What DBA interviews assess
DBA interviews assess knowledge across three domains: database design and modelling (schemas, normalisation, data types), administration and operations (backup, recovery, user management, performance monitoring), and performance tuning (query optimisation, indexing strategies, execution plans). The balance shifts by role: production DBAs are weighted towards operations and tuning; data architect roles are weighted towards design. Cloud DBA roles (AWS RDS, Azure SQL, Google Cloud Spanner) expect additional cloud platform knowledge on top of the fundamentals.
Database fundamentals questions
"Explain the difference between clustered and non-clustered indexes." A clustered index determines the physical order of data rows on disk. A table can have only one clustered index (the data can only be physically sorted one way). A non-clustered index is a separate structure with a pointer to the data row. Clustered indexes are most efficient for range queries on the clustering key. Non-clustered indexes are useful for queries filtering on columns that are not the clustered key. Choosing the right columns for each index type is a core tuning skill.
"What is a deadlock and how do you resolve one?" A deadlock occurs when two or more transactions each hold a lock that the other needs, creating a circular wait. SQL Server and PostgreSQL detect deadlocks automatically and kill one of the transactions (the deadlock victim). To prevent deadlocks: access objects in a consistent order across transactions, keep transactions short, use appropriate isolation levels, and minimise lock hold time. Analyse the deadlock graph in the error log or Extended Events to identify the specific lock contention pattern.
Performance tuning questions
"How do you identify and fix a slow-running query?" Process: capture the slow query (slow query log in MySQL, Query Store in SQL Server), examine the execution plan to find the expensive operators (table scans, sort operators, hash joins on large tables), check index usage (is the optimizer using the index you expect?), look at statistics freshness (outdated statistics cause poor plan choices), and check for parameter sniffing issues in stored procedures. Fixes in order of invasiveness: add a missing index, update statistics, rewrite the query, or restructure the schema.
"What is the difference between OLTP and OLAP and why does it matter for database design?" OLTP (Online Transaction Processing): high-concurrency, short transactions, normalised schema (3NF) to minimise write overhead, optimised for fast inserts and updates. OLAP (Online Analytical Processing): analytical queries, large data scans, denormalised schema (star schema, snowflake schema) to minimise joins on large datasets, optimised for read throughput. Mixing OLTP and analytical workloads on the same database creates contention: analytical queries lock resources and degrade transactional performance. The standard solution is a separate data warehouse or HTAP system.
Backup, recovery, and availability questions
"What is the difference between RPO and RTO?" RPO (Recovery Point Objective): the maximum acceptable data loss, measured in time. If RPO is 1 hour, you must be able to restore to within 1 hour of a failure. RTO (Recovery Time Objective): the maximum acceptable downtime before the system must be restored. RPO drives backup frequency; RTO drives recovery infrastructure investment (standby servers, automated failover). Both are business requirements, not technical ones — the DBA implements the technical solution to meet them.
Cloud database questions
Cloud-managed database services (AWS RDS, Azure SQL Database, Google Cloud SQL) handle the infrastructure layer: hardware provisioning, OS patching, and replication. DBAs working with cloud databases focus on: choosing the right service tier and instance size, configuring automated backups and point-in-time recovery, managing connection pooling (critical at scale since each connection costs memory), monitoring with cloud-native tools (CloudWatch, Azure Monitor), and understanding shared responsibility — you own the schema and queries; the cloud provider owns the underlying infrastructure. Know the difference between single-instance RDS and Aurora (distributed, replicated) for AWS interviews.