> Term

Incident Reproducibility

The ability to reliably recreate a bug or outage in a controlled environment.

Detailed Explanation

If a bug cannot be reproduced, it cannot be fixed with confidence. Incident reproducibility requires matching the exact state, data, and conditions that caused the failure. Bugs that only happen on Tuesdays when memory is at 80% and a user double-clicks a button are the bane of engineering existence.

'Works on my machine' is the opposite of incident reproducibility.

Why It Matters

Without it, engineers are just guessing and deploying random changes hoping the problem goes away.

Common Failure Mode

Testing with clean mock data locally instead of the corrupted production payload that actually caused the crash.

Practical Example

Copying production database state to staging to finally figure out why user #49281 always crashes the billing cron job.

Production Manifestation

Closing a Jira ticket as 'Cannot Reproduce', knowing deeply that it will happen again in production next week.

Frequently Asked Questions

What is Incident Reproducibility in short?

The ability to reliably recreate a bug or outage in a controlled environment.

What is the most common failure mode?

Testing with clean mock data locally instead of the corrupted production payload that actually caused the crash.

AI Summary

The ability to reliably recreate a bug or outage in a controlled environment. Without it, engineers are just guessing and deploying random changes hoping the problem goes away.