, LRU eviction, per-tenant isolation.
- **Service** invalidates — TTL expiry plus explicit invalidation on index

Question 5

How would you design a typeahead/autocomplete system?

Accepted Answer

**Target Audience:** Engineers preparing for a system design round; you know what a server and a database are, and data structures like tries are explained from scratch.
**Difficulty:** Hard
**Verified with:** Node 22.22, Python 3.12, Go 1.24 and Java 21 (all four print identical output). The capacity figures are worked arithmetic from stated assumptions, not measurements of a real service.

> **How to read this doc:** Every idea is explained in plain language first, then tagged with `📌 Intervi

Question 6

How would you design Instagram?

Accepted Answer

**Target Audience:** Engineers preparing for a system design round. You know what a server, a database and HTTP are; storage, CDN and sharding ideas are defined as we go.
**Difficulty:** Hard
**Verified with:** Node 22.22, Python 3.12, Go 1.24 and Java 21 (all four print identical output). Instagram engineering facts are cited inline, checked in October 2026.

> **How to read this doc:** Plain language first, then `📌 Interview term:` callouts. Numbers come from the example program and are quote

Question 7

Consistent hashing vs rendezvous (HRW) hashing — which for what scale?

Accepted Answer

**Target Audience:** Readers who know the basic hash ring and virtual nodes; see [What is consistent hashing?](/interview-question/what-is-consistent-hashing) and [What is consistent hashing with virtual nodes and when do you need it?](/interview-question/what-is-consistent-hashing-with-virtual-nodes-and-when-do-you-need-it) if not.
**Difficulty:** Hard
**Verified with:** Node 22.22, Python 3.12, Go 1.24 and Java 21 (all four print identical output). History is cited from Thaler and Ravishankar

Question 8

How does CDC with Debezium keep search and cache in sync with Postgres?

Accepted Answer

**Target Audience:** Back-end and platform engineers who keep derived stores (search indexes, caches, analytics) in sync with a primary database. You know what Postgres, Kafka and a cache are; CDC is explained from zero.
**Difficulty:** Hard
**Verified with:** Node 22.22 and Python 3.12 (both print identical output). Debezium and Postgres behaviour is from their documentation, cited inline; it was not run here.

> **How to read this doc:** Every idea is explained in plain language first, then ta

Question 9

How would you design Twitter's timeline and search? (Facebook feed)

Accepted Answer

## Step 1: Use Cases, Constraints, Assumptions, Back-of-Envelope

### Use cases
We'll scope to handle only the core flows that expose the bottlenecks; everything else is deferred.

- **User** posts a tweet — 280 chars + media refs, persisted, then fanned out to followers' timelines.
- **User** reads home timeline — GET merged, ranked feed (paginated, <200ms p95).
- **Service** searches tweets — inverted index over tokens/hashtags with ranking and recency.
- **Service** fans out / backfills — asy

Question 10

How would you design Google Drive?

Accepted Answer

**Target Audience:** Developers preparing for system design interviews. This page builds on the [Dropbox design](/interview-question/how-would-you-design-a-file-storage-service-like-dropbox) (block sync) and the [S3-like object storage design](/interview-question/how-would-you-design-s3-like-object-storage-572x) (blob storage) and focuses on what is different: an ID-based shared tree, permissions, the change feed and uploads.
**Difficulty:** Hard
**Verified with:** Node 22.22 and Python 3.12 (bo

SYSTEM DESIGNDECODED.

Medium