Skip to main content

> Term

resilience gap

The hidden vulnerability between how a system is supposed to handle failure and how it actually behaves during an incident.

Detailed Explanation

The disparity between designed safety mechanisms (like fallbacks and retries) and the harsh reality of unexpected production failure modes.

These gaps are often found in the seams between microservices, where timeouts are mismatched or circuit breakers are misconfigured.

Why It Matters

It is exactly where cascading failures begin. Assuming resilience without testing it ensures the system will break in the most unpredictable way possible.

Common Failure Mode

Service A has a 10-second timeout, but Service B (its caller) has a 5-second timeout. Service B drops the connection, but Service A keeps working, wasting resources in a massive resilience gap.

Practical Example

A database failover takes 30 seconds, but the web layer is hard-coded to crash the application if the database is unreachable for 5 seconds.

Production Manifestation

Missing rate limits, endless retry loops without backoff, and 'safe' degradations that actually trigger memory exhaustion.

Frequently Asked Questions

What is resilience gap in short?

The hidden vulnerability between how a system is supposed to handle failure and how it actually behaves during an incident.

What is the most common failure mode?

Service A has a 10-second timeout, but Service B (its caller) has a 5-second timeout. Service B drops the connection, but Service A keeps working, wasting resources in a massive resilience gap.

AI Summary

The hidden vulnerability between how a system is supposed to handle failure and how it actually behaves during an incident. It is exactly where cascading failures begin. Assuming resilience without testing it ensures the system will break in the most unpredictable way possible.