With ever-increasing storage capacity and escalating energy costs, the small to medium business (SMB) sector is looking for solutions to address both the capital and operational costs of storage.
It isn’t just about finding the budget to acquire storage hardware and software, for SMB’s it’s about the expense of managing storage at a time when digital information continues to be generated at breakneck speed.
The solution is green storage systems which focus on cutting storage management costs. While it would be nice to think that adopting green storage strategies is all about being a good corporate citizen, the truth is that it is all about savings. Most green storage solutions revolve around buying and operating less storage which is why it is so appealing to the SMB sector.
Supporting this green movement is the Storage Networking Industry Association (SNIA) which has introduced the Emerald program to certify the energy efficiency of different storage systems.
As part of its plans to develop storage metrics and standards, SNIA ANZ created a Green Storage Committee. SNIA ANZ chairman Craig Scroggie said the committee was formed to develop a coordinated IT industry response to rising energy costs.
Since its formation the committee has launched the Green Storage Power Measurement Specification which includes a Green Storage Taxonomy.
Scroggie said the taxonomy was designed to classify storage systems based on the application environments they support. The application environments are divided into five categories (classes) ranging from small home/office applications (SOHO) to larger enterprise–oriented applications. It covers data protection, component redundancy, serviceability, data access time, and energy consumption.
To help SMBs in the selection of Green storage solutions, Computerworld has created this two part feature which profiles green technologies and the main providers. Part one looks at the latest technologies to address this growing problem while Part two compares Green messaging and strategies by the major storage vendors including Dell, EMC, Fujitsu, Hitachi Data Systems, HP, IBM, NetApp and Oracle-Sun.
Part One: Green technologies and strategies
Tiered storage
Spinning disk drives require energy which generates heat which require cooling systems that use even more energy. Fortunately, not all disk drives consume the same amount of energy.
Fast, high performance disks such as 15,000 revolutions per minute (rpm) consume more energy than 7,200 rpm Serial Advanced Technology Attachment (SATA) disk drives. Solid state-drives (SSDs), while expensive, consume significantly less than their spinning counterparts.
As Gartner analyst Phil Sargeant explains, not all information needs to reside on the most expensive, highest performing, most energy hungry disk drives. Organisations can store a lot of their data on SATA devices. “Using multiple forms of disk drive in a storage infrastructure is a way to manage performance and cost,” he said. “The problem with tiered storage infrastructures is deciding what data resides where, making it potentially more complex to manage. This complexity is disappearing as automated tiering products come to market.”
Thin Provisioning
Storage is often over-allocated. Users may think they require 50-gigabyte data volumes but if they are never filled the storage goes to waste. Thin provisioning is a feature that allows systems to allocate storage on a use and needs basis to ensure there is no wastage.
Data reduction – compression, single instance store and deduplication
Compression, single-instance store (SIS) and deduplication have the potential to cut down the amount of data that is being stored – either primary data or in the backup process. Compression is simply the encoding of data to reduce size and has been around for many years, particularly for data on tape. It is now available for disk-based data.
SIS is a form of compression, usually applied to a large collection of files in a shared data store. Only one instance of a file is retained in the data store, and duplicate instances of the file reference the stored instance.
Data deduplication is yet another form of compression which eliminates redundant sub elements, chunks or blocks.
Over the page: Deleting and archiving data to disk and tape, Cloud storage, Massive Array of Idle Disks, Heterogeneous Storage Virtualization