This directory contains 74 system design scenario files (S1.md – S74.md). Each file presents one or more interview-style problems at varying difficulty levels (Easy → Hard / Lead Architect), followed by a detailed, structured solution.
📖 What Each Scenario Contains
Problem statement — a realistic engineering challenge framed as an interview question.Difficulty levels — Easy, Medium, and Hard variants within the same topic.Solution walkthrough — requirements, data flow, component choices, trade-offs, and bottlenecks.Design rationale — explains why a particular technology or pattern was chosen.
🎯 How to Use This Section
Read the problem statement only — close the file and design the system yourself first.Sketch your design — draw the components, define data flows, estimate capacity.Reopen and compare — identify gaps, missed trade-offs, or better alternatives.Work through difficulty levels in order — the Easy variant builds intuition; Hard and Lead Architect variants add operational depth.
Suggested order: Work through S1 → S30 sequentially to build a strong base before tackling the advanced scenarios (S31+).
Foundations of Scalability & Storage
File
Topic
S1.md Scalability foundations & high-throughput log ingestion
S2.md Read scalability — caching strategy for extreme read load
S3.md Database internals & indexing (OLTP vs. OLAP vs. Geo-spatial)
S18.md Back-of-the-envelope capacity estimation
S19.md Database sharding — horizontal partitioning at scale
API Design & Communication Patterns
File
Topic
S4.md API patterns — REST, GraphQL, gRPC, synchronous vs. asynchronous
S29.md API Gateway & Backend for Frontend (BFF) pattern
Message Queues & Event Streaming
File
Topic
S5.md Message queues — SQS, RabbitMQ, and Apache Kafka compared
S39.md Kafka internal storage & zero-copy architecture
S40.md Kafka exactly-once semantics (EOS)
Batch & Stream Processing
File
Topic
S6.md Batch vs. stream processing architectures
S21.md Lambda vs. Kappa architectures
S37.md Apache Flink & unbounded stream processing
S38.md Stateful checkpointing & the Chandy-Lamport algorithm
S62.md Complex Event Processing (Flink CEP) for fraud detection
S67.md Flink exactly-once guarantees & failure recovery
Observability & Reliability
File
Topic
S7.md Observability — metrics, logs, and distributed tracing
S17.md Distributed consensus — Raft, ZooKeeper, etcd, split-brain problem
S27.md Data quality & pipeline observability
S63.md Data platform monitoring & alerting
S64.md Incident response & on-call runbooks
Container Orchestration & Infrastructure
File
Topic
S8.md Kubernetes — stateless deployments, services, stateful apps
S42.md Container orchestration at scale
S43.md Infrastructure as Code with Terraform
S44.md Cloud networking — VPCs, subnets, and security groups
S45.md CI/CD pipelines for distributed systems
S47.md Cloud cost optimization
File
Topic
S9.md Authentication, authorization & zero-trust networks
S16.md Advanced rate limiting strategies
S46.md IAM & cloud security (least-privilege access)
S48.md Data encryption & row-level security
Multi-Region & Global Systems
File
Topic
S10.md Multi-region architecture & global state synchronization
File
Topic
S11.md Deconstructing Uber — location ingestion, spatial indexing, matching
S12.md Deconstructing Netflix — video ingestion, transcoding, CDN streaming
S13.md Deconstructing WhatsApp & Discord — 2 billion concurrent connections
S30.md Design Ticketmaster — high-concurrency seat reservation
S50.md Uber's real-time surge pricing engine
E-Commerce & Distributed Transactions
File
Topic
S14.md Distributed shopping cart, inventory concurrency & idempotency
S20.md Distributed transactions — Two-Phase Commit & Saga pattern
ML, AI & Generative AI Systems
File
Topic
S15.md ML feature store & recommendation system architecture
S57.md ML platform & LLMOps for data-driven products
S58.md Retrieval-Augmented Generation (RAG) & vector search
S60.md LLM agent orchestration with memory and tool use
Data Warehouse, Lakehouse & Storage Formats
File
Topic
S23.md Data Warehouse vs. Data Lake vs. Lakehouse
S31.md Medallion Architecture (Bronze / Silver / Gold layers)
S52.md Lakehouse transactional formats — Delta Lake & Apache Iceberg
File
Topic
S32.md Slowly Changing Dimensions (SCDs)
S33.md Fact table design patterns
S61.md Data Vault modeling for multi-source enterprise warehouses
File
Topic
S24.md Apache Airflow — DAG design & dependency management
S53.md Advanced Airflow — dynamic task generation & scaling
Data Transformation with dbt
File
Topic
S26.md dbt (data build tool) & ELT pipelines
S51.md Advanced dbt — Jinja macros & DRY SQL
S59.md dbt testing & data quality enforcement
Data Governance, Catalog & Contracts
File
Topic
S25.md Data Mesh architecture & domain ownership
S49.md Data catalog & metadata governance
S54.md Data contracts — schema evolution & upstream coordination
S56.md Organizational data strategy & decentralized data teams
Change Data Capture & Real-Time OLAP
File
Topic
S22.md Change Data Capture (CDC) with Debezium
S28.md Real-Time OLAP — ClickHouse & Druid for customer-facing analytics
File
Topic
S34.md Spark physical execution — driver, executors, DAG
S35.md Advanced Spark joins — broadcast, sort-merge, shuffle
S36.md Spark memory management & OOM debugging
S65.md Spark Catalyst Optimizer & query planning
S66.md Spark shuffle — data movement across the network
Systems Programming & Low-Level I/O
File
Topic
S68.md Zero-copy I/O & kernel/user space optimization
S69.md Memory-mapped files, page cache, and DPDK
Event-Driven Architecture
File
Topic
S41.md Event Sourcing & CQRS
Data Security & Compliance
File
Topic
S48.md Data encryption & column-level access control
S74.md GDPR Right to be Forgotten in append-only systems
Data Activation & Cloud Platforms
File
Topic
S55.md Serving layer — making warehouse data queryable by applications
S70.md Reverse ETL & operational data activation
S71.md Snowflake & Databricks — enterprise cloud data platforms
S72.md CI/CD for data — testing and deploying warehouse changes safely
S73.md End-to-end system design interview framework (Lambda/Kappa)
Spotted an error or want to suggest a new scenario? Open an issue or pull request on the root repository.
Content is shared for educational purposes. Please credit the source if you reuse or adapt any material.
← Back to repository root · by Sepuri Sai Krishna