Why Are Data Warehouses Necessary?

Status Quo and Requirements

Much data (orders, stocks, materials, and payments) is generated in the business operation systems and background (transactional) database of enterprises.

Decision makers categorize and analyze the data for business decision-making.

Difficulties

Data categorization and analysis involve the concurrent access to the data in multiple database tables. That is, multiple tables being updated by different transactions may be locked at the same time, which may cause complications to the database systems during peak hours.

  • Locking multiple tables increases the latency of complex query.

  • The transactions that are updating the database tables are blocked, causing delays or interruptions.

Solution

Data warehouses excel in data aggregation and association, so users mine more data, get more information, and make better decisions. The mining requires complex queries that involve data on multiple tables.

The ETL process copies data in business operation databases to data warehouses for analysis and computing. Data can be aggregated from multiple business operation systems into one data warehouse for better association, analysis, and actionable insights.

Data warehouses and standard transaction-oriented databases such as Oracle, Microsoft SQL Server, and MySQL use different design modes. Data warehouses are optimized in terms of data aggregation and association but the transaction or data adding and deleting functions or performance may not be guaranteed. Therefore, data warehouses and databases apply to different scenarios. Transactional databases are dedicated to transaction processing (business operation of enterprises) whereas data warehouses excel at complex data analysis. In conclusion, databases apply to data updates whereas data warehouses apply to data analysis.