Can most operational (OLTP) databases be directly reused for a wide range of BI analytics without remodeling, or do BI applications typically require dedicated analytical structures?

Introduction / Context:
Operational databases are optimized for fast writes and transaction integrity, whereas BI workloads emphasize aggregations, historical trends, and complex joins across large datasets. Hence, BI commonly relies on warehouses, marts, or lakehouses designed for analytics rather than directly querying OLTP schemas for most use cases.

Given Data / Assumptions:

• OLTP focuses on normalized schemas, small transactions, strict concurrency.
• BI focuses on read-heavy, scan-heavy queries and time-series analysis.
• Analytical stores are denormalized or columnar and include historical snapshots.

Concept / Approach:
Attempting BI on OLTP strains performance and risks lock contention. Analytical structures (star schemas, columnstores, materialized views) and curated semantic layers drastically improve performance and usability for analysts.

Step-by-Step Solution:
Extract from OLTP via ETL/ELT.Transform to analytic models (facts/dimensions, conformed dimensions).Optimize storage and indexes for reads and aggregations.Expose via BI tools for dashboards and ad hoc analysis.

Verification / Alternative check:
Measure query latency and lock contention when running heavy reports directly on OLTP versus on a warehouse; the dedicated analytic store will generally outperform and isolate workloads.

Why Other Options Are Wrong:
Claiming direct reuse for most BI ignores differing workload patterns (option b). Columnar OLTP is uncommon and still not ideal for complex analytics (option c). Lack of historical needs is rare in BI (option d). Disabling ETL is not a sensible precondition (option e).

Common Pitfalls:
Overloading OLTP with reporting; skipping data quality and conformance; misunderstanding why star schemas and columnstores exist.

Final Answer:
BI typically requires dedicated analytical structures; direct reuse is limited