NetApp CEO Tom Georgens recently told the world that he believed the entire concept of tiering was dying. The comment received quite a bit of tweets and blog responses from the talking mouths. (Is tweeting fingers more appropriate?) Perhaps we need to take his comments in full context, and giving him the benefit of the doubt, assume he was just referring to primary storage and tiering. If that’s the case, Tom isn’t taking into account the immense benefits that tiering offers to secondary storage.
We all agree that there are two forms of data: active and inactive. Active data is a work in progress, so performance and availability are important. Active data is housed within primary storage, where information is born. However, in the second part of the data’s lifecycle, the time comes for it to relocate to secondary storage. You can’t freeload in your parent’s mansion forever. Once data leaves the comfy confines of primary storage, it will now live in a secondary storage system. The information is still important, but the data is now reference information and only accessed periodically.
In the secondary storage layer, critical storage decisions must be made. It’s the last stop for data, where information either finds a permanent resting place on the lowest tier or faces death by deletion. Compliance mandates often require that information be saved for a length of time or purged after a period of time, making the secondary archive mission-critical. (These highly sensitive environments should be tiered, but more on this later.) Name the industry and legal requirements exist for keeping inactive data accessible. Sarbanes-Oxley, SEC 17-4a, 21 CFR Part 11, and DoD 5015.2 are only a handful of existing regulations.
Responsible data archiving is easily accessible, efficiently stored, and reliably secured from loss or tampering. The easiest path to accessibility, efficiency, and security is via tiering to the most appropriate storage media. As mentioned above, compliance-driven industries like health, government, and banking need to have processes in place that assure information can be accessed quickly, but also that it’s safeguarded against sabotage or accidental deletion. Tiering, commonly defined as the assignment of different categories of data to different types of storage media, ensures compliancy because data lies on accessible tiers and often includes built-in redundancy to protect against deletions or disasters.
Smart folks are always engineering the latest and greatest storage technology. The concept of tiering allows customers to use the best solutions on the market to house their data. Why wouldn’t customers string together top-rate storage platforms that can talk to each other within the tier?
What would the world look like without tiering? IT administrators would buy much more storage, lots more expensive disk arrays, and then use the older, even more expensive disk arrays for archive storage. And as for compliance, they’d cross their fingers and hope to discover and find information when called upon to do so.
Here’s the bottom line: tiering is easy, stub-less, transparent, extremely scalable, and highly secure. Show me an industry that doesn’t utilize tiering. It’s everywhere: hospitals, banks, and brokerage firms, schools, universities, and research labs. Tiering is not dying, nor is it on the way out; it’s just heading into its prime.
- Jim Moulton, CEO of Seven10 Storage Software
