There is much debate in the industry over what is meant by so-called 'green software.' Generally, practitioners should start by cleaning up stranded data sets and delete information that should not be retained. This frees up storage space, lowers costs and reduces the risk of a plaintiff's attorney finding a 'smoking gun' in data that should not be retained out of context.
The housecleaning effort starts with a data classification initiative and an effort to migrate inactive data from primary to secondary or tertiary storage.
In addition, there are a number of software tools and products beyond this sound practice that can contribute to lowering energy consumption. The basic concept of the software is to reduce the number of spinning disks, the most consumptive component of a disk array and thus lower energy consumption.
These software components include:
- Thin provisioning - which allows IT managers to 'fool' the application into thinking it has more storage than is actually physically provisioned.
- Storage virtualization - which per se does not necessarily lower power consumption but supports thin provisioning and the facile movement of data between tiers, thereby lowering energy consumption.
- Data de-duplication which reduces redundant data stored.
- MAID - Massive Array of Idle Disk - software that identifies disk drives with inactive data and slows down the spin speed, parks the drive's heads, idles the media or shuts the drive down completely.