What you will learn from this tip: Rick Cook outlines a number of approaches to tiered data migration.
As storage becomes more sophisticated, and information lifecycle management (ILM) becomes a fact of life for more enterprises, the data migration among storage tiers becomes more complex.
Ultimately, at least some of the data will be migrated to tape, magneto-optical (MO) or some other form of archival storage, a function normally handled by the backup software. However, before then, larger amounts of data may need to be shifted three or four times to progressively cheaper storage.
One goal in all of this is migrating data with minimum disruption. Ideally, the applications using the data should require either minimal or no changes as the data is migrated until, perhaps, it is finally archived. However, this isn't easy, especially if you want to do it automatically in response to established business rules.
There are a number of approaches to the process of tiered migration, depending on budgets and specific needs.
A number of storage management systems, including Hewlett-Parckard Co.'s StorageWorks, offer file migration agents -- either built-in or as add-ons -- that are designed to work closely with the main management application.
For many vendors, tiered migration is part of storage virtualization. Companies such as EMC Corp., (which bought Rainfinity) and Neopath Networks Inc., offer NAS virtualization front ends which can shuffle data between storage devices, and hence storage tiers, automatically. Ideally, the process is invisible to users and applications alike. Likewise, companies like Incipient Inc., offer migration as part of their virtualization offerings for storage area networks (SAN). Incipient's NSP software runs on the SAN switches to handle moving data.
A potentially cheaper approach, especially for small and medium-sized enterprises, is to keep the data on the same storage array, but shift it among RAID levels. Something like RAID-5 requires less redundancy, and hence makes cheaper storage, than RAID-10, which requires 2 megabytes (MB) of storage capacity for each MB of data. This saves the expense of additional storage devices, but usually means that the system can't take advantage of lower-cost technologies such as SATA.
Do you know…
About the author: Rick Cook has been writing about mass storage since the days when the term meant an 80 K floppy disk. The computers he learned on used ferrite cores and magnetic drums. For the last 20 years, he has been a freelance writer specializing in storage and other computer issues.