Comparison questionMedium~7 min

Interview question

During an incident, how would you decide between rollback, roll-forward, a kill switch, or degraded mode?

Choose rollback, roll-forward, or kill switch during incidents

Compares rollback, roll-forward, feature kill switches, scaling, and degraded mode as incident mitigations.

1Say itAnswer out loud first

2CompareCompare with Core answer

3RefineRefine with Stronger and Advanced

TL;DR

I would choose based on impact, confidence, and reversibility. Rollback is good when the last version is compatible and the release caused the issue. Roll-forward is better when rollback is unsafe because of migrations, data changes, or external contracts. A kill switch or feature flag is useful when one feature path is bad and the rest of the system can keep serving users.

Full answer below

Practice focus

Mitigation choice, reversibility, data compatibility, feature flags, user impact, blast radius, and recovery verification.

Answer first

Answer path

Core answer first, then checkpoints toward a stronger answer.

Start
Core answer
Start here. This is the direct spoken answer to practice first.
Why this question matters
Incident mitigation is about reducing user impact safely. The fastest option is not always the safest option.
I would choose based on impact, confidence, and reversibility. Rollback is good when the last version is compatible and the release caused the issue. Roll-forward is better when rollback is unsafe because of migrations, data changes, or external contracts. A kill switch or feature flag is useful when one feature path is bad and the rest of the system can keep serving users.
Checkpoint
Stronger answer
Optional upgrade with examples, trade-offs, and sharper phrasing.
Degraded mode fits when the product can temporarily skip optional work, serve cached data, delay background processing, or reject expensive actions. Scaling can help when the problem is capacity, but it can make dependency pressure worse if the bottleneck is SQL, Redis, or a provider. I would avoid changing several things at once because that makes recovery harder to attribute.
Finish
Advanced answer
Optional deeper angle for production judgment and nuance.
In production I define mitigation options before incidents. Critical features have known kill switches, rollback checks, migration compatibility rules, ownership, and runbooks. I choose one change at a time where urgency allows, record the decision and expected signal, and verify recovery through user-facing success, latency, backlog, and business-state metrics rather than deployment status alone. Traffic is restored gradually so a recovering dependency or database is not hit by the full backlog at once. If mitigation requires replay or data repair, recovery includes reconciliation and explicit confirmation that duplicate, missing, or inconsistent business effects have been resolved.

Review signals

Use these to catch weak phrasing, shallow reasoning, and risky claims.

Common mistakesCandidateCommon mistakesMistakes that weaken the answer or signal shallow understanding.Open details

Details

Rolling back through incompatible database changes.
Trying several mitigations at once with no timeline.
Scaling out when the bottleneck is an overloaded dependency.
Leaving kill switches undocumented until an incident.

Red flagsCandidateRed flagsSignals that an interviewer may treat as risky.Open details

Details

No rollback safety criteria.
No degraded-mode thinking.
No verification after mitigation.
No runbook for common mitigation paths.

Follow-up prompts

Quick interviewer prompts inside this drill.

Follow-up questionsCandidateFollow-up questionsLikely next questions after the main answer.Open details

Details

Treat these as the next interviewer turn. Answer one out loud, then save the gap in your practice notes.

When is roll-forward safer than rollback?
What makes a good kill switch?
How would you verify degraded mode is working?

Continue practice

Move to the linked follow-up, next path step, prerequisite, or deeper variant.

2 linked

Go deeper

Practice the next layer of the same subject.

Rollback vs Roll Forward

Try a variant

Answer a similar question in a different context.

Hotfix vs Rollback

Comparison questionMedium~7 min

Interview question

During an incident, how would you decide between rollback, roll-forward, a kill switch, or degraded mode?

Choose rollback, roll-forward, or kill switch during incidents

Compares rollback, roll-forward, feature kill switches, scaling, and degraded mode as incident mitigations.

1Say itAnswer out loud first

2CompareCompare with Core answer

3RefineRefine with Stronger and Advanced

TL;DR

Full answer below

Practice focus

Mitigation choice, reversibility, data compatibility, feature flags, user impact, blast radius, and recovery verification.

Answer first

Answer path

Core answer first, then checkpoints toward a stronger answer.

Start
Core answer
Start here. This is the direct spoken answer to practice first.
Why this question matters
Incident mitigation is about reducing user impact safely. The fastest option is not always the safest option.
I would choose based on impact, confidence, and reversibility. Rollback is good when the last version is compatible and the release caused the issue. Roll-forward is better when rollback is unsafe because of migrations, data changes, or external contracts. A kill switch or feature flag is useful when one feature path is bad and the rest of the system can keep serving users.
Checkpoint
Stronger answer
Optional upgrade with examples, trade-offs, and sharper phrasing.
Degraded mode fits when the product can temporarily skip optional work, serve cached data, delay background processing, or reject expensive actions. Scaling can help when the problem is capacity, but it can make dependency pressure worse if the bottleneck is SQL, Redis, or a provider. I would avoid changing several things at once because that makes recovery harder to attribute.
Finish
Advanced answer
Optional deeper angle for production judgment and nuance.
In production I define mitigation options before incidents. Critical features have known kill switches, rollback checks, migration compatibility rules, ownership, and runbooks. I choose one change at a time where urgency allows, record the decision and expected signal, and verify recovery through user-facing success, latency, backlog, and business-state metrics rather than deployment status alone. Traffic is restored gradually so a recovering dependency or database is not hit by the full backlog at once. If mitigation requires replay or data repair, recovery includes reconciliation and explicit confirmation that duplicate, missing, or inconsistent business effects have been resolved.

Review signals

Use these to catch weak phrasing, shallow reasoning, and risky claims.

Common mistakesCandidateCommon mistakesMistakes that weaken the answer or signal shallow understanding.Open details

Details

Rolling back through incompatible database changes.
Trying several mitigations at once with no timeline.
Scaling out when the bottleneck is an overloaded dependency.
Leaving kill switches undocumented until an incident.

Red flagsCandidateRed flagsSignals that an interviewer may treat as risky.Open details

Details

No rollback safety criteria.
No degraded-mode thinking.
No verification after mitigation.
No runbook for common mitigation paths.

Follow-up prompts

Quick interviewer prompts inside this drill.

Follow-up questionsCandidateFollow-up questionsLikely next questions after the main answer.Open details

Details

Treat these as the next interviewer turn. Answer one out loud, then save the gap in your practice notes.

When is roll-forward safer than rollback?
What makes a good kill switch?
How would you verify degraded mode is working?

Continue practice

Move to the linked follow-up, next path step, prerequisite, or deeper variant.

2 linked

Go deeper

Practice the next layer of the same subject.

Rollback vs Roll Forward

Try a variant

Answer a similar question in a different context.

Hotfix vs Rollback

During an incident, how would you decide between rollback, roll-forward, a kill switch, or degraded mode?

Practice focus

Answer path

Why this question matters

Review signals

Follow-up prompts

Continue practice

Go deeper

Try a variant

During an incident, how would you decide between rollback, roll-forward, a kill switch, or degraded mode?

Practice focus

Answer path

Why this question matters

Review signals

Follow-up prompts

Continue practice

Go deeper

Try a variant