Skip to main content

Explore Problems

Pricing Discuss

Distributed Systems & Reliability

Course Contents

Welcome to Hard Mode

The Fallacies of Distributed Computing

Failure Models: Crash, Omission, Timing

Network Partitions & Split Brain

Heartbeats & Failure Detection

Checkpoint: Why Distribution Is Hard

The Clock-Sync Problem

Physical vs Logical Time

Lamport Timestamps

Ordering in Practice

Checkpoint: Time & Order

The Consensus Problem

Quorum Intuition, Revisited

Leader Election

Paxos, the Intuition

Gossip Protocols

Distributed Locks & Leases

ZooKeeper & etcd in Practice

Codelab: Leader Election with etcd

Checkpoint: Consensus & Coordination

The Conflict Problem

Operational Transformation

Conflict Resolution Strategies

Offline-First Sync

Checkpoint: Collaborative & Replicated State

Timeout Hierarchies

Retries, Budgets & Retry Storms

Circuit Breakers

Load Shedding & Graceful Degradation

Tail Latency: Living at p99

Cascading Failures

Chaos Engineering

Codelab: Run a Chaos Experiment

Checkpoint: Resilience & Overload Control

SLIs, SLOs & Error Budgets

Alerting Philosophy

Incident Response & Postmortems

Disaster Recovery: RTO & RPO

Multi-Region Patterns

Cell-Based Architecture & Deployment Stamps

Data Integrity: Backups That Restore

Shadow Traffic & Dark Launches

Checkpoint: The Reliability Discipline

Facebook, 2021: The BGP Self-Lockout

AWS Kinesis, 2020: The Thread-Limit Cascade

us-east-1: The Region the Internet Leans On

Anatomy of a Cascading Failure

How to Read a Postmortem

Checkpoint: Learning from Real Outages

Capstone Pass 3: The Marketplace Survives

Loading...

Loading content...