While there have been many innovations that have become de-facto standards in terms of storage developments, certain much publicized technologies have failed to take off. Due to lack of commercial interest, there is little talk as well. But there is an upside to this. The less hype about these technologies, the less premium you pay for incorporating these in your data center. These storage developments also deliver a lot of value. Here’s a quick overview of these storage technologies that are yet to take off.
Data centers are considered as major energy consumers. For example, when EMC opened its big lab in Bengaluru, the city administration faced issues trying to cater to the facility’s new power demand. A major portion of the lab’s energy consumption is attributed to large storage devices. Huge arrays storing petabytes of information, with thousands of continuous spinning disks are a common sight in such facilities.
Although most of the data on these spindles is rarely accessed, the disks keep spinning due to design. Storage developments such as MAID tackle this challenge of storing data that applications would write once and read occasionally (WORO). An example on this front is SATA storage with petabytes of backup data. Imagine if you could power select disks only when you needed a restore, and the huge power savings.
I believe most companies do not have a green IT strategy. Sometimes it is not even a priority, and hence technologies like MAID haven’t caught their interest. This is similar to the fact that thin provisioning became popular not because of the power it could save, but because it offered to stripe data over multiple disks, in turn delivering better performance.
Useful storage development 2: Data compression appliance for NAS
Storage developments have resulted in technologies such as data compression that aim to reduce the number of spindles required to store your primary data. The compression appliance does this on the fly and transparently. The only challenge is integrating this with your NAS device (coupled with minor changes in the network). Such appliances can reduce existing storage utilization by up to 80%. There are techniques to compress your existing data by routing it once again through the appliance.
Again, compression is viewed as an additional overhead and complication, and hence this storage development hasn’t become popular. With its recent acquisition of Storwize (a leading player in real time storage compression technology), can IBM’s marketing make storage compression appliances a must buy with NAS?
Useful storage development 3: Storage resource management (SRM) software
You might argue that SRM software is present in most organizations, but I still want to feature this storage development here. The reason is a lack of widespread use by storage administrators. The lack of common standards is posing a challenge in developing unified storage management tools.
Every storage vendor differentiates itself by incorporating patented features into its products. This makes it impossible for other vendors to integrate them into a common management framework. In the meanwhile, SNIA continues to drive common standards in the industry via the SMI-S framework. It even tried to create its own software for unified storage management, but gave up after some industry partners pulled out. My personal experience is that none of the SRM software that exists today can truly offer unified storage management.
About the author: Biju Krishnan is a DCT consultant who has worked with IT majors in APAC. His specialties include cloud computing, grid computing, virtualization and storage technologies. He can be reached on [email protected]