Summary
Load shedding rejects lower-priority work to protect the core system. Graceful degradation keeps the product usable with reduced functionality.
Interview Points
- Load shedding is protective rejection during overload.
- Graceful degradation is reduced capability instead of total failure.
- Prioritize critical user paths over optional features.
- Use cached data, simpler responses, disabled recommendations, or read-only modes.
- Make degradation explicit and observable.
2-3 Minute Interview Script
“Load shedding and graceful degradation are both overload strategies. Load shedding means intentionally dropping or rejecting work to keep the system alive. Graceful degradation means serving a reduced experience instead of failing completely.
For example, during high load an e-commerce site might preserve checkout, but disable recommendations, expensive personalization, or non-critical analytics. If necessary, it may reject low-priority traffic with clear retry behavior.
The key is to define priority before the incident. Critical paths get capacity first. Optional features should have fallbacks, cached responses, or feature flags that can be turned off quickly.
In an interview, I would say load shedding protects the service, while graceful degradation protects the user experience.”
Follow-Ups
- How do you decide what to shed?
- What should clients see during degradation?