System Design Notes

Personal reference notes on distributed systems, data structures, and Java internals.

---

Folder Structure

system-design-notes/
├── README.md
│
├── design/                           — distributed systems & architecture topics
│   ├── 01-scalability-fundamentals.md
│   ├── 02-cap-theorem.md
│   ├── 03-consistent-hashing.md
│   ├── 04-caching.md
│   ├── 05-load-balancing.md
│   ├── 06-database-sharding.md
│   ├── 07-replication.md
│   ├── 10-redis.md
│   ├── 11-message-queues.md
│   ├── 12-graph-databases.md
│   ├── 13-search-engines.md
│   ├── 14-rate-limiting.md
│   ├── 15-quorum-and-consensus.md
│   ├── 16-leader-election-gossip.md
│   ├── 17-circuit-breaker.md
│   ├── 18-api-gateway-service-discovery.md
│   ├── 19-sql-vs-nosql-indexes.md
│   ├── 20-distributed-transactions.md
│   ├── 21-back-of-envelope-cdn.md
│   ├── 22-leases.md
│   ├── 23-stream-processing.md
│   ├── 24-observability-operations.md
│   ├── 25-apache-iceberg-data-lake.md
│   ├── 26-netflix-data-platform.md
│   ├── 27-infrastructure-k8s.md
│   ├── 28-java-sizes-boe.md          — primitive/object/collection sizes for back-of-envelope
│   └── 29-system-design-questions.md — common design questions with thought process + solutions
│
├── data-structures/                  — internals: how storage engines work
│   ├── 08-bloom-filters.md
│   ├── 09-probabilistic-data-structures.md
│   ├── 28-skip-lists.md
│   ├── 29-lsm-trees.md
│   ├── 30-b-trees.md
│   └── 31-red-black-trees.md
│
└── java/                             — Java language deep dives
    ├── 01-concurrency.md             — CountDownLatch, CyclicBarrier, Semaphore, ExecutorService, Virtual Threads, CompletableFuture
    ├── 02-message-brokers.md         — Redis Streams vs Kafka vs Kinesis vs SQS: mental models, when to use which
    └── 03-redis-commands-cheatsheet.md — all Redis commands by type: strings, hashes, sets, ZSETs, streams, bitmaps, HLL, Lua

---

Design Topics

Priority	File	Topic
Core	design/02-cap-theorem.md	Consistency vs availability tradeoffs
Core	design/03-consistent-hashing.md	Sharding, KV stores, Kafka partitioning
Core	design/23-stream-processing.md	Kafka, Flink, exactly-once semantics
Core	design/04-caching.md	Cache strategies, Redis vs Memcached, EVCache
Core	design/15-quorum-and-consensus.md	Replication, consistency models
Core	design/20-distributed-transactions.md	Saga, 2PC, workflow orchestration
Deep	design/25-apache-iceberg-data-lake.md	Table format, ACID on object storage
Deep	design/26-netflix-data-platform.md	Kafka, Flink, Iceberg, EVCache, Maestro
Deep	design/22-leases.md	Partition ownership, fencing tokens
Supporting	design/07-replication.md	Leader-follower vs leaderless
Supporting	design/27-infrastructure-k8s.md	Kubernetes, service mesh

---

Data Structures

Know well enough to explain and compare:

File	When It Comes Up
data-structures/29-lsm-trees.md	RocksDB, Cassandra, DynamoDB internals
data-structures/30-b-trees.md	Database index internals
data-structures/28-skip-lists.md	Redis ZSET, ConcurrentSkipListMap, MemTable
data-structures/31-red-black-trees.md	TreeMap, why skiplist wins for concurrency
data-structures/08-bloom-filters.md	SSTable reads, membership checks
data-structures/09-probabilistic-data-structures.md	HyperLogLog, Count-Min Sketch

---

Quick Reference — When to Use What

Need	Use
Fast reads, reduce DB load	Cache (Redis / EVCache)
Distribute load across nodes	Consistent Hashing
Approximate membership test	Bloom Filter
Count unique items (approx)	HyperLogLog
Sorted map, concurrent	ConcurrentSkipListMap
Write-heavy storage engine	LSM Tree (RocksDB, Cassandra)
Read-heavy storage engine	B+ Tree (PostgreSQL, MySQL)
Async decoupling at scale	Kafka
Stateful stream processing	Flink (keyed state + checkpoints)
Exactly-once semantics	Kafka transactions + Flink checkpoints
ACID on object storage	Apache Iceberg
Distributed coordination	ZooKeeper / etcd
Handle partial failures	Circuit Breaker (Hystrix)
Partition ownership	Leases + fencing tokens
Producer-consumer backpressure	ArrayBlockingQueue + Virtual Threads