Before launch: backups tested, indexes reviewed, monitoring alerts, least-privilege users, and a migration plan for schema changes.
Checklist
- Replica set (or Atlas) with backup restore drill completed
- Indexes on hot queries; no accidental COLLSCAN on APIs
- Connection string secrets in vault; TLS enforced
- Validators for critical collections
- Explain slow queries; Performance Advisor on Atlas
- Document shard key strategy before data exceeds one RS
- Runbook for failover and connection storm (pool sizes)
Next steps
Deepen relational skills on PostgreSQL or MySQL when your stack is polyglot. Practice Node.js with Mongoose for full-stack apps.
Important interview questions and answers
- Q: Backup without restore test?
A: Unknown RTO—treat as incomplete disaster planning. - Q: Production on M0?
A: Upgrade tier before real traffic and SLA commitments.
Self-check
- List three pre-launch checks.
- Why restore drill?
Tip: A backup you never restored is hope—not a disaster plan.
Interview prep
- Restore drill?
- Proves backups work—defines real RTO/RPO.
- Validator in prod?
- Protects data from rogue writers and shells.