I have been thinking about data storage and protection recently and how our behaviour is driving massive growth in cost and complexity.  The issue seems to eminate from the fact that we focus on solving backup when the real problem is restoring data:

• We need to be able to deal with data corruption and incorrectly added transactions (rollback)
• We need to be able to deal with partual or complete data loss caused by hardware malfunction, malicious acts or human error (data recovery)
• We need to be able to deal with total data centre loss resulting in loss of data (site recovery)
• We need to provide archive data to comply with regulatory and legal requirements (compliance)
• We need to be able to take snapshots of point in time data for testing (copy data)

Each of these requirements tends to drive a separate solution involving multiple copies of data in seperate towers. Complexity increases, volumes grow out of hand and we singularly fail to achieve the key objectives:

• Rapid data recovery with engineered recovery point objective
• Safe and secure business continuity protection
• Compliance with regulations and laws
• Clear understanding of where our data is and what version it is at

 If we were looking to solve these problems and meet these requirements then tape backup, deduplication and archiving proceedures that we commonly use today would not be where to start.  Storage, Server and Network Virtualisation has freed us from the tyrany of phyical connections between hardware and application, yet data protection pulls us back again.  We need a new protection and availability architecture.