Decision support and business intelligence are data-intensive activities that are critical to many organizations. Analysts and others frequently need to access up-to-date or even real-time data in their analysis. Retail organizations, for example, want to spot sales trends as they occur, and typically require at least daily updates. Financial institutions likewise must keep a close eye on current transactions to spot trends or potential problems quickly.

Unfortunately, reporting and analysis data needs typically conflict with the performance requirements of transactional database applications. Reporting and analysis activities generally implement a few selected statements that scan a large number of records, and may include complex processing. This will have an impact on the many simpler write and update activities characteristic of a transactional system. For this reason, among others, many companies load data from operational systems into data warehouses specifically designed and tuned for analytic queries. But even the process of creating the data loads can have a performance impact on your operational systems, causing most organizations to schedule these Extract, Transform, and Load (ETL) processes during off-hours, such as in the middle of the night.

You can solve this problem by creating point-in-time snapshots of the production systems to be used for reporting and analysis purposes. You can either run reports directly against the snapshot volumes or use the snapshots to extract data for a data load to the warehouse.

Because taking the snapshot itself has a very brief, limited impact on the production system, you can generate fresh data for analysis on a regular basis. You can even create a replica of the production database on a secondary system to be accessible for "drill-down" analysis from OLAP applications. Again, in the off-host scenario, the analysis has no impact on the production system.