An organisation runs a proprietary warehouse and is evaluating a lakehouse. Leadership asks what they gain architecturally by moving to open table formats on object storage. What is the strongest answer?
Select an answer to reveal the explanation.
Short Explanation
Open table formats let you keep one governed copy of the data and read it with multiple engines. You're not locked into the engine that wrote the data, and you can add another tool next year without migrating everything. That's the architectural win, not a promise of faster queries.
Full Explanation
The strongest answer is the one saying one governed copy of the data can be read by multiple engines and tools, since the format is open rather than tied to a single vendor's engine. When tables are defined in an open format on object storage, no single engine owns the data. Different engines suited to different workloads read the same governed copy, and adopting a new engine does not require a migration.
The claim that queries become faster than the warehouse in every workload, because object storage has lower latency, is incorrect because it overstates the case. A tuned proprietary warehouse can outperform a lakehouse on some workloads, and object storage does not have lower latency than local storage. Performance comes from table layout, partitioning and engine capability.
The claim that governance becomes unnecessary, since open formats include their own permission model, is incorrect because open formats define structure and transactional semantics, not entitlement. Governance remains a distinct and necessary layer.
The claim that schema design becomes unnecessary, since open formats accept arbitrary structures, is incorrect because these formats are schema-aware by design. They support schema evolution, which is not the same as having no schema.
For the exam, express the benefit as shared governed data across engines, and be suspicious of any answer promising universally better performance.