Rolling disasters aren't a new problem, but like a lot of things they have been made more acute by the increased...
pace of IT life. Specifically new methods of marinating backups with minimal time lags (real-time or near-real time copy) have made rolling disasters a bigger concern.
In a rolling disaster, the failure or data corruption takes place over time rather than instantaneously. The time can be seconds to hours, but data stored during the disaster is suspect at best and quite possibly corrupt. Unless the real-time or near-real-time backup system is designed to handle rolling disasters, the backup can be corrupted as well.
Several vendors offer real-time and near-real-time products that can recover from rolling disasters. Among the products available on 360-class computers are IBM's Extended Remote Copy and Hitachi Data Systems. Such products time stamp all writes to the remote copy, allowing the system to recreate the sequence of writes. EMC offers Adaptive Copy, which can be configured to maintain a vaulted copy of changed tracks at the local site that is updated to the remote site only at the end of a given period of time, such as every 24 hours. Analogous products are available for open systems.
Hitachi Data Systems has a white paper discussing approaches to rolling disasters in terms of its own products at: www.eu.hds.com/pdf/NanoWhtPpr2.pdf.
Rick Cook has been writing about mass storage since the days when the term meant an 80K floppy disk. The computers he learned on used ferrite cores and magnetic drums. For the last twenty years he has been a freelance writer specializing in storage and other computer issues.