Data Storage: Mistaking Backups for Archives

Companies often make the mistake of confusing archives with backups.  That’s an observation made by Laura Guio, IBM Vice President, Storage Sales, Systems & Technology Group, at the 2011 Implementing Information Infrastructure Symposium (IIIS) in Sydney sponsored by ComputerWorld.

Backups and Archive data serve different purposes within the business and should be treated as such, but many companies attempt to use their backups for archival.  While costs of archiving tend to be higher, Guio suggests that the use of virtualized storage could help in keeping costs down by increasing storage utilization by as much as 30 percent.

At the same IIIS conference, IBRS advisor, Kevin McIsaac emphasized again that “Backup is not archive.  The problem of trying to use backup as archive is you get a discovery request for email and now you have to go back and dump your backup tapes to a system, get them back, load them up into Exchange, only to find that the emails you are looking for are not there. You then have to go back and do it again and again.  The process is hopeless.”

Mark Nielsen, HP StorageWorks sales and category manager, pointed out that many users fail to fully understand the concept that not all data is of equal value to an organization and that storage management solutions should factor in those differences.  Guio said the first step in building an effective storage solution is to understand the range of data that the organization must manage and the characteristics of the different data types.  Data often quickly goes stale, for example, over a three month window most organizations access less than 30 percent of all the data which they have stored.  Other pieces of data and support information may be infrequently accessed but may be required to be produced for regulatory or business management purposes.

Nielsen said when valuing data to consider what its relevance is today and how the relevance is likely to change in the future.  In this way you can build an understanding of the lifecycle for types of data and then construct a corresponding strategy for migrating the data.  Clive Gold, EMC ANZ marketing CIO, countered this point of view by pointing out that classifying data and assigning weighted values to the importance can be biased and that different weightings will be assigned by different people.  Value weightings derived by actual data-type access statistics and from data identified as being required for regulatory purposes seems to be fairly non-biased ways of assessing the relative importance of different data.

Guio commented that archiving and backup processes are complementary processes.   Effective archiving can reduce backup costs by 60 percent and reduce the time to backup by 80 percent.  She summarized three recommendations that she typically gives to her clients:

• Try to stop storing so much
• Try to fully utilize the storage that you currently own
• Try to partition data into data types with defined lifecycles and automate data to be migrated through the lifecycle tiers