Buckle up, intrepid coder! The Art of Scalable Systems is your passport to unraveling the mysteries of services. Explore the hilarious mishaps and ingenious solutions to everyday problems.
📚 NB: This is an attempt at documenting all my system design notes from reading and hands-on experience, starting with explaining the basic components of a service to providing examples of real world systems and discussing tradeoffs of possible solutions.
| System components | System concepts |
|---|---|
| Load Balancer | ACID |
| API Gateway | CAP Theorem |
| Caching | PACEL Theorem |
| DNS | |
| Message Queues | |
| Proxies | |
| Databases |
| Problem | Solution |
|---|---|
| Check set contains | Bloom Filters |
| Where to insert data | Consistent Hashing |
| Ensure strong consistency | Quorum Consensus |
| Ensure high availability | Leader And Follower |
| Atomicity in data transfers | |
| Concurrent updates to file | Locking Lease |
| Concurrent node writes | Vector Clocks |
| Node liveness | |
| Deal with dead nodes coming back | |
| Update dead nodes with missed data | Hinted Handoff |
| Update stale nodes | |
| Check data inconsistency | Checksum |
| Chapters |
|---|
| CQRS |
| Two Phase Commit |
| Transactional Outbox |
| Change Data Capture (CDC) |
| Chapters |
|---|
| Domain Driven Design |
| The Strangler Fig Migration Pattern |
| Chapters |
|---|
| Distributed Message Queue & Event Streaming Platform |
Infrastructure Bits and Bobs
| Chapters |
|---|
| K8S - ClusterIP vs NodePort vs LoadBalancer |
| Chapters |
|---|
| Parties: Acquirer, Issuer |
| Events: Chargebacks, Reversals |