Data virtualization as a solution — Part 2: the hidden cost of copying data

Copying data is expensive — and the storage hardware is usually the smallest part of the bill. The larger cost comes from everything that follows once data is duplicated. This second part looks at those follow-on costs and the role data virtualization plays in avoiding them.

The most cost-effective aspect of copying data is the hardware on which the data is stored.

Governance, ownership and infrastructure

Copied data still has to meet every compliance, confidentiality and privacy requirement, with authorization concepts, auditing, integrity and encryption maintained around it. Copying also raises the question of data ownership and the "one version of truth": business logic gets copied alongside the data, and those duplicated strands drift apart over time, producing inconsistent key figures. On top of that come application servers, network links, firewalls, licenses, updates and monitoring for the whole redundant stack.

Maintaining load processes — and the cost of neglecting them

ETL routines must be monitored for the entire retention period, because a change in a source or target can break them. Hardware fails, so disaster recovery is mandatory and adds its own cost. Because all of this is effortful, it is often deprioritized — and neglect is risky: missed updates open security holes, unnoticed load failures yield outdated or wrong data, and a privacy or reporting failure can mean penalties and lasting reputational damage.

No hardware has eternal life.

Data virtualization removes the effort at its root

It is one small, inconspicuous step that triggers all of this: the data is copied. Because virtualization leaves data in its source, governance stays in the source system, authenticated and authorized via Single Sign-On; there are no load processes to break unnoticed; nothing compliant has to be re-implemented elsewhere; and business logic can live centrally with no disputes over interpretation.

This holds as long as the virtual space keeps its discipline: no persistence (not even "just for performance"), and consistent Single Sign-On with no technical-user shortcuts. Within those rules, data virtualization significantly reduces effort and cost while keeping enterprise data secure, reliable and consistent.