System Design Notes All designs

Fundamentals

Concepts & Technologies Used

Every technology and concept below is used somewhere on this site — this index is generated by scanning the content of all 41 pages, not pulled from outside. 120 concepts are catalogued; each row links to the designs that use it.

Datastores & storage

Concept / technology#Used in
Redis 16 Data Structures for System Design, Database Indexing, Bit.ly — URL Shortener, Web Crawler, LeetCode — Online Judge, YouTube — Video Platform, Facebook News Feed, Twitter / X Timeline, Facebook Post Search, WhatsApp — Chat & Messaging, Tinder — Matching & Swipes, Uber — Ride Hailing, Yelp — Local Search & Reviews, Ticketmaster — Ticketing, Online Auction, Ad Click Aggregator
Memcached 2 Data Structures for System Design, Database Indexing
Cassandra 5 Data Structures for System Design, Database Indexing, Bit.ly — URL Shortener, Facebook Live Comments, Tinder — Matching & Swipes
ScyllaDB 2 Data Structures for System Design, Database Indexing
DynamoDB 10 Interview Toolkit & Cheat Sheet, Data Structures for System Design, Database Indexing, Bit.ly — URL Shortener, Web Crawler, Dropbox — File Storage & Sync, Facebook News Feed, WhatsApp — Chat & Messaging, Uber — Ride Hailing, Online Auction
PostgreSQL 11 Interview Toolkit & Cheat Sheet, Data Structures for System Design, Database Indexing, Bit.ly — URL Shortener, YouTube — Video Platform, Facebook Post Search, Tinder — Matching & Swipes, Yelp — Local Search & Reviews, Ticketmaster — Ticketing, Online Auction, Experiment Tracking & HPO
PostGIS 4 Interview Toolkit & Cheat Sheet, Data Structures for System Design, Database Indexing, Yelp — Local Search & Reviews
MySQL / InnoDB 3 Data Structures for System Design, Database Indexing, Bit.ly — URL Shortener
SQLite 1 Data Structures for System Design
MongoDB 2 Data Structures for System Design, Database Indexing
HBase 2 Data Structures for System Design, Database Indexing
Bigtable 1 Data Structures for System Design
RocksDB 2 Data Structures for System Design, Database Indexing
LevelDB 2 Data Structures for System Design, Database Indexing
ClickHouse 1 Database Indexing
Druid 1 Data Structures for System Design
Spanner 1 Online Auction
BigQuery 2 Data Structures for System Design, Database Indexing
Snowflake (warehouse) 1 Database Indexing
S3 / object storage 20 Kafka Deep Dive, Web Crawler, Dropbox — File Storage & Sync, YouTube — Video Platform, Twitter / X Timeline, Facebook Post Search, WhatsApp — Chat & Messaging, Yelp — Local Search & Reviews, Ad Click Aggregator, Distributed GPU Training Cluster, LLM Inference Serving Platform, Multi-Tenant GPU Cluster, Distributed Checkpointing, Model Registry & Artifact Store, Petabyte ML Data Pipeline, GPU Workload Observability, ML-Optimized Distributed File System, Experiment Tracking & HPO, CUDA Kernel Profiling & Auto-Tuning, Multi-Datacenter Model Serving
Glacier / cold storage 5 YouTube — Video Platform, Facebook Post Search, Distributed Checkpointing, ML-Optimized Distributed File System, Experiment Tracking & HPO
OLAP store 2 Database Indexing, Ad Click Aggregator

Messaging, streaming & compute

Coordination, infra & protocols

Concept / technology#Used in
ZooKeeper 5 Kafka Deep Dive, YouTube Top-K Videos, Facebook Live Comments, WhatsApp — Chat & Messaging, GPU Cluster Scheduler
etcd 2 Facebook Live Comments, GPU Cluster Scheduler
KRaft 1 Kafka Deep Dive
CDN 12 Bit.ly — URL Shortener, LeetCode — Online Judge, Dropbox — File Storage & Sync, YouTube — Video Platform, Facebook News Feed, Twitter / X Timeline, Facebook Post Search, Yelp — Local Search & Reviews, Ticketmaster — Ticketing, Model Registry & Artifact Store, Experiment Tracking & HPO, Multi-Datacenter Model Serving
WebSocket 4 LeetCode — Online Judge, Facebook Live Comments, WhatsApp — Chat & Messaging, Online Auction
Server-Sent Events (SSE) 5 LeetCode — Online Judge, Facebook Live Comments, Online Auction, LLM Inference Serving Platform, Continuous Batching Inference Engine
Long polling 1 Facebook Live Comments
API gateway 15 Bit.ly — URL Shortener, LeetCode — Online Judge, Dropbox — File Storage & Sync, YouTube — Video Platform, Facebook News Feed, Twitter / X Timeline, Facebook Live Comments, Facebook Post Search, Tinder — Matching & Swipes, Uber — Ride Hailing, Yelp — Local Search & Reviews, Ticketmaster — Ticketing, Online Auction, Ad Click Aggregator, LLM Inference Serving Platform
Load balancer 6 YouTube Top-K Videos, Twitter / X Timeline, WhatsApp — Chat & Messaging, LLM Inference Serving Platform, Continuous Batching Inference Engine, Multi-Datacenter Model Serving
Docker containers 1 LeetCode — Online Judge
gVisor 1 LeetCode — Online Judge
Firecracker microVM 1 LeetCode — Online Judge
seccomp / cgroups 2 LeetCode — Online Judge, Multi-Tenant GPU Cluster
Stripe (payments) 3 Ticketmaster — Ticketing, Distributed Checkpointing, ML-Optimized Distributed File System
Push (APNs / FCM) 3 WhatsApp — Chat & Messaging, Tinder — Matching & Swipes, Uber — Ride Hailing
Presigned URLs 5 Dropbox — File Storage & Sync, YouTube — Video Platform, WhatsApp — Chat & Messaging, Model Registry & Artifact Store, Experiment Tracking & HPO

Data structures & sketches

Patterns & techniques

Concept / technology#Used in
CAP theorem 1 Interview Toolkit & Cheat Sheet
PACELC 1 Interview Toolkit & Cheat Sheet
Fan-out (write vs read) 14 Data Structures for System Design, Database Indexing, Kafka Deep Dive, Web Crawler, LeetCode — Online Judge, Facebook News Feed, Twitter / X Timeline, Facebook Live Comments, Facebook Post Search, WhatsApp — Chat & Messaging, Online Auction, Model Registry & Artifact Store, Experiment Tracking & HPO, Multi-Datacenter Model Serving
Sharding / partitioning 36 Interview Toolkit & Cheat Sheet, Data Structures for System Design, Kafka Deep Dive, Behavioral Questions & Answers, Bit.ly — URL Shortener, Web Crawler, LeetCode — Online Judge, Dropbox — File Storage & Sync, YouTube — Video Platform, YouTube Top-K Videos, Facebook News Feed, Twitter / X Timeline, Facebook Live Comments, Facebook Post Search, WhatsApp — Chat & Messaging, Tinder — Matching & Swipes, Uber — Ride Hailing, Yelp — Local Search & Reviews, Ticketmaster — Ticketing, Online Auction, Ad Click Aggregator, Distributed GPU Training Cluster, LLM Inference Serving Platform, GPU Cluster Scheduler, Continuous Batching Inference Engine, Multi-Tenant GPU Cluster, Distributed Checkpointing, GPU Health & Failure Detection, Petabyte ML Data Pipeline, Distributed Parameter Server, GPU Workload Observability, ML-Optimized Distributed File System, GPU Cluster Network Topology, Experiment Tracking & HPO, Disaggregated GPU Memory, Multi-Datacenter Model Serving
Replication 16 Interview Toolkit & Cheat Sheet, Data Structures for System Design, Kafka Deep Dive, Bit.ly — URL Shortener, YouTube Top-K Videos, Twitter / X Timeline, Online Auction, Distributed GPU Training Cluster, GPU Cluster Scheduler, Distributed Checkpointing, Model Registry & Artifact Store, Petabyte ML Data Pipeline, Distributed Parameter Server, ML-Optimized Distributed File System, Disaggregated GPU Memory, Multi-Datacenter Model Serving
Quorum 2 Tinder — Matching & Swipes, Online Auction
In-sync replicas (ISR) 1 Kafka Deep Dive
Leader / follower 1 Kafka Deep Dive
Idempotency 17 Interview Toolkit & Cheat Sheet, Kafka Deep Dive, Behavioral Interview Framework, Behavioral Questions & Answers, Web Crawler, LeetCode — Online Judge, YouTube — Video Platform, Facebook News Feed, Facebook Post Search, WhatsApp — Chat & Messaging, Tinder — Matching & Swipes, Yelp — Local Search & Reviews, Ticketmaster — Ticketing, Online Auction, Ad Click Aggregator, Model Registry & Artifact Store, Experiment Tracking & HPO
Write-ahead log (WAL) 3 Data Structures for System Design, Database Indexing, Tinder — Matching & Swipes
MVCC 2 Interview Toolkit & Cheat Sheet, YouTube — Video Platform
TTL (auto-expiry) 16 Interview Toolkit & Cheat Sheet, Kafka Deep Dive, Bit.ly — URL Shortener, LeetCode — Online Judge, YouTube Top-K Videos, Facebook News Feed, Twitter / X Timeline, Facebook Post Search, WhatsApp — Chat & Messaging, Tinder — Matching & Swipes, Uber — Ride Hailing, Ticketmaster — Ticketing, Ad Click Aggregator, Model Registry & Artifact Store, Experiment Tracking & HPO, Multi-Datacenter Model Serving
Conditional write / CAS 5 Interview Toolkit & Cheat Sheet, Uber — Ride Hailing, Ticketmaster — Ticketing, Online Auction, GPU Cluster Scheduler
Distributed lock 2 Uber — Ride Hailing, Ticketmaster — Ticketing
Virtual waiting queue 3 Kafka Deep Dive, Ticketmaster — Ticketing, Continuous Batching Inference Engine
Saga pattern 2 Tinder — Matching & Swipes, Ticketmaster — Ticketing
Kappa architecture 1 Ad Click Aggregator
Scatter-gather 2 Facebook News Feed, Facebook Post Search
Rate limiting 10 Bit.ly — URL Shortener, Web Crawler, LeetCode — Online Judge, Dropbox — File Storage & Sync, Twitter / X Timeline, Tinder — Matching & Swipes, Uber — Ride Hailing, Ticketmaster — Ticketing, LLM Inference Serving Platform, Multi-Datacenter Model Serving
Chunking 6 Dropbox — File Storage & Sync, YouTube — Video Platform, LLM Inference Serving Platform, Continuous Batching Inference Engine, Model Registry & Artifact Store, GPU Workload Observability
Deduplication 14 Data Structures for System Design, Kafka Deep Dive, Bit.ly — URL Shortener, Web Crawler, Dropbox — File Storage & Sync, Facebook Live Comments, WhatsApp — Chat & Messaging, Online Auction, Ad Click Aggregator, Model Registry & Artifact Store, GPU Health & Failure Detection, Petabyte ML Data Pipeline, Distributed Parameter Server, Experiment Tracking & HPO
Read-through / write-through cache 2 Bit.ly — URL Shortener, Online Auction
Backpressure 5 LeetCode — Online Judge, Facebook Post Search, Ticketmaster — Ticketing, Petabyte ML Data Pipeline, Multi-Datacenter Model Serving
Delivery semantics (exactly/at-least-once) 7 Kafka Deep Dive, Web Crawler, Facebook Live Comments, WhatsApp — Chat & Messaging, Tinder — Matching & Swipes, Online Auction, Ad Click Aggregator
Geospatial indexing 3 Database Indexing, Uber — Ride Hailing, Yelp — Local Search & Reviews
Adaptive bitrate (HLS / DASH) 4 YouTube — Video Platform, Ad Click Aggregator, GPU Health & Failure Detection, CUDA Kernel Profiling & Auto-Tuning
Transcoding pipeline 2 Kafka Deep Dive, YouTube — Video Platform
Short codes (base62 / sqids / Hashids) 1 Bit.ly — URL Shortener
Time-sortable IDs (Snowflake / KSUID) 1 WhatsApp — Chat & Messaging
UUID 3 Facebook Live Comments, Online Auction, Multi-Tenant GPU Cluster

GPU / ML infrastructure

Concept / technology#Used in
NCCL collectives 5 Distributed GPU Training Cluster, GPU Health & Failure Detection, Distributed Parameter Server, GPU Workload Observability, GPU Cluster Network Topology
InfiniBand 5 Distributed GPU Training Cluster, Multi-Tenant GPU Cluster, Distributed Checkpointing, GPU Cluster Network Topology, Disaggregated GPU Memory
NVLink / NVSwitch 9 Distributed GPU Training Cluster, LLM Inference Serving Platform, GPU Cluster Scheduler, LLM KV-Cache Management, Multi-Tenant GPU Cluster, GPU Health & Failure Detection, GPU Workload Observability, GPU Cluster Network Topology, Disaggregated GPU Memory
RDMA 3 Multi-Tenant GPU Cluster, GPU Cluster Network Topology, Disaggregated GPU Memory
RoCE 3 Multi-Tenant GPU Cluster, GPU Cluster Network Topology, Disaggregated GPU Memory
CXL 1 Disaggregated GPU Memory
Multi-Instance GPU (MIG) 2 GPU Cluster Scheduler, Multi-Tenant GPU Cluster
MPS (time-slicing) 1 Multi-Tenant GPU Cluster
CUDA 8 Distributed GPU Training Cluster, Continuous Batching Inference Engine, Distributed Checkpointing, Petabyte ML Data Pipeline, GPU Workload Observability, Experiment Tracking & HPO, CUDA Kernel Profiling & Auto-Tuning, Disaggregated GPU Memory
Nsight / CUPTI 4 WhatsApp — Chat & Messaging, Continuous Batching Inference Engine, GPU Workload Observability, CUDA Kernel Profiling & Auto-Tuning
DCGM 3 Multi-Tenant GPU Cluster, GPU Health & Failure Detection, GPU Workload Observability
Triton 2 Continuous Batching Inference Engine, CUDA Kernel Profiling & Auto-Tuning
vLLM 3 LLM Inference Serving Platform, LLM KV-Cache Management, Continuous Batching Inference Engine
PagedAttention 3 LLM Inference Serving Platform, LLM KV-Cache Management, Continuous Batching Inference Engine
KV cache 5 LLM Inference Serving Platform, LLM KV-Cache Management, Continuous Batching Inference Engine, Disaggregated GPU Memory, Multi-Datacenter Model Serving
Continuous batching 3 LLM Inference Serving Platform, LLM KV-Cache Management, Continuous Batching Inference Engine
Data / tensor / pipeline parallelism 4 Distributed GPU Training Cluster, LLM Inference Serving Platform, LLM KV-Cache Management, Continuous Batching Inference Engine
All-reduce 10 Distributed GPU Training Cluster, LLM Inference Serving Platform, GPU Cluster Scheduler, Multi-Tenant GPU Cluster, GPU Health & Failure Detection, Petabyte ML Data Pipeline, Distributed Parameter Server, GPU Workload Observability, GPU Cluster Network Topology, CUDA Kernel Profiling & Auto-Tuning
FSDP / ZeRO 3 Distributed GPU Training Cluster, Distributed Parameter Server, Disaggregated GPU Memory
Fat-tree / Dragonfly 2 Model Registry & Artifact Store, GPU Cluster Network Topology
Roofline model 1 CUDA Kernel Profiling & Auto-Tuning
Model FLOPs Utilization (MFU) 5 Distributed GPU Training Cluster, LLM Inference Serving Platform, Distributed Checkpointing, GPU Workload Observability, GPU Cluster Network Topology