Driven to Tiers
Tiered storage is real, but it isn't the ILM of two years' back
May 16, 2007
![Network Computing logo Network Computing logo](https://eu-images.contentstack.com/v3/assets/bltde8121fc52c5c8f3/blt3f3d0318f746b1c2/65a530e4187606040a1d8b8c/placeholder.png?width=1280&auto=webp&quality=95&format=jpg&disable=upscale)
Where has ILM gone? Check the warehouse that's also home to artificial intelligence, midlevel management systems, and the lights-out data center.
Like other mildewed marketing terms, ILM hasn't gone anywhere, because it never arrived. Information lifecycle management was really a castle in the air, a rosy picture painted by vendors, in which data would move efficiently (and apparently without user intervention) from arrival on the network to progressively cheaper storage locations as it aged.
"A lot of vendors wanted a vision of real-time data activities going on, with storage as a sort of stateless entity," says analyst Brad O'Neill of the Taneja Group consultancy.
Pure pipe dream, given the facts of storage life. But like most pipe dreams, it grew from a tiny bit of truth. In this case, the truth was that SATA and SAS were rising to meet the requirements of a cheap but effective alternative to Fibre Channel. At the same time, the level of data companies needed to keep was escalating. (See SAS Wave Breaks Big.) All this led to the adoption of tiered storage, in which data is moved (instead of moving itself) from one platform to another, depending on factors like its age and its accessibility requirements.
"It's an absolute reality that users have deployed multiple tiers of disk storage and classes of data," O'Neill says. And every storage vendor has made its metadata accessible to help users get organized. "It's well in hand," O'Neill says.This isn't to say it's automatic or even streamlined. As O'Neill notes, there are tools for parsing metadata, tools for setting storage management policies, and tools for moving data from one array to another nearly always on a scheduled basis, not automatically. Trouble is, none of these tools are necessarily in the same package — or from the same vendor.
Improved tiered storage is progressing on several fronts. Vendors are talking about tiering data inside the box, instead of buying extra storage. Startups are talking about distributed storage that makes data accessible from any machine in a secure way. (See Njini Participates in FAN Group, and Moonwalk Takes Small Steps.)
But technology alone can't get the job done. Storage managers have to formulate a plan for tiered storage. While many suggestions have been made, it's always a lot of work and involves cooperation, management, and diplomacy. (See What's the Best Way to Set Up Tiered Storage?)
According to George Crump, founder of the Storage Switzerland consultancy, it's often these human issues that stall tiered storage. "It's a big project, and IT staffs are strapped and stretched too thin," he says. There's just not time to tackle a big project that involves a lot of end user buy-in. "It's easier to keep buying more storage," he says. "There are plenty of tools to classify the data, but you can't move someone else's data without them knowing about it."
Tiered storage realities have replaced the ILM visions promoted by vendors. But tiered storage is evolving too, and it's likely we'll see plenty of technology — and technobabble — as it morphs.— Mary Jander, Site Editor, Byte and Switch
Forsythe Solutions Group Inc.
Moonwalk Inc.
Njini Inc.
Taneja Group
You May Also Like
Maximizing cloud potential: Building and operating an effective Cloud Center of Excellence (CCoE)
September 10, 2024Radical Automation of ITSM
September 19, 2024Maximizing Manufacturing Efficiency with Real-Time Production Monitoring
September 25, 2024