Tags

CDN

CDN

0 lessons
6 system designs
2 community items

cdn

System Design

6 articles
System Design

Blob Storage, Object Stores & CDNs

Databases are wrong for storing large unstructured files - photos, videos, backups, logs. Object stores like S3 give you cheap, durable, infinitely scalable storage for blobs, while CDNs cache that content at edges close to users. This lesson covers the object-storage data model, multi-part upload, storage classes, presigned URLs, and how a CDN turns a globally slow origin into a globally fast experience. By the end you can design the media layer for any social, video, or e-commerce system.

blob-storage
object-storage
cdn
content-delivery-network
s3
caching
system-design
intermediate

698

20

Medium
System Design

Design a URL Shortener (TinyURL)

Design a URL shortening service like TinyURL or bit.ly that maps a long URL to a 7 character code, redirects clicks in under 50 ms, and survives a 100:1 read-to-write ratio. This lesson walks through capacity estimation, the choice between counter based and hash based key generation, the database split between a key store and an analytics store, and the caching strategy that lets a single mid-tier service handle 10K redirects per second on commodity hardware.

design-tinyurl
url-shortener
case-study
social-content-platforms
base62-encoding
read-heavy
consistent-hashing
caching
cdn
system-design
beginner
free

690

4

Easy
System Design

Design Pastebin

Design a service like Pastebin or GitHub Gist where users dump up to 10 MB of text and share a link. The interview twist over a URL shortener: pastes are big, so you store them in object storage (S3) and only keep metadata in your database. This lesson covers the metadata vs blob split, expiration via S3 lifecycle policies, presigned URLs for direct uploads, syntax highlighting strategy, and how to handle the read pattern when most pastes are read once and never again.

design-pastebin
case-study
social-content-platforms
blob-storage
cdn
expiration-policy
presigned-urls
read-heavy
system-design
beginner
free

601

15

Easy
System Design

Design Instagram (Photo Sharing)

Design a photo sharing service like Instagram with 500M daily active users uploading 100M photos a day, served as personalized feeds at sub-200 ms p99. The interview centerpiece is the news feed: fan-out on write versus fan-out on read, the celebrity problem, and the hybrid pull-on-read model that real Instagram uses. We also cover photo upload pipelines (presigned URLs, multi-resolution generation, CDN), the metadata data model, and how to scale follow graphs that go from a few friends to hundreds of millions of followers.

design-instagram
case-study
social-content-platforms
photo-sharing
fan-out-on-write
fan-out-on-read
hybrid-fan-out
celebrity-problem
feed-ranking
media-storage
thumbnail-generation
cdn
social-media
system-design
intermediate
free

797

18

Medium
System Design

Design a Content Delivery Network

Design a Cloudflare/Akamai/Fastly-style content delivery network that offloads 95%+ of static traffic from origin servers, brings latency from hundreds of milliseconds down to single digits, and absorbs DDoS attacks at the edge. The interview centerpiece is the cache hierarchy and routing: hundreds of edge POPs anycast-routed to the user's nearest location, a regional shield layer that consolidates fetches, and the origin only seeing the long tail of misses. We cover cache key design with Vary headers, the TTL lifecycle and purge model, stale-while-revalidate for resilience under origin outages, and the moves CDNs make to keep dynamic content fast (programmable edge functions, smart routing).

design-cdn
case-study
infrastructure-storage
cdn
edge-caching
origin-shield
anycast
cache-invalidation
stale-while-revalidate
ddos-protection
system-design
intermediate
premium

865

15

Medium
System Design
Premium

Design Google Maps

Design Google Maps: a global mapping service that renders the Earth from 256x256 tiles, computes the shortest driving route in under 200 ms, and folds live traffic into routing for 1B users issuing 5B route requests per day. The interview centerpiece is the routing engine: how Dijkstra is too slow on a continent-scale graph and how Contraction Hierarchies (CH) precompute shortcuts so the live query is logarithmic. We cover the tile pyramid (zoom 0-20, ~1 trillion possible tiles at zoom 20), how live traffic from 100M Android phones updates edge weights every minute, and how to keep navigation latency under 1 second when re-routing.

design-google-maps
case-study
ride-sharing-and-maps
google-maps
graph-algorithms
dijkstra
a-star
contraction-hierarchies
routing-engine
map-tiles
tile-rendering
real-time-traffic
cdn
geospatial
h3-hex-grid
system-design
advanced
premium

584

10

Hard