Unitiv Blog

Storage capacity has boomed over the past decade, but so has the need for that capacity. The answer used to be to just throw more and more hardware at the problem, but the fact is that today’s highly-specialized and often-virtualized environments require fewer and fewer potential bottlenecks, and simply adding drives is not a good solution. Smart folks recognize that handling today’s data loads involves not only increasing capacity, but finding creative ways to reduce data requirements.

Here are the top 5 things your organization to begin doing today in order to reduce data requirements:

1. Compress. Compression is probably the data reduction process that first comes to mind, and is certainly the most commonly discussed. Compression looks for repeated byte patterns and eliminates them. This works wonders with things like file storage, email and databases, but not nearly as well with images. Unfortunately, some of the biggest data requirements today are image-based.
2. Deduplicate. Tracking down and getting rid of duplicate data can drop your storage requirements significantly. While experts disagree as to how significant this reduction can be, most suspect that it is 30 percent or more, with the more optimistic estimating as much as 90 percent. Perhaps the best illustration of deduplication is the email attachment that is sent out from an officer to the hundreds of employees in his area. That attachment is duplicated hundreds of times, while it’s only necessary truly to store a single copy.
3. Create tiering policies. Tiering of data based on policies is another way to reduce data. This is a highly specialized method, however, and depends greatly on things like retention requirements and compliance. Still, you can tier data such that the data that is infrequently used or doesn’t require quick access is stored on cheaper, slower media.
4. Provisioning thinly. Thin provisioning lets you set up a particular application server to only utilize a specific volume of space on a given drive, but to not access it until it’s actually necessary. This won’t reduce your data requirements, but it does allow you to delay increasing capacity.
5. Virtualize. In the same way that server virtualization brings multiple servers onto fewer hardware devices, so storage virtualization pulls your storage devices into a single storage pool. This lets storage admins put data where it’s best utilized, I terms of tiering.