# Interview — Replicated ledger

**Q1. What problem does this mechanism solve?**
How do three nodes agree on the same append-only transaction history? Use an ordered, validated log and deterministic replay; do not confuse replication with Byzantine consensus.

**Q2. What is the naive failure mode?**
A single centralized database is a single trust authority; copying data without deterministic ordering produces divergent histories.

**Q3. Is this enough for a secure public blockchain?**
No. This lab isolates one mechanism; production needs cryptographic identities, network rules, adversarial validation, consensus and operational safeguards.

**Q4. What would you test?**
Happy path, malformed input, tampering and disagreement or competing histories where relevant.