Anyone who has worked in a data center that’s been cobbled together over time has had the same dream: what would I do if I could start from scratch? How would I design a data center and strategy that’s more efficient, easier to use and less of a burden to IT?
For data enthusiasts like us, this is a fun exercise: what would your greenfield data storage strategy be? Of course, you can’t build a data storage strategy from scratch -- you already have too much data, too many systems, too much inertia. That doesn’t mean you’re powerless to change, though. If you could start from scratch, you’d probably do the following three things. Luckily there are ways to get similar benefits, even if you’re not living your exact data center dream.
Make your data center flexible.
One thing about on-prem data centers is they require a significant amount of upfront commitment and investment. You have to purchase much more capacity than you need right now, based on growth expectations. The problem is, you’re paying to manage and run that capacity before you ever put it to use. The reverse is also true; if you can reduce the amount of space you’re using, you still have the hardware, so your savings are limited.
The first thing we’d do with our hypothetical from-scratch data storage model, then, is make it truly flexible. When we need more capacity, it appears. When we need less, it disappears. We’re never paying for more than we need, and we’re never rushed to add capacity in a pinch.
What you can do: OK, we’re not fooling anyone here. If you haven’t at least looked at the many cloud options for data storage, and other IT needs, you are in the smallest minority; companies have moved all sorts of workloads and all types of data to the cloud to gain flexibility and performance, and reduce costs. There are many as-a-service options out there that give you almost the same flexibility that our magical appearing and disappearing servers would.
Make your data center self-protecting.
Losing data to a disaster, attack or user error is high on the list of things that keep IT up at night. Most data centers have security, backup and disaster recovery built on top of the other functionality. If we were starting from scratch, we’d build these features right into the fabric of the infrastructure. That way we wouldn’t have to worry about separate management, separate upgrades or compatibility issues.
What you can do: Take a look at self-protecting storage. Enterprise storage that has data protection, backup and recovery functionality built in isn’t a dream; it’s being considered by more than 70 percent of organizations, according to a Storage Swiss study. Self-protecting storage is also available as a service, which allows us to keep our first wish as a reality -- provisioning as much as you need, and reducing the management burden on IT.
Pay for only one copy of your data.
Take a moment and think about how many copies you have of each piece of data that exists in your system. If you’re like most companies, you have one primary storage copy in your database. You also have a secondary copy for test and development purposes. Then there’s a backup and disaster recovery copy. Then there’s the overall backup (think tapes) of everything, that’s likely stored offline, far away, doubling everything again. When you add it all up, you’re probably in the range of five, six or seven copies of each.
For our ground-up data strategy, we want to pay for one master copy of each piece of data and we want to access it anywhere it’s needed. This should be possible; the amount of your data you actually access on any given day is a small percentage of the data footprint of any enterprise. It makes sense to keep that data physically close, and on relatively expensive media, like flash. The data that’s not accessed frequently? Let’s keep that as far away as possible, stored safely and inexpensively.
What you can do: Put your data into a single network and think about access, not replication. All data is written to the cloud, yet the most accessed data is kept close. Our Smart Tiered Caching™ automatically puts the data you use most right at the edge of your applications, and it gradually moves warm and cold data farther away, to more appropriate storage media -- all still accessible when you need it.
No business is going to let IT press the “red button” on its infrastructure investments, systems and procedures, and start from scratch. By using these three options, you can start thinking beyond the data center today, and make your business much more efficient in the process.