System Design

System Design Advanced

Go from "I know what this is" to "I could implement, operate, and defend this in production." 16 conceptual-depth modules plus 20 technology deep-dives — Postgres, Kafka, Cassandra, Kubernetes, Flink, and more.

Advanced1764161 lessons17 free lessonssystem-designdistributed-systemsscalabilitydatabaseskafka

System Design Advanced: Production Depth & Technology Deep Dives

This is the depth course. It assumes you already have the System Design Fundamentals foundation (or equivalent experience) and takes you from recognising a concept to being able to implement it, operate it, and defend your decisions about it in a senior interview or a production incident review.

It splits into two halves:

Conceptual Depth (16 modules). The topics where a surface treatment isn't enough for production work — advanced capacity planning for tail latency, networking below the socket (TLS 1.3, HTTP/2 and /3, QUIC, kernel bypass), PACELC and linearizability vs causal consistency, FLP and production Raft, database internals (MVCC, WAL, storage engines), stream processing (event time, watermarks, exactly-once), ML systems design, edge computing, advanced observability and security, and the modern hardware layer.

Technology Deep Dives (20 modules). Each major production technology covered with the depth that lets you make defensible decisions and operate it for real: PostgreSQL, MySQL, MongoDB, Redis, Memcached, DynamoDB, Cassandra, Elasticsearch, Kafka, RabbitMQ, SQS, Flink, Spark, S3, Lambda, Nginx, ZooKeeper, Docker, Kubernetes, and Prometheus — history and design philosophy, architecture, the decisions that make each tool different, when to use it (and when not to), scaling patterns, and the operational realities that only show up in production.

The first module is free to preview. The full course is a one-time purchase with lifetime access.

What you will learn

• Capacity planning for tail latency, failure, and multi-region — not just the average case
• Networking at depth: TLS 1.3 internals, HTTP/2 & HTTP/3 / QUIC, anycast, kernel bypass
• The real distributed-systems theory: PACELC, linearizability, causal consistency, HLCs, consensus (Multi-Paxos, Raft, BFT)
• Database internals — storage engines, MVCC, WAL, distributed query execution, modern cloud DBs
• Stream processing done right: event time, watermarks, windowing, streaming joins, exactly-once
• ML systems, advanced architectural patterns (event sourcing, CQRS, sagas), search & recommendations
• Edge computing, advanced observability (SLOs, cardinality, tracing) and security (zero-trust, mTLS, threat modeling)
• Operate 20 production technologies with confidence — Postgres, Kafka, Cassandra, Kubernetes, Flink, S3, and more

Prerequisites

• System Design Fundamentals (or equivalent production experience)

Course outline

Module 1: Advanced Capacity Planning

Advanced Capacity Planning — 4 lessons.

• P99 Sizing vs Average Sizing 10 min read
• Capacity Planning for Failure 11 min read
• Multi-Region Capacity 12 min read
• Cost-Aware Capacity Planning 13 min read Pro lesson

Module 2: Advanced Networking

Advanced Networking — 5 lessons.

• TLS Internals at Depth 14 min read
• HTTP/2 Internals 12 min read Pro lesson
• HTTP/3 and QUIC 12 min read Pro lesson
• Anycast Routing in Production 11 min read Pro lesson
• Kernel Bypass and High-Performance Networking 16 min read Pro lesson

Module 3: Advanced Load Balancing

Advanced Load Balancing — 4 lessons.

• L4 vs L7 Deep Dive 12 min read Pro lesson
• Advanced Load Balancing Algorithms 14 min read Pro lesson
• Service Mesh Architecture 13 min read Pro lesson
• Tail Latency Mitigation 16 min read Pro lesson

Module 4: Advanced Consistency and Distribution

Advanced Consistency and Distribution — 6 lessons.

• PACELC Theorem - The Refinement of CAP 7 min read
• Linearizability in Depth 9 min read Pro lesson
• Sequential Consistency - A Weaker But Useful Model 9 min read Pro lesson
• Causal Consistency in Depth - The Often-Right Balance 9 min read Pro lesson
• Hybrid Logical Clocks - The Modern Combination of Physical and Logical Time 9 min read Pro lesson
• Multi-Leader Replication and Conflicts 11 min read Pro lesson

Module 5: Advanced Consensus

Advanced Consensus — 5 lessons.

• FLP Impossibility in Depth 12 min read
• Multi-Paxos 13 min read Pro lesson
• Byzantine Fault Tolerance 11 min read Pro lesson
• Production Raft 12 min read Pro lesson
• Distributed Locks in Practice 14 min read Pro lesson

Module 6: Advanced Caching Patterns

Advanced Caching Patterns — 4 lessons.

• Probabilistic Early Refresh 13 min read Pro lesson
• Hedged Caching 12 min read Pro lesson
• Negative Caching 13 min read Pro lesson
• Multi-Tier Cache Hierarchies 14 min read Pro lesson

Module 7: Database Internals at Depth

Database Internals at Depth — 6 lessons.

• Storage Engines Compared 16 min read Pro lesson
• MVCC Implementations 13 min read Pro lesson
• Write-Ahead Logging in Depth 13 min read Pro lesson
• Distributed Query Execution 13 min read Pro lesson
• Modern Cloud Databases 13 min read Pro lesson
• Vector Database Internals 14 min read Pro lesson

Module 8: Advanced Database Scaling

Advanced Database Scaling — 5 lessons.

• Resharding Without Downtime 12 min read Pro lesson
• Hot Shard Mitigation 10 min read Pro lesson
• Cross-Shard Queries 12 min read Pro lesson
• Read-After-Write Consistency in Replicated Systems 12 min read Pro lesson
• Database Failover Patterns 16 min read Pro lesson

Module 9: Stream Processing at Depth

Stream Processing at Depth — 6 lessons.

• Event Time vs Processing Time 10 min read Pro lesson
• Watermarks 10 min read Pro lesson
• Windowing Strategies 10 min read Pro lesson
• Streaming Joins 11 min read Pro lesson
• State Management in Streaming 12 min read Pro lesson
• Exactly-Once Semantics in Streaming 16 min read Pro lesson

Module 10: ML Systems Design

ML Systems Design — 5 lessons.

• ML Serving Architectures 12 min read Pro lesson
• Feature Stores 11 min read Pro lesson
• Model Monitoring 10 min read Pro lesson
• A/B Testing for ML Models 10 min read Pro lesson
• RAG and Vector Search Architectures 13 min read Pro lesson

Module 11: Advanced Architectural Patterns

Advanced Architectural Patterns — 5 lessons.

• Event Sourcing at Scale 11 min read Pro lesson
• CQRS Implementation Patterns 10 min read Pro lesson
• Saga Orchestration in Production 10 min read Pro lesson
• Service Mesh and the xDS API 10 min read Pro lesson
• Sidecar Pattern Beyond Service Mesh 10 min read Pro lesson

Module 12: Advanced Search and Recommendations

Advanced Search and Recommendations — 4 lessons.

• Search Relevance Engineering 10 min read Pro lesson
• Inverted Index Internals 12 min read Pro lesson
• Recommendation System Architectures 11 min read Pro lesson
• Personalisation at Scale 11 min read Pro lesson

Module 13: Edge Computing and Global Systems

Edge Computing and Global Systems — 4 lessons.

• Edge Architecture Patterns 12 min read
• Multi-Region Active-Active 11 min read Pro lesson
• Geo-Replication Strategies 12 min read Pro lesson
• Data Sovereignty and Compliance 13 min read Pro lesson

Module 14: Advanced Observability

Advanced Observability — 4 lessons.

• Cardinality Explosion and Mitigation 11 min read
• SLO/SLI Design 11 min read Pro lesson
• Distributed Tracing at Scale 12 min read Pro lesson
• AIOps and Anomaly Detection 13 min read Pro lesson

Module 15: Advanced Security

Advanced Security — 5 lessons.

• Zero-Trust Architecture 10 min read
• mTLS at Scale 11 min read Pro lesson
• Threat Modelling 10 min read Pro lesson
• Compliance Frameworks 11 min read Pro lesson
• Secrets Rotation at Scale 12 min read Pro lesson

Module 16: Modern Hardware Considerations

Modern Hardware Considerations — 3 lessons.

• NVMe Storage at Scale 10 min read Pro lesson
• Memory Hierarchies at Scale 9 min read Pro lesson
• Network Hardware - DPUs, SmartNICs, and Hardware Offload 11 min read Pro lesson

Module 17: PostgreSQL

PostgreSQL — 5 lessons.

• Postgres Origin and Design Philosophy 9 min read
• Postgres Architecture 9 min read Pro lesson
• MVCC in Postgres (and the Vacuum/Bloat Problem) 9 min read Pro lesson
• Postgres at Scale 8 min read Pro lesson
• Postgres Operational Realities 9 min read Pro lesson

Module 18: MySQL

MySQL — 4 lessons.

• MySQL History and the InnoDB Story 11 min read
• InnoDB Architecture 12 min read Pro lesson
• MySQL Replication 11 min read Pro lesson
• MySQL at Scale (Vitess, PlanetScale, and Sharding Patterns) 13 min read Pro lesson

Module 19: MongoDB

MongoDB — 4 lessons.

• MongoDB History and the Document Model 10 min read
• MongoDB Architecture 11 min read Pro lesson
• MongoDB Consistency (Read Concerns, Write Concerns, and Transactions) 11 min read Pro lesson
• When MongoDB Fits, When It Doesn't - The Honest Framing 13 min read Pro lesson

Module 20: Memcached

Memcached — 3 lessons.

• Memcached vs Redis - The Honest Comparison 10 min read
• Memcached Architecture 11 min read Pro lesson
• When Memcached Still Wins 10 min read Pro lesson

Module 21: DynamoDB

DynamoDB — 5 lessons.

• DynamoDB Origin - The Dynamo Paper and the Shopping Cart Story 10 min read Pro lesson
• DynamoDB Data Model - Partition Keys, Sort Keys, and Secondary Indexes 10 min read Pro lesson
• DynamoDB Consistency and Capacity 10 min read Pro lesson
• DynamoDB Patterns - Single-Table Design and the Rick Houlihan School 10 min read Pro lesson
• DynamoDB in Production - Hot Partitions, Throttling, and GSI Gotchas 10 min read Pro lesson

Module 22: Cassandra

Cassandra — 5 lessons.

• Cassandra Design Philosophy - Dynamo-Derived, Eventually Consistent by Design 11 min read Pro lesson
• Cassandra Data Model - Partition Keys, Clustering Columns, and Query-Driven Design 10 min read Pro lesson
• Cassandra Architecture - Consistent Hashing, Replication, and Tunable Consistency 10 min read Pro lesson
• Cassandra at Scale - What Works, What Breaks, and Operational Reality 11 min read Pro lesson
• When Cassandra Fits - Time Series, Write-Heavy, and Geographically Distributed Workloads 10 min read Pro lesson

Module 23: Elasticsearch

Elasticsearch — 5 lessons.

• Elasticsearch and Lucene - Inverted Indexes and Full-Text Fundamentals 10 min read Pro lesson
• Elasticsearch Architecture - Clusters, Nodes, Shards, and Replicas 10 min read Pro lesson
• Search Relevance in Elasticsearch - Scoring, BM25, and the Query DSL 10 min read Pro lesson
• Elasticsearch Beyond Search - Logs, Observability, and the ELK Stack 10 min read Pro lesson
• Elasticsearch Operational Realities - Heap Pressure, Split-Brain History, and the Complexity Tax 13 min read Pro lesson

Module 24: Apache Kafka

Apache Kafka — 6 lessons.

• Kafka Design Philosophy - The Append-Only Log and Why It Changed Messaging 10 min read Pro lesson
• Kafka Architecture - Brokers, Partitions, Replicas, and KRaft 10 min read Pro lesson
• Kafka Guarantees - At-Most-Once, At-Least-Once, Exactly-Once, and What Each Really Means 10 min read Pro lesson
• Kafka Producers and Consumers - Consumer Groups, Offset Management, and Rebalancing 10 min read Pro lesson
• Kafka at Scale - Partition Strategy, Retention, and Tiered Storage 10 min read Pro lesson
• Kafka Operational Realities - Lessons from Netflix-Scale Deployment 11 min read Pro lesson

Module 25: RabbitMQ

RabbitMQ — 4 lessons.

• RabbitMQ and AMQP - The Standards Heritage, Protocol Model, and Message Lifecycle 11 min read Pro lesson
• RabbitMQ Patterns - Fanout, Direct, Topic, and Headers Exchanges 10 min read Pro lesson
• RabbitMQ vs Kafka - The Honest Comparison 10 min read Pro lesson
• RabbitMQ Operational Realities - Memory Pressure, Quorum Queues, and Federation 11 min read Pro lesson

Module 26: AWS SQS

AWS SQS — 3 lessons.

• SQS Design Philosophy - The Managed Queue and What You Trade For It 11 min read Pro lesson
• SQS Standard vs FIFO - When Each Queue Type Is Right 11 min read Pro lesson
• SQS in Production - Dead Letter Queues, Visibility Timeout, Long Polling, and Lambda Integration 13 min read Pro lesson

Module 27: Apache Flink

Apache Flink — 5 lessons.

• Flink Design Philosophy - True Streaming, Not Micro-Batch 10 min read Pro lesson
• Flink Architecture - JobManager, TaskManagers, and State Backends 10 min read Pro lesson
• Flink Stateful Processing - Keyed State, Checkpoints, and Savepoints 10 min read Pro lesson
• Flink Semantics - Event Time, Watermarks, and Windowing in Production 10 min read Pro lesson
• Flink in Production - What Can Go Wrong 10 min read Pro lesson

Module 28: Apache Spark

Apache Spark — 5 lessons.

• Spark Origin and Design - From MapReduce to RDDs to DataFrames 10 min read Pro lesson
• Spark Architecture - Driver, Executors, and Cluster Manager 10 min read Pro lesson
• Spark Streaming and Structured Streaming - Micro-Batch and Beyond 10 min read Pro lesson
• Spark SQL and Catalyst Optimizer - How Spark Makes Your Queries Fast 10 min read Pro lesson
• Spark at Scale - Tuning, Shuffle, and Operational Realities 11 min read Pro lesson

Module 29: Amazon S3

Amazon S3 — 6 lessons.

• S3 Design Philosophy - Why S3 Is Built the Way It Is 10 min read Pro lesson
• S3 Architecture - What's Known Publicly and the Durability Story 10 min read Pro lesson
• S3 Consistency - The 2020 Shift to Strong Consistency 10 min read Pro lesson
• S3 Storage Classes - Standard, IA, Glacier, Deep Archive and the Economics 10 min read Pro lesson
• S3 Access Patterns - Multipart Upload, Presigned URLs, and Requesters-Pay 10 min read Pro lesson
• S3 at Extreme Scale - Patterns from Direct Experience 10 min read Pro lesson

Module 30: AWS Lambda

AWS Lambda — 4 lessons.

• Lambda Design Philosophy - Why Serverless Exists and What Lambda Optimises For 10 min read Pro lesson
• Lambda Architecture - Firecracker, Cold Starts, and the Execution Model 10 min read Pro lesson
• Lambda Integration Patterns - Triggers, Step Functions, and EventBridge 10 min read Pro lesson
• When Lambda Fits, When It Doesn't - The Honest Framing 10 min read Pro lesson

Module 31: Nginx

Nginx — 4 lessons.

• Nginx Design Philosophy - The C10K Problem and Event-Driven Architecture 10 min read Pro lesson
• Nginx Architecture - Master/Worker Processes and the Event Loop 10 min read Pro lesson
• Nginx Configuration Patterns - Reverse Proxy, Load Balancing, SSL, Caching, and Rate Limiting 10 min read Pro lesson
• Nginx at Production Scale - What Goes Wrong and How to Fix It 10 min read Pro lesson

Module 32: Apache ZooKeeper

Apache ZooKeeper — 4 lessons.

• ZooKeeper Purpose and Design - Coordination as a Service and the ZAB Protocol 10 min read
• ZooKeeper Architecture - Ensemble, Leader, Followers, and Observers 10 min read Pro lesson
• ZooKeeper APIs and Patterns - Locks, Leader Election, and Configuration Management 10 min read Pro lesson
• ZooKeeper Alternatives - etcd, Consul, and When Each Is Right 11 min read Pro lesson

Module 33: Docker

Docker — 4 lessons.

• Docker Design Philosophy - Why Containers Matter 11 min read
• Docker Architecture - Daemon, Runtime, Images, and Layers 10 min read Pro lesson
• Docker Networking and Storage - How Containers Communicate and Persist Data 11 min read Pro lesson
• Docker in Production - Anti-Patterns, Security, and Operational Lessons 12 min read Pro lesson

Module 34: Kubernetes

Kubernetes — 6 lessons.

• Kubernetes Origin and Philosophy - The Borg Lineage 10 min read
• Kubernetes Architecture - API Server, etcd, Controllers, Scheduler, and Kubelets 10 min read Pro lesson
• Kubernetes Workloads - Pods, Deployments, StatefulSets, DaemonSets, and Jobs 10 min read Pro lesson
• Kubernetes Networking - Services, Ingress, CNI, and Network Policies 10 min read Pro lesson
• Kubernetes Storage - Volumes, PVCs, StorageClasses, and CSI 10 min read Pro lesson
• Kubernetes Operational Realities - What's Hard, What Breaks, and What's Still Being Figured Out 11 min read Pro lesson

Module 35: Prometheus

Prometheus — 4 lessons.

• Prometheus Design Philosophy - Pull-Based Metrics for a Dynamic World 10 min read
• Prometheus Architecture - Scrape Model, TSDB, and PromQL 10 min read Pro lesson
• Prometheus at Scale - Sharding, Federation, Thanos, and Cortex 10 min read Pro lesson
• PromQL in Production - Query Patterns, Alerting Rules, and Recording Rules 10 min read Pro lesson

Browse all courses