Key to Low-Cost Backup/Recovery: Putting Idle Resources to Work
Why is the cost of online storage so expensive when the cost of physical hard drives is falling as capacity rises? Why are some cloud storage vendors shutting down their service?
According to the executives I talked to at Symform, it’s because the pricing model and technology are all wrong.
Think about it: at least half of your hard drive’s capacity is likely empty -- on my desktop it’s closer to 75 percent. The drive just sits there spinning and costing you money. It’s a waste. That’s just the waste in your own organization. Now imagine a storage provider, which must maintain a data center with terabytes to spare, all sitting around unused. The biggest cost isn’t the hardware -- it’s the cost of electricity to run (and cool) drives that aren’t being used to their fullest.
The big benefit of server virtualization is that multiple applications could run on the same server, thus pushing server utilization up and overall costs down. That same principle -- making use of idle resources -- is what’s behind Symform’s storage-in-the-cloud service, which today is announcing a new pricing program that is a radical departure from traditional pay-by-the-terabyte plans.
Symform doesn’t maintain central data centers, which are money sinks. Instead, it puts to use the idle space on its customers’ hard drives. Using a small agent that runs on individual systems in your environment, your data is compressed, encrypted (using AES-256 algorithms), and stored on other systems within Symform’s customer base -- much like Hadoop uses unused CPU capacity among all the systems in a resource pool.
Using Symform’s model, you pay a flat fee based on a combination of number of employees and servers; then it’s an all-you-can-store model, with capacity determined by how much storage space you make available to other Symform customers. In other words, if you can provide 1 TB of data for storage for others, you can store 1 TB of your own data on other systems within the resource pool.
There’s a lot of sophistication behind that agent and Symform’s technology. The agent splits files apart into smaller chunks, which it then geographically disburses around the world (the company calls this geo-spacing), giving you the benefit of geographic decentralization to a much greater degree than most offsite storage providers offer. What if some storage locations are offline? Symform adds redundancy. It claims that only 64 of 96 pieces of a file need to be available for the company to be able to reassemble the pieces and recover a file. As machines fail it automatically repopulates the fragments. This means 33 unrelated machines would have to fail at about the same time for Symform to lose customer data. Compare that to a data center. Security is assured; there is no way for a company to recover your file because (a) it’s encrypted and (b) no single site stores all of the pieces of your file.
Symform executives, including Praerit Garg (the company’s president and co-founder) and Bassam Tabbara (chief technology officer and the other co-founder) are disarmingly honest. Although security and reliability are critical, choosing a storage service, they admit, comes down to price. In that regard, the company has a strong case to make, claiming that its service can save an enterprise up to 50 percent when compared to other cloud storage providers, and in many cases, much more than that. For example, for an enterprise with 100 employees and five servers, Symform says savings of $3,900 to $136,400 are possible, depending on your provider.
According to a company statement, “all corporate laptops and desktops are included in this low price so CIOs don’t have budgets driving the choice of devices and data to protect.” That’s smart.
-- James E. Powell
Editorial Director, ESJ
Posted by Jim Powell on 04/26/2011 at 11:53 AM