Reliability in Distributed Systems

Explore reliability in distributed systems through this 42-minute conference talk from EuroPython 2018. Delve into essential aspects of system stability, dependency management, and problem identification in microservices and APIs. Learn about multi-level monitoring techniques, including hardware, application, external, and anomaly detection. Discover standard methods for preventing system failures during dependency outages and strategies for recovering from inconsistent states. Gain valuable insights on monitoring resources, client-side monitoring, functionality checks, and anomaly detection. Understand the importance of error tracking, strategic deployment, environment management, and conducting effective postmortems. Suitable for beginners and intermediates, this talk provides a comprehensive overview of techniques to enhance the reliability and resilience of distributed systems.

Intro
Systems
Goal
What do we communicate with?
Timeouts
Circuit breakers
Async
Disaster recovery
Monitoring - resources
Monitoring - from client side
Monitoring - functionality
Monitoring - anomaly detection
Alerting - error tracking
Deployment where to put it?
Environments
Monitoring - web server
Postmortems
Summary - prevent failing
Summary - monitoring
Summary - testing
Summary - architecture
Summary - logging
Summary - alerting