Hotfix = emergency production fix. Faster than normal release; QA still required.
Hotfix process:
1. Incident triage.
- Severity confirmed.
- Decision to hotfix vs wait.
2. Root cause analysis.
- Quick understanding of issue.
- Identify minimal fix.
3. Fix development.
- Dev makes change in hotfix branch.
- Reviewed by lead.
4. QA testing.
- Targeted tests: confirm fix works.
- Regression on critical paths: confirm fix doesn't break.
- Limited time — usually hours, not days.
5. Release readiness.
- Critical defect resolved.
- No major regressions.
- Risk of release < risk of incident.
6. Deploy.
- Production hotfix deployment.
- Monitor closely.
7. Hypercare.
- Heightened monitoring post-deploy.
8. Post-mortem.
- Why did this happen?
- What can prevent recurrence?
Hotfix QA realities:
- Less testing than normal — accept some risk.
- Faster turnaround — hours vs days.
- Pressure — but rigor matters.
Tools:
- Targeted regression suite for hotfix scenarios.
- Quick automated tests in CI.
- Manual verification of fix.
Common pitfalls:
- Insufficient testing — fix breaks something else.
- No rollback plan — fix worse than original.
- No post-mortem — same issue recurs.
Senior QA insight: hotfixes are stressful but important. Have a hotfix process; don't improvise.
The senior framing: hotfixes save customers; bad hotfixes lose them.