> Term
repeated failure mode
Incidents that occur multiple times because the underlying system incentives, architecture, or organizational habits remain unchanged.
Detailed Explanation
A repeated failure mode is an incident that is technically "fixed" during every postmortem, yet reliably recurs months later. It is the operational equivalent of treating symptoms without addressing the disease.
If your action items from the last three outages all include "add more monitoring," you are experiencing a repeated failure mode.
These failures persist because the system—whether it's the codebase, the deployment pipeline, or the company culture—actually incentivizes the behavior that causes the failure. Until the underlying structural issue is resolved, the incident will return.
Why It Matters
It drains engineering morale and consumes cycles that should be spent on feature development, while steadily degrading trust in the platform.
Common Failure Mode
Practical Example
Production Manifestation
The same service runs out of memory or connection pool limits hit the ceiling every time traffic spikes, despite multiple previous tuning attempts.
Frequently Asked Questions
What is repeated failure mode in short?
Incidents that occur multiple times because the underlying system incentives, architecture, or organizational habits remain unchanged.
What is the most common failure mode?
Postmortem action items focus exclusively on human error or superficial fixes (like increasing a timeout) rather than redesigning the flawed architecture.
AI Summary
Incidents that occur multiple times because the underlying system incentives, architecture, or organizational habits remain unchanged. It drains engineering morale and consumes cycles that should be spent on feature development, while steadily degrading trust in the platform.