Is redundancy wasteful?
It trades short-term efficiency for long-term resilience. The cost of redundancy is almost always less than the cost of a catastrophic single-point failure.
Mental Models
Redundancy
The inclusion of extra components or pathways that serve as backups when primary ones fail.
Redundancy trades efficiency for resilience by ensuring that no single point of failure can bring down the whole system.
Core question
If this one thing breaks, does everything else still function?
Example
A SaaS company stores data across multiple availability zones. When one data center has an outage, the service stays live because redundant copies handle the load.
How to apply
- 1.Identify single points of failure in the system.
- 2.Rank them by blast radius — how much breaks if this one thing fails.
- 3.Add backup paths for the highest-blast-radius components.
- 4.Test failovers regularly to confirm redundancy actually works.
Pitfalls
- ·Adding redundancy everywhere and creating unmanageable complexity.
- ·Never testing backup paths, so they fail silently when needed.
- ·Treating redundancy as a substitute for fixing root causes.
Questions
Where should redundancy be applied outside engineering?
Key-person risk in teams, revenue concentration on one client, and single-channel marketing are all business vulnerabilities that benefit from redundancy.
Related
- Margin of Safety
Build a buffer between your plan and the worst plausible outcome.
- Antifragility
Some systems get stronger from stress and disorder.
- Bottleneck
The narrowest constraint sets the throughput of the entire system.