Resilience for System Designs
Resilience is the ability of a system to keep working when parts fail. It is built with redundancy, health checks, timeouts, retries with backoff, circuit breakers and graceful degradation. Observability, meaning logs, metrics and traces, lets engineers see failures early and find their cause in production.
Stage: Reliability · Part of the System Design roadmap · Last reviewed 19 November 2026
What to learn in Resilience
- Add timeouts and exponential backoff with jitter to remote calls
- Explain how a circuit breaker prevents cascading failure
- Design for failover across zones with health checks
- Define availability targets using SLIs, SLOs and error budgets
- Choose which signals to collect as metrics, logs and traces
- Degrade a feature gracefully when a dependency is down
Resilience interview questions
Why add jitter to retry backoff?
If many clients retry on the same schedule after an outage, they hit the recovering service together and knock it over again. Random jitter spreads retries across time, which reduces these synchronised waves of load.
What is the difference between a metric, a log and a trace?
Metrics are numeric measurements over time, good for alerts and trends. Logs are timestamped event records with detail. Traces follow a single request across services, showing where time was spent and where it failed.
Where Resilience fits in the System Design roadmap
Resilience belongs to the Reliability stage (advanced level), which takes about 32 hours. Design for failure.
Other units in this stage
Back to the full System Design roadmap